AI transcript
0:00:04 Hey, welcome to the Next Wave podcast.
0:00:04 I’m Matt Wolfe.
0:00:06 I’m here with Nathan Lanz.
0:00:09 And today we’re talking about the future of investing.
0:00:15 We’re going to be talking about how you can leverage AI using tools built on top of Crunchbase
0:00:21 to figure out what to invest in, what your sales team should go and focus on, all sorts
0:00:26 of really cool strategies to leverage data and AI around the world of investing.
0:00:32 Nathan, in this episode, had one of the most brilliant ideas I’ve ever heard for a salesperson.
0:00:36 If you’re actually out there trying to sell your product to companies, you need to stick
0:00:40 around because this idea, I think, could totally change the game for your business.
0:00:42 But this is an amazing episode.
0:00:45 It’s with Jagger McConnell, the CEO of Crunchbase.
0:00:47 And I’m not going to say any more.
0:00:49 Let’s just go ahead and jump in and talk to Jagger.
0:00:54 Cutting your sales cycle in half sounds pretty impossible.
0:00:58 But that’s exactly what Sandler Training did with HubSpot.
0:01:03 They used Breeze, HubSpot’s AI tools, to tailor every customer interaction without losing
0:01:04 their personal touch.
0:01:06 And the results were pretty incredible.
0:01:09 Click-through rates jumped 25%.
0:01:13 And get this, qualified leads quadrupled.
0:01:14 Who doesn’t want that?
0:01:18 People spent three times longer on their landing pages.
0:01:18 It’s incredible.
0:01:23 Go to HubSpot.com to see how Breeze can help your business grow.
0:01:29 Thank you so much for joining us, Jagger.
0:01:30 It’s great to have you on the show.
0:01:31 Great to be here.
0:01:35 Well, let’s just dive straight into it and talk a little bit about Crunchbase.
0:01:40 I know that Crunchbase over the last, I don’t know, however long, maybe the last year or so,
0:01:45 has really, really gone deep into the AI world and sort of shifted what Crunchbase is into like
0:01:47 an AI-first platform.
0:01:48 So let’s talk about that.
0:01:51 Like, what is the sort of grand vision with Crunchbase?
0:01:52 Who is it for?
0:01:53 What’s the plans with it?
0:01:54 Let’s just get right into it.
0:01:55 Yeah.
0:01:57 I mean, I could take the whole time just talking about that.
0:02:05 Look, like we’ve been for the last 14, 15, forever years, this sort of historical record
0:02:07 of what’s happened with a company.
0:02:12 And we realized with AI happening and like all this, yes, it’s structured data, but huge
0:02:13 amounts of structured data.
0:02:17 Were there insights that maybe were more interesting than what’s happened in the past with a company?
0:02:21 Can we use AI to figure out what’s going to happen next with a company?
0:02:25 How accurate would it be if we go and try to figure out, you know, what next funding round
0:02:29 is going to happen with what company or what company is going to get acquired next or who’s
0:02:29 going to go public?
0:02:34 So we took all of this data, all of the historical data and combined it with data that no one
0:02:35 else has access to.
0:02:40 Things like our usage data, anonymized, of course, sort of looking at trends of investor flow or
0:02:44 corporate flows and said, what can we learn from all this?
0:02:47 And that’s where we launched this sort of prediction engine.
0:02:52 It just wasn’t possible a couple of years ago, just because given that petabytes of data that
0:02:53 we have sort of behind the scenes here.
0:02:55 So it’s pretty exciting times.
0:02:55 Yeah.
0:02:59 I’m curious, so are the models that you’re using, are they like external models or is
0:03:02 this stuff that Crunchbase is sort of developing internally?
0:03:04 Yeah, it’s sort of a combination.
0:03:08 Obviously, we don’t have the tens of billions of dollars to go and build our own stuff.
0:03:11 So we certainly are leveraging the latest and greatest tech.
0:03:16 As you might imagine, open AI as part of that equation, TensorFlow as part of that equation.
0:03:20 So there’s a lot of these, the sort of latest and greatest tech, but we’re building our own
0:03:21 stuff on top of it, right?
0:03:26 So we’re going and taking a lot of the proprietary stuff that no one else would build because
0:03:30 they’re not Crunchbase and we’re leveraging the pieces that we need to, to pull the right
0:03:30 spaces.
0:03:35 So in other words, how do you go and generate the content of how we present to the user?
0:03:36 Sure.
0:03:37 We’re going to use open AI for that.
0:03:41 But at the end of the day, there’s still a massive machine learning problem that’s hiding
0:03:46 in the Crunchbase side that it takes more than just uploading it up to ChatGPT to sort of
0:03:47 answer for us.
0:03:47 Gotcha.
0:03:52 It’s really fascinating to me, like what sort of data points would you even look at
0:03:54 to sort of predict the future?
0:03:56 Because that’s kind of what Crunchbase is trying to do, right?
0:04:00 It’s kind of trying to predict what companies are going to IPO next, which ones are sort of
0:04:01 acquisition targets.
0:04:02 What are we looking at?
0:04:04 People already kind of use Crunchbase that way, right?
0:04:06 Like in the past, that’s, I think it’s like a genius evolution of the platform.
0:04:10 Because, you know, for context, I lived in San Francisco for 13 years, did some tech startups,
0:04:11 know a bunch of VCs.
0:04:15 And people always check Crunchbase and they’d look at who’s in the round or when did they raise
0:04:17 around or what’s recent news.
0:04:18 Right.
0:04:19 So they’re already using it that way.
0:04:20 And it’s exciting what you guys are doing.
0:04:21 Yeah, it’s really different.
0:04:22 So you’re absolutely right, though.
0:04:25 The use case before was, are they going to raise money soon?
0:04:29 So they would go and look at a profile and say, well, it’s been about 18 months.
0:04:30 It’s probably about time for fundraising.
0:04:33 And that would be the one data point that they’re using.
0:04:37 To answer the question, we’re using thousands of feature vectors to go and figure this out.
0:04:41 So like an easy to understand example would be, is a company going to go and fundraise soon?
0:04:45 So we’ll go and say, sure, is it about time for them to fundraise?
0:04:49 Then we’ll go and look at the entire industry and we’ll say, okay, well, like how long does
0:04:52 it usually take for a company to fundraise in this space for this size?
0:04:57 But then we’ll go a little deeper and we’ll go and say, well, has anyone updated that company
0:04:59 profile recently?
0:05:04 And if the answer is yes, that gives us a little signal that maybe there’s something
0:05:05 going on at the company.
0:05:12 And then another signal might be, has investor flow to this profile changed significantly
0:05:13 compared to the past?
0:05:16 So if there’s more investors looking at the profile, well, why?
0:05:19 What would drive them to go and look at this profile?
0:05:20 And then how are they looking at it?
0:05:23 Are they searching for this organically came upon it?
0:05:24 Or was it a link?
0:05:26 Or did they come from a Gmail account?
0:05:31 So you have this sort of like two-way sort of conversation happening on our profile.
0:05:34 Then are the entrepreneur looking at those same investors who are looking at them?
0:05:36 Well, that’s a little signal, right?
0:05:41 So each one of these steps along the way gets us more and more confident that a funding round
0:05:44 is maybe happening behind the scenes.
0:05:48 And if it is, we can sort of signal at least some level of confidence out to our end users.
0:05:52 And now that’s just one, again, of thousands of these things.
0:05:53 What’s happening in the news?
0:05:56 Just how is traffic to the site in general?
0:06:00 There’s a lot of signals that we can go and combine together to sort of find the right
0:06:04 pattern to say, this is a company that’s going to fundraise soon, as an example.
0:06:05 Yeah, that makes a lot of sense.
0:06:07 I mean, just thinking about it.
0:06:07 Yeah.
0:06:12 If a company’s in there updating their crunch-based profile and adding new information, they’re
0:06:16 probably doing that because they’re expecting, you know, investors or people to be going and
0:06:17 looking at that page.
0:06:18 So that’s interesting.
0:06:22 It can’t be the only signal, but it certainly is one of more signals, right?
0:06:25 And the nice thing about AI is we don’t really need to figure out exactly the right
0:06:27 combination that means it.
0:06:31 It’s just like, it can look at every company that’s ever raised funding and historically look
0:06:35 at all the data we had collected at that moment to say, well, here’s the 16 different paths
0:06:38 of a company that might lead to a funding round.
0:06:39 And that gets me excited, right?
0:06:41 Because this is stuff that no one else can do.
0:06:42 It doesn’t matter which competitor we’re talking about.
0:06:47 They don’t have that 80 million people using our site to go and drive and inform those prediction
0:06:48 decisions.
0:06:49 For sure.
0:06:54 Since you’ve actually pivoted to more sort of AI-based analytics of like figuring out when
0:06:58 there’s going to be an IPO or who’s going to raise or those kinds of things, have you sort
0:07:00 of figured out the accuracy level of it?
0:07:02 Like how accurate has it been so far?
0:07:03 Yeah.
0:07:06 And this is maybe the biggest challenge is that when anyone ever tells you, hey, I’ve got a
0:07:10 prediction engine, you’re like, that’s garbage because every prediction engine ever is
0:07:10 garbage.
0:07:11 So this is very different.
0:07:13 So we do do a lot of analysis.
0:07:16 We can do a lot of back testing and sort of figure out how would we have done.
0:07:20 So you take sort of two thirds of all of our data and you build the models on that.
0:07:25 And then the remaining third you use to sort of test to see would it have done it correctly
0:07:27 had this model existed.
0:07:30 And by using that framework and looking at the kind of companies that we’re trying to make
0:07:33 these predictions against, there’s precision and recall.
0:07:38 And we have a 95% precision on fundraisers and 99% recall.
0:07:44 So in other words, when we make a prediction about a fundraise, it’s 95% correct.
0:07:49 And we make predictions against 99% of the companies that match the criteria we’re looking
0:07:49 for.
0:07:52 So it’s ridiculously high how accurate we are.
0:07:54 Now, that’s the easy answer.
0:07:59 The more complicated answer is as you add time scale to it, it gets much harder.
0:08:04 So who’s going to fundraise tomorrow is an impossible question to answer.
0:08:08 Crunchbase will do a better job than you guessing, but it’s still going to be fundamentally
0:08:08 a guess.
0:08:10 This is actually what I was going to say.
0:08:13 So before we got on here, I checked a few companies that I know when they’re fundraising.
0:08:18 You were super accurate on they were going to be fundraising, a bit off on the timelines.
0:08:19 Right.
0:08:20 And so we put that in there.
0:08:24 So we say, look, here’s what we think, but we’re never going to be 100% confident.
0:08:26 Oh, in the next six months, it’s going to happen.
0:08:30 But we might say like 80% chance it’s going to happen in the next six months because that
0:08:34 timing scale, there’s so many factors that are impossible for us to know unless we’re inside
0:08:37 the brain of the founder to know if they’re going to fundraise.
0:08:42 But that signal, getting back to a question you asked earlier, helps the use case of I’m
0:08:42 an investor.
0:08:46 I’m not looking at this company, but maybe I should be because it looks like Crunchbase
0:08:47 thinks they’re going to be fundraising soon.
0:08:50 Or I’m a large public company.
0:08:51 I want to acquire this company.
0:08:56 They might be going to fundraise soon and I want to get them before they raise that money
0:08:57 or increase their valuation.
0:09:02 There’s a lot of different uses for even just that one fundraising prediction among the almost
0:09:04 dozen different insights and predictions that we have.
0:09:05 Gotcha.
0:09:05 That makes sense.
0:09:10 I mean, I think it would be great to just jump in and sort of get a little demo, a little
0:09:11 tour of what it’s capable of.
0:09:14 And, you know, this is kind of a YouTube first podcast.
0:09:18 We like to be really visual and show what we can, so love to jump in and just sort of
0:09:21 get a sneak peek and give people a little demo of what it can do.
0:09:22 Yeah.
0:09:25 So let me see if I can work the internets here.
0:09:27 So this is the new homepage.
0:09:30 That’s very different than the old homepage of yesteryear.
0:09:34 And you’ll notice that we’re right up front saying sort of the new data is coming in.
0:09:39 I think we’ve got 12 and a half million new predictions in just the last 30 days.
0:09:44 So this engine is constantly running, constantly updating, trying to find not just the next
0:09:48 funding round, but also, you know, what is the next acquisition?
0:09:49 What is the next IPO?
0:09:51 Which companies do we predict to grow?
0:09:55 So we’re really looking at like a lot of different aspects of what a company is.
0:09:58 We’ve got this new sort of AI agent that can help you sort of navigate crunch space, but
0:10:01 you can just type in, you know, blue sky as an example.
0:10:06 And now we’re looking at the blue sky profile and you’re going to see right up front, this
0:10:07 didn’t used to be there.
0:10:10 Now up front, we put some of the biggest predictions up at the top.
0:10:14 So in blue sky’s case, we think it’s probable that they’re going to raise another round of
0:10:14 funding.
0:10:18 We think that it’s likely that they’re going to get acquired at some point, which not a
0:10:19 lot of people are talking about.
0:10:21 And we don’t think blue sky is going to go public.
0:10:23 Certainly there’s no signals that they are.
0:10:27 And then as we scroll down, we sort of took this approach to the profile page of, you know,
0:10:31 what is the stock ticker equivalent of a profile?
0:10:35 So there’s no way we can put valuation day by day over time, which is what a stock ticker
0:10:36 does.
0:10:38 But what is the private company equivalent of that?
0:10:41 So we’ve got these things called heat score and growth score.
0:10:45 And there’s a little definition of what these things are, but like looking at how the company
0:10:50 is interplaying with the public web, how it’s interacting with us, how does it rank among
0:10:55 all the other companies within our corpus of companies that we track that gives us these axes of
0:10:57 data like this heat score and growth score.
0:11:01 And that, again, drives other pieces of the application and even some of our predictions.
0:11:05 So we understand what’s happening and what’s going to happen next to the company.
0:11:09 You can play around with this, you know, and sort of make it do different whiz bang things.
0:11:13 So if you’re a data nerd, you can kind of get into this and all the raw data is available
0:11:14 in the API, of course.
0:11:18 And then as I scroll down, you’re going to find predictions and insights.
0:11:21 So here’s where I can see we predict they’re going to be growing.
0:11:22 We don’t just say it’s going to be growing.
0:11:28 We actually explain with our own words why we think that this company is currently growing.
0:11:31 And if there’s a growth prediction, why we think it’s going to grow in the future.
0:11:34 And then here’s some of these predictions like we were talking about.
0:11:38 Like here’s we think there’s a 37 percent chance that they’re going to go and fundraise in the
0:11:40 next six to 18 months.
0:11:42 There’s a good chance they don’t fundraise, right?
0:11:48 So we’re kind of transparent that these numbers don’t necessarily lead up to 100 percent because
0:11:51 there’s still some percent chance that they’re not going to raise at all.
0:11:53 So we’ll go and put that in there.
0:11:54 Are they going to get acquired?
0:11:55 We give reasons as to why.
0:11:59 In the API, we give all of the detailed reasons, right?
0:12:04 So we actually give percentages and we go and say, here are the drivers that we believe lead
0:12:04 up to this thing.
0:12:09 So as an API user, you can discount things that we think are true that maybe you don’t
0:12:13 want to sort of incorporate into the prediction, or you can just use the prediction score on top
0:12:15 of your own prediction algorithms, which a lot of VCs do.
0:12:19 They sort of use as an input into their own proprietary algorithms.
0:12:21 So this is some of the stuff that we’re doing nowadays.
0:12:26 There’s a lot around like we rock all the news that is happening on a company and sort of summarize
0:12:27 it for you.
0:12:30 So you don’t have to read 30 news articles to figure it out.
0:12:31 We sort of bring it all together.
0:12:34 So there’s just a lot of different pieces that help you understand.
0:12:40 And of course, we still have the funding data, but that’s all sort of just drivers now into
0:12:43 these sort of bigger, meatier questions that we’re trying to answer.
0:12:43 Yeah.
0:12:45 There’s a lot around this.
0:12:49 There’s a lot around, even on that homepage, you know, sort of seeing what’s important,
0:12:52 what’s trending, what’s happening in all the data we’re tracking.
0:12:56 And you can decide, I want to look at these particular predictions for these types of industries
0:13:00 and sort of get a daily feed of all the stuff that’s happening and what we think is going
0:13:03 to happen next in these companies, which is pretty exciting as well.
0:13:05 So lots of interesting use cases.
0:13:10 And again, you can always go and have a conversation with a scout, which is our little sort of dog
0:13:14 fetching thing that will go and sort of do some of the logic stuff that you couldn’t do in
0:13:15 Crunchbase before, right?
0:13:21 Before you couldn’t figure out sort of what’s the business model of this company, or how does
0:13:26 this compare to another company or how do public events impact these particular private
0:13:26 companies?
0:13:30 Now you have a way to have that conversation with Crunchbase programmatically, which is kind
0:13:30 of cool.
0:13:35 Have you found any like questions that have been like really, really valuable, like any sort
0:13:38 of like best practice questions where you’re like, if you ask this, you’re going to get some
0:13:39 really good stuff out of it?
0:13:40 Yeah.
0:13:42 I mean, there’s meaty like policy questions, right?
0:13:46 Like, so we see whenever something happens in the government, you know, people come to Crunchbase
0:13:49 and go type in, how are tariffs going to affect this company?
0:13:53 And it will do a pretty good job of sort of speculating and sort of figuring out what’s
0:13:53 going to happen next.
0:13:55 I think those are some of the interesting ones.
0:13:57 And then just the analysis, right?
0:14:00 It can be hard, like if we go over to AI Search Builder.
0:14:05 So now like you can just natural language in your query and it’s going to go and build sort
0:14:10 of these very complex searches because, you know, you think about multi-joint searches and
0:14:11 how to build those.
0:14:12 It’s always been sort of cumbersome.
0:14:18 Now you can say, show me all the CEOs at companies where they used to work at Salesforce and
0:14:23 then they went to Stanford, you know, like you could type that all in, in a huge run-on
0:14:28 sentence and it will go and show you exactly which companies do it because it builds a joint
0:14:28 for you.
0:14:31 And just that alone is a huge time saver for our users.
0:14:33 I wonder what this is going to do to like startups.
0:14:37 It feels like it’s going to like really increase like the velocity of rounds, like how fast rounds
0:14:40 will close because, you know, you guys are kind of creating like the ultimate like FOMO
0:14:41 machine, right?
0:14:44 People are like, oh my God, look, it’s a hot company.
0:14:45 Crunchbase just told me it’s hot.
0:14:47 I got to get in.
0:14:47 Yeah.
0:14:49 I’m curious if you guys have seen that or like what your thoughts are.
0:14:50 Yeah.
0:14:54 I mean, it’s hard for us to have data on this, but anecdotally we’ve heard that when we go
0:14:59 and signal that is very imminent that a company is fundraising, that they get a lot of inbound
0:15:03 interest from investors because there’s now awareness that it’s happening.
0:15:05 Oh, you notify people or?
0:15:07 We don’t, but people set up their own alerts, right?
0:15:14 I can just say, you know, show me biotech companies that are very likely to fundraise,
0:15:15 right?
0:15:16 I can just make that search.
0:15:19 It’s going to go and, you know, this is a live demo.
0:15:20 We’ll see what actually happens.
0:15:21 But there we go.
0:15:22 So it did.
0:15:25 It said industry is biotechnology, funding predictions here is very likely.
0:15:28 So these are all the companies that are very likely to fundraise soon.
0:15:28 Yeah.
0:15:30 And we can go and create an alert off of this, right?
0:15:35 So you’ve got investors who are subscribers of ours say, when a new company shows up on
0:15:37 this list, shoot me an email, right?
0:15:39 Go and let me know that that’s happened.
0:15:41 And so we don’t need to send them emails.
0:15:44 They’ll get their emails themselves because they’ve set up their alerts the right way.
0:15:45 Yeah, that’s awesome.
0:15:46 So I can do like AI coding or something like that.
0:15:50 And I can just like, as soon as you guys have a new prediction in that category, you guys
0:15:51 will email me or something.
0:15:52 That’s amazing.
0:15:52 Yeah.
0:15:54 And that’s just one of the predictions.
0:15:57 Another very common one is I’m looking for these sorts of companies to acquire.
0:16:02 Let me know when a company of this size, no bigger than Series C, I don’t want them who
0:16:05 raised more than a hundred million dollars, whatever it is, when a new one shows up is
0:16:07 very likely to get acquired.
0:16:10 Let me know because I’m in the space of acquiring those companies.
0:16:13 So a lot of corp dev departments get excited about that.
0:16:16 Hey, we’ll be right back to the show.
0:16:19 But first I want to tell you about another podcast I know you’re going to love.
0:16:21 It’s called Marketing Against the Grain.
0:16:24 It’s hosted by Kip Bodner and Kieran Flanagan.
0:16:26 And it’s brought to you by the HubSpot Podcast Network.
0:16:29 The Audio Destination for Business Professionals.
0:16:33 If you want to know what’s happening now in marketing, especially how to use AI marketing,
0:16:35 this is the podcast for you.
0:16:41 Kip and Kieran share their marketing expertise, unfiltered in the details, the truth, and like
0:16:42 nobody else will tell it to you.
0:16:49 They recently had a great episode called Using Chat TVT 03 to Plan Our 2025 Marketing Campaign.
0:16:56 It was full of like actual insights as well as just things I had not thought of about how
0:16:57 to apply AI to marketing.
0:16:59 I highly suggest you check it out.
0:17:02 Listen to Marketing Against the Grain wherever you get your podcasts.
0:17:10 Now, are all the companies that are in here, are they all like self-added or is all the data
0:17:11 sort of pulled by Crunchbase?
0:17:15 Sort of absolutely is the short answer.
0:17:20 We get data from our users, but in 2014, it was 100% from our users.
0:17:22 Today, it’s about 5%.
0:17:23 So we’ve sort of transitioned.
0:17:28 But the brand belief is that still, like if you ask our users, where do we get our data from?
0:17:29 They’re mostly, we’ll say it’s user-generated content.
0:17:34 But really, that’s only for the smallest companies, ones that haven’t had a news article yet.
0:17:35 No one knows they exist.
0:17:37 They go and put themselves in.
0:17:39 They sort of announce themselves on Crunchbase.
0:17:43 But the reality is we invest tens of millions of dollars now every year into getting data from
0:17:44 a lot of different sources.
0:17:47 We have 5,000 partnerships of data coming in.
0:17:49 We, of course, go look at government filings.
0:17:52 We’ve got lots of data partnerships to go and give us seed data.
0:17:55 And then we have our own AI systems that go out and find the data.
0:18:00 So if we don’t hear about a company that isn’t in our data set, we very actively go and fill
0:18:03 out the profile as best as we can, assuming it fits a certain set of criteria.
0:18:08 And that plus the engagement data, plus the generated data, right?
0:18:12 Like the biggest source of data now is Crunchbase generating its own data on the data that we have.
0:18:17 That is that huge, huge, huge data set that we’ve collected over the years now.
0:18:19 But not to devalue the user-generated data.
0:18:20 That’s still very important stuff.
0:18:21 Right.
0:18:26 It just seems like it could be such a really good discovery engine for small new startups.
0:18:29 You want to make sure you’re in Crunchbase because then you’re sort of in that algorithm.
0:18:33 You’re in that system where now people might discover you if they’re looking for,
0:18:36 you know, small startups in X niche, right?
0:18:38 So that’s why I was curious.
0:18:43 Can companies just go and sort of input their data in there to make sure that they get discovered
0:18:44 when people are making those sort of queries?
0:18:46 Yeah, they absolutely can.
0:18:50 And we do a bad job of this, like giving reasons why you should, because they ask,
0:18:53 you get discovered by VCs who wouldn’t normally have found you.
0:18:54 And if you’re like, I’m a bootstrap company.
0:18:55 I never want venture funding.
0:18:56 I don’t need to be in Crunchbase.
0:19:02 Job seekers is like a good solid 10% of our users are going and researching your company
0:19:03 to see if it’s a company they want to work at.
0:19:07 If you’re not in there and it’s not up to date, they’re like, this isn’t something I want
0:19:09 to go and participate in because they couldn’t grok your website or they couldn’t find your
0:19:11 website or whatever the case happens to be.
0:19:16 There’s a good chance that Crunchbase’s profile comes up higher than the website of the smaller
0:19:16 company.
0:19:21 So it’s usually a good idea to have that data correct in Crunchbase.
0:19:26 Now, it used to be traditionally like VC-backed companies that were mostly on Crunchbase, right?
0:19:27 Is that still the case?
0:19:29 Are there companies that are private companies that are non-VC?
0:19:31 Like you just said, you suggested that they should do that.
0:19:34 But also like, I can see this working for even like public companies, like just like a
0:19:39 general tool to help me guide my investments in companies in general.
0:19:42 Yeah, another brand challenge that we have.
0:19:44 There’s about 300,000 companies that have ever received funding.
0:19:46 We have about 4 million companies in Crunchbase.
0:19:51 So that math is very different because there’s a lot of companies that never get to funding.
0:19:52 There are a lot of bootstrap companies.
0:19:56 There are a lot of companies that will do acquisitions that we had not seen before that will go and
0:19:57 put in Crunchbase.
0:19:59 And then every public company is in Crunchbase.
0:20:01 So we do track all those things.
0:20:05 Now, are you going to get all the information you want to about a public company?
0:20:05 Probably not.
0:20:09 Like you should go to Yahoo Finance or Google Finance or whatever the latest degree this is,
0:20:12 because there’s so much public data out there that we just aren’t going to track because
0:20:13 it doesn’t apply to private companies.
0:20:16 But for the data, it does overlap for sure.
0:20:22 So for instance, when like Virgin America back in the day, that was a hot startup airline.
0:20:24 So it was definitely tracked in Crunchbase.
0:20:27 Alaska Airlines was not tracked in Crunchbase.
0:20:31 But when Alaska Airlines bought Virgin, then we started tracking Alaska Airlines.
0:20:36 So like this ecosystem and sort of the spider web of how companies are interconnected with one
0:20:38 another expands, expands, expands, and we’ll just keep adding the companies.
0:20:43 But we’re not going to put in, you know, Joe’s Pizza Shop on the corner because it probably
0:20:46 isn’t relevant to our community unless it’s got some cool tech.
0:20:50 Let’s say you’re not a VC or an angel investor or something like that.
0:20:54 Do you see value in using Crunchbase for just general people that are interested in investing?
0:20:59 What’s the value of Crunchbase to the people that can’t write big checks and get into like
0:20:59 early startups?
0:21:04 Actually, it’s a minority part of our users are actually VCs who have funding that want
0:21:04 to go invest.
0:21:08 So that is an important use case to us, but it’s certainly not the biggest, you know, you’ll
0:21:09 see use cases across sales.
0:21:13 Let me go and find the companies that are going to have money soon.
0:21:15 I’m going to go and start a sales cycle with them.
0:21:16 It’s a good time.
0:21:19 And it’s better than waiting for the fundraise to happen too, right?
0:21:20 Because everyone knows when the fundraise happens.
0:21:25 If you can know it six months, a year in advance, maybe you can sort of get entrenched earlier
0:21:26 than that.
0:21:28 So that’s a pretty big use case for us.
0:21:30 I mentioned CorpDev, right?
0:21:31 Anyone who’s buying companies, that’s important.
0:21:35 On the self-service side, we do have job seekers who go and are paying us and say, I want to
0:21:40 find hot-level companies in my area because I want to be there at their early stage, right?
0:21:42 So you’ve got that use case pretty heavy.
0:21:48 A lot of researchers and analysts from the largest consulting firms all the way down to students
0:21:50 who are trying to figure out sort of some interesting trends.
0:21:53 So you’ve got a lot of that sort of use case lurking in Crunchbase.
0:21:58 Those are some of the big ones, but it always is surprising to me to see that pie chart of
0:22:02 all the use cases of Crunchbase and saying, first, how the heck are we going to build a product
0:22:04 for all these different use cases?
0:22:07 But more generally, it’s really exciting to sort of see how many different people have
0:22:10 different uses for that private company data.
0:22:13 I was curious, I know there’s Crunch Fund, which has no connection.
0:22:13 No connection.
0:22:17 But it feels like you guys, all this data, you know which companies are going to raise money.
0:22:20 Somebody should be like piggybacking off of that and like making a lot of money off this.
0:22:23 Like you guys have all the data, find out the right companies, get into them, even like
0:22:24 small allocations.
0:22:25 Yeah.
0:22:28 I mean, we’ve toyed with the idea of doing it ourselves, honestly.
0:22:32 Like one of the ideas that we have lurking out there, it’s on the roadmap for not this
0:22:37 year, but maybe some future year, is I showed you a profile that has our equivalent of a stock
0:22:37 ticker, right?
0:22:39 Which is this growth and heat score.
0:22:41 What if we aggregate those, right?
0:22:43 What if we do that across entire industries?
0:22:47 So now you’ve got the AI heat score and growth score over time.
0:22:48 We’re using our stuff.
0:22:51 We’re predicting, is this moving up or down in the right direction?
0:22:58 What if we worked with maybe a secondary provider and create a little index for those companies?
0:23:01 So you could invest in some subset, right?
0:23:04 And letting the retail investor maybe start playing around with this.
0:23:06 It’s still an idea.
0:23:10 We’ve sort of had some early chats about it, but you sort of stumble into a lot of regulatory
0:23:13 issues pretty quickly, but you know, you’ll never know.
0:23:15 It could be distracting for you guys.
0:23:16 It’s not part of your core mission.
0:23:18 And the ROI is way out there, right?
0:23:20 So like, yeah, we could start a fund.
0:23:21 It’s a 10-year thing.
0:23:24 We’re a little bit more focused on the present than that far out.
0:23:24 Right.
0:23:25 But so intriguing.
0:23:27 Yeah.
0:23:31 I’m kind of curious about like what the sort of future of investing looks like.
0:23:35 And I don’t necessarily know the exact question to ask because I don’t know what I don’t
0:23:37 know when it comes to like AI and investing.
0:23:43 But I’m trying to figure out like if the general population has access to the information and
0:23:45 like what’s likely to sell next.
0:23:49 And, you know, this information is, for lack of a better term, democratized, right?
0:23:53 I’m curious about what the world looks like as we move closer and closer to that reality.
0:23:54 And I’m just curious if you have thoughts on that.
0:23:55 Yeah.
0:23:58 I think there’s a lot of potential disruption across a lot of different industries.
0:23:59 I think data is included in that.
0:24:02 And that’s honestly why we moved the way we did.
0:24:06 Like I would argue funding data is already commoditized, right?
0:24:08 So like, yes, we think we have the best.
0:24:09 Yes, there’s a lot.
0:24:14 But if we just kept resting on those laurels, like that company goes out of business when
0:24:19 all of the data gets absorbed into our LLM masters, you know, what else is there?
0:24:20 Once it goes in, it’s not going to come back out.
0:24:23 So facts are a dangerous business to be in.
0:24:27 So speculation and predictions is at least dynamic and changing.
0:24:29 And I think a lot of data companies are going to be thinking like that.
0:24:34 If you deal, even if it’s hard to get facts or you lose it all, if someone takes it all
0:24:36 and uploads it, is your business in trouble or not?
0:24:38 That’s the question I think everyone should be asking.
0:24:43 Now to the broader question of how does it affect the entire industry?
0:24:49 You know, I think generic tools that do not have proprietary pieces of the story are going
0:24:50 to be very, very hard.
0:24:52 The first mover to manage isn’t going to be a thing.
0:24:54 It’s always going to be a challenge.
0:24:58 So how do you go and build a thing that is uniquely yours?
0:25:02 I don’t know of how to do that unless you are building the foundational models, right?
0:25:07 Like you are the open AI and everyone’s on top of you, or you’ve got something that truly
0:25:10 changes all the time and is only available to you.
0:25:13 And it’s critical to people’s business workflows.
0:25:15 I don’t know how else you survive.
0:25:20 You know, I go to a lot of these AI conferences and I see a lot of people like building AI on top
0:25:21 of their product.
0:25:25 But I really think there’s a day, I mean, Replit’s almost there and Cursor’s getting
0:25:30 there as well, where you can just describe the thing and it’s going to build as good as the
0:25:30 thing that you have.
0:25:36 As long as you’ve got a good product manager with a good set of ideas, that tech is not
0:25:36 that far away.
0:25:38 And then you just fast forward five years.
0:25:41 It’s going to suggest things to do to beat the competition, right?
0:25:43 It’s going to code it for you and build a feature.
0:25:45 It’s going to raise the money for you too, or it’s going to…
0:25:46 Yeah, I mean, maybe.
0:25:49 So all of that becomes commoditized, essentially.
0:25:50 So there is…
0:25:54 Now it’s just like companies are going to go back to building their own internal tools
0:25:59 because their customers bespoke for what they need rather than trying to fit into someone
0:26:00 else’s package.
0:26:05 So in those scenarios, you’ve got to bring some other value other than that into the equation.
0:26:08 And that’s why being a data company that has some stuff that no one else has feels pretty
0:26:09 good for that long-term vision.
0:26:10 I’m biased.
0:26:13 You also mentioned that your Crunchbase has an API as well.
0:26:18 So I mean, that API can sort of work into your own sort of proprietary stuff.
0:26:21 I’m sure there’ll be people out there that figure out some really good prompts and really
0:26:25 good data points to look at to sort of make their own predictions and then not want to share
0:26:29 them with the world because that’s their sort of little secret sauce that they figured out.
0:26:30 Totally.
0:26:33 Almost every major VC now has their own data science team.
0:26:37 And we have conversations with them saying, hey, how would that API feed into your team?
0:26:39 We don’t want to replace that team.
0:26:40 We’re just going to supplement them.
0:26:43 That API also is used in lots of different applications.
0:26:44 So a lot of people don’t know this.
0:26:49 Like we power most private company data you’re going to find on any site out there.
0:26:56 So when you’re using, you know, a major financial tool or a CRM tool, our data often is in there
0:26:59 because we have a partnership with them, which is code for the same.
0:27:02 They’re a customer of ours who have taken our API and putting it in their product.
0:27:06 And that was a strategic decision because we want to make sure people don’t build competing
0:27:08 databases to ours.
0:27:13 But also it’s really exciting to sort of see the innovation and like see how people incorporate
0:27:15 our data into their own tools and make them successful.
0:27:16 For sure.
0:27:21 Nathan, I’m curious, like the sort of investing, you know, venture capital world is sort of more
0:27:24 the world that you’ve played in and less the world that I’ve played in.
0:27:28 So I’m curious if there’s any ground that we haven’t covered that you want to make sure
0:27:28 we cover.
0:27:31 I mean, just still, like I said, the thing that keeps going through my mind is like, okay,
0:27:34 if everyone has access to this data, then you have to find the outliers.
0:27:38 Somebody probably should be making like a really great newsletter on top of this data
0:27:42 and like giving their own opinion about what this means and what they’re saying, you know,
0:27:43 beyond just the data.
0:27:44 Yeah, that’s true.
0:27:44 That’s my take.
0:27:45 That’s true.
0:27:50 I mean, we are thinking about how to build sort of value reports on top of the data.
0:27:54 So it’s not just like reporting on the raw data, but can we put a narrative to the data?
0:27:54 Right.
0:27:56 So that’s one angle that we’re thinking about.
0:27:59 Another angle we’re thinking about is how do we take what’s happening in public markets
0:28:02 and interlock it with private market data?
0:28:08 So for instance, if, you know, a certain set of companies, let’s say biotech companies are
0:28:14 suddenly their stock market is tanking and they’re doing really poorly, how does that affect the
0:28:15 VC market?
0:28:17 How does that affect our predictions, right?
0:28:19 Like there’s these external influences.
0:28:24 And then can we report on that and say, look, based on what we’re seeing in these sort of
0:28:29 external sort of public markets, we predict there’s a cooling happening on this side of
0:28:32 the house and those dollars are going to get redirected to, you know, whatever, whatever
0:28:34 the other hot trend is at the time.
0:28:38 Like we can start making more of a commentary on what’s happening in the world than just leave
0:28:40 it to others to interpret.
0:28:43 There’s sort of countless opportunities lurking out there, especially when people are willing
0:28:48 to pay for sort of like these very deep analyses of what’s happening in a certain industry
0:28:50 or a certain micro-league industry.
0:28:54 I do think out of the use cases you mentioned, like for me, the sales one is super interesting.
0:28:59 Like if I could talk with AI and like, okay, I’ve got a marketing agency and maybe, you know,
0:29:00 you could even get really detailed.
0:29:04 Like I went to Stanford, maybe like look up startups that they went to Stanford so I can
0:29:05 like bond over that.
0:29:07 Like if you could get like really detailed like that, right?
0:29:12 And then reach out like, hey, we both went to Stanford, we both went wherever, and then
0:29:12 start conversation.
0:29:15 I think that could be like super powerful for a lot of people.
0:29:15 Yeah.
0:29:18 We’re trying to change the definition of what an ICP is, right?
0:29:21 Look, here’s exactly who I always sell to successfully.
0:29:22 And we also went to school together.
0:29:24 Those are still historical facts.
0:29:29 So the thing we’re trying to change in people’s minds are, is it the right time for you to talk
0:29:29 to them?
0:29:29 Cool.
0:29:33 Here’s this list of companies, which is the right one to talk to you right now for this
0:29:35 particular accounting executive, right?
0:29:38 So this accounting executive historically has been great at selling these sorts of deals.
0:29:42 Here’s companies that match the kind of companies that they would successfully sell to.
0:29:45 But here’s ones that we are predicting are going to grow, growing quickly.
0:29:47 They’re probably fundraising in the next 18 months.
0:29:51 They’re not going public and they’re not going to get acquired because that would distract the
0:29:52 sales cycle.
0:29:54 So you can kind of like tee up.
0:29:57 This is the right time for this ICP is the one you want to talk to you right now.
0:30:02 And I think that immediacy and urgency, hopefully we’ll start changing with the definition
0:30:05 a little bit of who we talk to you next as a sales team.
0:30:05 Yeah.
0:30:06 That’s worth a lot.
0:30:07 Hope so.
0:30:08 We’ll find out.
0:30:14 Now, can you use Crunchbase to sort of discover emerging markets as a whole?
0:30:18 Like, obviously you can look at in a specific market and find the companies that are sort
0:30:21 of, you know, making moves in those industries.
0:30:27 But can you find like, you know, you mentioned biotech and, you know, had you known like what
0:30:30 AI was going to do over the last, you know, six or seven years?
0:30:35 Like, are there any ways to sort of see that stuff coming a little bit sooner?
0:30:35 Yeah.
0:30:39 It is one of the harder to find features in Crunchbase, honestly.
0:30:41 So, so it’s really possible to do.
0:30:46 There’s things called hubs and not many people know what even the hub is, but we basically
0:30:50 took every major piece of metadata that has data.
0:30:51 So for instance, like industries.
0:30:54 So we have every industry that’s one access.
0:30:56 Geo is another access.
0:31:01 There’s gender, there’s founder, there’s stage of company.
0:31:02 How much they raise.
0:31:03 And we basically intermixed up.
0:31:04 So we made these pages.
0:31:09 So somewhere on Crunchbase, there is a sort of female founder in crypto in Europe.
0:31:11 There’s a page for that.
0:31:13 And we originally did it just for SEO reasons.
0:31:15 And it collects all the data.
0:31:16 So it’s like, here are all the latest funding rounds.
0:31:18 Here are all the people.
0:31:19 Here are the companies.
0:31:22 And they’re all ranked by which ones are trending the most in Crunchbase.
0:31:24 So you take that.
0:31:26 So every single thing in Crunchbase has a rank.
0:31:28 So hubs have ranks.
0:31:34 So which is the hottest one right now, today, based on what’s happening in Crunchbase?
0:31:37 There’s a first, second, third, fourth, all the way down for every single combination of
0:31:41 the, I don’t even know, 100,000 of these different hub pages that we’ve created.
0:31:44 Then you can go and say, well, which ones are trending?
0:31:48 So if you start looking at which ones are trending and which ones had the low rank that are trending
0:31:52 upwards quickly, that’s where you get to see which of those combinations is the hottest.
0:31:55 And you’ll find some really just fascinating things lurking in there.
0:32:00 Some of it’s going to be weird, you know, but like Taiwan artificial intelligence companies
0:32:01 are like a hot thing right now.
0:32:02 I love that.
0:32:02 I used to live in Taiwan.
0:32:03 It’s great.
0:32:04 Okay.
0:32:04 Yeah.
0:32:08 So there’s these little pieces of, that you wouldn’t normally otherwise know.
0:32:12 And I think if you were a savvy investor who was really trying to figure out what is an
0:32:17 emerging trend, or even just again, an analyst or even a journalist who also, they also use
0:32:20 Crunchbase, you can find some interesting things lurking in hub pages.
0:32:23 I think it’s a sleeper feature that we have.
0:32:24 Very cool.
0:32:26 Well, this has been an absolutely fascinating conversation.
0:32:30 I don’t want to be like a salesperson for you, but I’m actually a subscriber of Crunchbase.
0:32:34 I actually do have a subscription and I get in there and I play around with the data
0:32:35 from time to time.
0:32:40 So like I’m actually a user and it sounds like Nathan’s fairly excited as well.
0:32:42 I used to be a user a long time ago.
0:32:43 I had no idea what you guys were doing now.
0:32:44 I checked the website.
0:32:45 I’m like, oh, this is awesome.
0:32:46 It’s different.
0:32:47 It’s definitely different than it used to be.
0:32:48 Very different.
0:32:49 Awesome.
0:32:53 But no, this has been a great conversation and I really appreciate you taking the time
0:32:55 to hang out and give us the demo and everything like that.
0:32:57 Obviously Crunchbase is the place to go.
0:32:58 Crunchbase.com.
0:33:02 If anybody listening wants to go check it out, is there anything else that they should
0:33:02 know?
0:33:05 Any other places they can maybe follow along with you?
0:33:08 Anything like that that you want to shout out before we wrap it up?
0:33:08 Yeah.
0:33:09 I mean, follow Crunchbase on LinkedIn.
0:33:12 I think that’s probably our top channel of sort of sharing stuff out.
0:33:16 And you can follow me on LinkedIn as well because I usually leak roadmap stuff.
0:33:20 So if you want to see what’s coming before it does, my product team hates it.
0:33:22 But I usually will post stuff about what’s coming soon.
0:33:23 Cool.
0:33:24 Awesome, Jagger.
0:33:24 This has been great.
0:33:27 Thank you so much for hanging out with us today.
0:33:32 And for anybody listening, if you like content like this, make sure you like this video
0:33:34 and subscribe wherever you listen to podcasts.
0:33:35 And thank you so much for tuning in.
0:33:37 Hopefully we’ll see you in the next one.
0:33:37 Thank you.
0:33:38 Awesome.
0:33:38 Thanks.
0:00:04 I’m Matt Wolfe.
0:00:06 I’m here with Nathan Lanz.
0:00:09 And today we’re talking about the future of investing.
0:00:15 We’re going to be talking about how you can leverage AI using tools built on top of Crunchbase
0:00:21 to figure out what to invest in, what your sales team should go and focus on, all sorts
0:00:26 of really cool strategies to leverage data and AI around the world of investing.
0:00:32 Nathan, in this episode, had one of the most brilliant ideas I’ve ever heard for a salesperson.
0:00:36 If you’re actually out there trying to sell your product to companies, you need to stick
0:00:40 around because this idea, I think, could totally change the game for your business.
0:00:42 But this is an amazing episode.
0:00:45 It’s with Jagger McConnell, the CEO of Crunchbase.
0:00:47 And I’m not going to say any more.
0:00:49 Let’s just go ahead and jump in and talk to Jagger.
0:00:54 Cutting your sales cycle in half sounds pretty impossible.
0:00:58 But that’s exactly what Sandler Training did with HubSpot.
0:01:03 They used Breeze, HubSpot’s AI tools, to tailor every customer interaction without losing
0:01:04 their personal touch.
0:01:06 And the results were pretty incredible.
0:01:09 Click-through rates jumped 25%.
0:01:13 And get this, qualified leads quadrupled.
0:01:14 Who doesn’t want that?
0:01:18 People spent three times longer on their landing pages.
0:01:18 It’s incredible.
0:01:23 Go to HubSpot.com to see how Breeze can help your business grow.
0:01:29 Thank you so much for joining us, Jagger.
0:01:30 It’s great to have you on the show.
0:01:31 Great to be here.
0:01:35 Well, let’s just dive straight into it and talk a little bit about Crunchbase.
0:01:40 I know that Crunchbase over the last, I don’t know, however long, maybe the last year or so,
0:01:45 has really, really gone deep into the AI world and sort of shifted what Crunchbase is into like
0:01:47 an AI-first platform.
0:01:48 So let’s talk about that.
0:01:51 Like, what is the sort of grand vision with Crunchbase?
0:01:52 Who is it for?
0:01:53 What’s the plans with it?
0:01:54 Let’s just get right into it.
0:01:55 Yeah.
0:01:57 I mean, I could take the whole time just talking about that.
0:02:05 Look, like we’ve been for the last 14, 15, forever years, this sort of historical record
0:02:07 of what’s happened with a company.
0:02:12 And we realized with AI happening and like all this, yes, it’s structured data, but huge
0:02:13 amounts of structured data.
0:02:17 Were there insights that maybe were more interesting than what’s happened in the past with a company?
0:02:21 Can we use AI to figure out what’s going to happen next with a company?
0:02:25 How accurate would it be if we go and try to figure out, you know, what next funding round
0:02:29 is going to happen with what company or what company is going to get acquired next or who’s
0:02:29 going to go public?
0:02:34 So we took all of this data, all of the historical data and combined it with data that no one
0:02:35 else has access to.
0:02:40 Things like our usage data, anonymized, of course, sort of looking at trends of investor flow or
0:02:44 corporate flows and said, what can we learn from all this?
0:02:47 And that’s where we launched this sort of prediction engine.
0:02:52 It just wasn’t possible a couple of years ago, just because given that petabytes of data that
0:02:53 we have sort of behind the scenes here.
0:02:55 So it’s pretty exciting times.
0:02:55 Yeah.
0:02:59 I’m curious, so are the models that you’re using, are they like external models or is
0:03:02 this stuff that Crunchbase is sort of developing internally?
0:03:04 Yeah, it’s sort of a combination.
0:03:08 Obviously, we don’t have the tens of billions of dollars to go and build our own stuff.
0:03:11 So we certainly are leveraging the latest and greatest tech.
0:03:16 As you might imagine, open AI as part of that equation, TensorFlow as part of that equation.
0:03:20 So there’s a lot of these, the sort of latest and greatest tech, but we’re building our own
0:03:21 stuff on top of it, right?
0:03:26 So we’re going and taking a lot of the proprietary stuff that no one else would build because
0:03:30 they’re not Crunchbase and we’re leveraging the pieces that we need to, to pull the right
0:03:30 spaces.
0:03:35 So in other words, how do you go and generate the content of how we present to the user?
0:03:36 Sure.
0:03:37 We’re going to use open AI for that.
0:03:41 But at the end of the day, there’s still a massive machine learning problem that’s hiding
0:03:46 in the Crunchbase side that it takes more than just uploading it up to ChatGPT to sort of
0:03:47 answer for us.
0:03:47 Gotcha.
0:03:52 It’s really fascinating to me, like what sort of data points would you even look at
0:03:54 to sort of predict the future?
0:03:56 Because that’s kind of what Crunchbase is trying to do, right?
0:04:00 It’s kind of trying to predict what companies are going to IPO next, which ones are sort of
0:04:01 acquisition targets.
0:04:02 What are we looking at?
0:04:04 People already kind of use Crunchbase that way, right?
0:04:06 Like in the past, that’s, I think it’s like a genius evolution of the platform.
0:04:10 Because, you know, for context, I lived in San Francisco for 13 years, did some tech startups,
0:04:11 know a bunch of VCs.
0:04:15 And people always check Crunchbase and they’d look at who’s in the round or when did they raise
0:04:17 around or what’s recent news.
0:04:18 Right.
0:04:19 So they’re already using it that way.
0:04:20 And it’s exciting what you guys are doing.
0:04:21 Yeah, it’s really different.
0:04:22 So you’re absolutely right, though.
0:04:25 The use case before was, are they going to raise money soon?
0:04:29 So they would go and look at a profile and say, well, it’s been about 18 months.
0:04:30 It’s probably about time for fundraising.
0:04:33 And that would be the one data point that they’re using.
0:04:37 To answer the question, we’re using thousands of feature vectors to go and figure this out.
0:04:41 So like an easy to understand example would be, is a company going to go and fundraise soon?
0:04:45 So we’ll go and say, sure, is it about time for them to fundraise?
0:04:49 Then we’ll go and look at the entire industry and we’ll say, okay, well, like how long does
0:04:52 it usually take for a company to fundraise in this space for this size?
0:04:57 But then we’ll go a little deeper and we’ll go and say, well, has anyone updated that company
0:04:59 profile recently?
0:05:04 And if the answer is yes, that gives us a little signal that maybe there’s something
0:05:05 going on at the company.
0:05:12 And then another signal might be, has investor flow to this profile changed significantly
0:05:13 compared to the past?
0:05:16 So if there’s more investors looking at the profile, well, why?
0:05:19 What would drive them to go and look at this profile?
0:05:20 And then how are they looking at it?
0:05:23 Are they searching for this organically came upon it?
0:05:24 Or was it a link?
0:05:26 Or did they come from a Gmail account?
0:05:31 So you have this sort of like two-way sort of conversation happening on our profile.
0:05:34 Then are the entrepreneur looking at those same investors who are looking at them?
0:05:36 Well, that’s a little signal, right?
0:05:41 So each one of these steps along the way gets us more and more confident that a funding round
0:05:44 is maybe happening behind the scenes.
0:05:48 And if it is, we can sort of signal at least some level of confidence out to our end users.
0:05:52 And now that’s just one, again, of thousands of these things.
0:05:53 What’s happening in the news?
0:05:56 Just how is traffic to the site in general?
0:06:00 There’s a lot of signals that we can go and combine together to sort of find the right
0:06:04 pattern to say, this is a company that’s going to fundraise soon, as an example.
0:06:05 Yeah, that makes a lot of sense.
0:06:07 I mean, just thinking about it.
0:06:07 Yeah.
0:06:12 If a company’s in there updating their crunch-based profile and adding new information, they’re
0:06:16 probably doing that because they’re expecting, you know, investors or people to be going and
0:06:17 looking at that page.
0:06:18 So that’s interesting.
0:06:22 It can’t be the only signal, but it certainly is one of more signals, right?
0:06:25 And the nice thing about AI is we don’t really need to figure out exactly the right
0:06:27 combination that means it.
0:06:31 It’s just like, it can look at every company that’s ever raised funding and historically look
0:06:35 at all the data we had collected at that moment to say, well, here’s the 16 different paths
0:06:38 of a company that might lead to a funding round.
0:06:39 And that gets me excited, right?
0:06:41 Because this is stuff that no one else can do.
0:06:42 It doesn’t matter which competitor we’re talking about.
0:06:47 They don’t have that 80 million people using our site to go and drive and inform those prediction
0:06:48 decisions.
0:06:49 For sure.
0:06:54 Since you’ve actually pivoted to more sort of AI-based analytics of like figuring out when
0:06:58 there’s going to be an IPO or who’s going to raise or those kinds of things, have you sort
0:07:00 of figured out the accuracy level of it?
0:07:02 Like how accurate has it been so far?
0:07:03 Yeah.
0:07:06 And this is maybe the biggest challenge is that when anyone ever tells you, hey, I’ve got a
0:07:10 prediction engine, you’re like, that’s garbage because every prediction engine ever is
0:07:10 garbage.
0:07:11 So this is very different.
0:07:13 So we do do a lot of analysis.
0:07:16 We can do a lot of back testing and sort of figure out how would we have done.
0:07:20 So you take sort of two thirds of all of our data and you build the models on that.
0:07:25 And then the remaining third you use to sort of test to see would it have done it correctly
0:07:27 had this model existed.
0:07:30 And by using that framework and looking at the kind of companies that we’re trying to make
0:07:33 these predictions against, there’s precision and recall.
0:07:38 And we have a 95% precision on fundraisers and 99% recall.
0:07:44 So in other words, when we make a prediction about a fundraise, it’s 95% correct.
0:07:49 And we make predictions against 99% of the companies that match the criteria we’re looking
0:07:49 for.
0:07:52 So it’s ridiculously high how accurate we are.
0:07:54 Now, that’s the easy answer.
0:07:59 The more complicated answer is as you add time scale to it, it gets much harder.
0:08:04 So who’s going to fundraise tomorrow is an impossible question to answer.
0:08:08 Crunchbase will do a better job than you guessing, but it’s still going to be fundamentally
0:08:08 a guess.
0:08:10 This is actually what I was going to say.
0:08:13 So before we got on here, I checked a few companies that I know when they’re fundraising.
0:08:18 You were super accurate on they were going to be fundraising, a bit off on the timelines.
0:08:19 Right.
0:08:20 And so we put that in there.
0:08:24 So we say, look, here’s what we think, but we’re never going to be 100% confident.
0:08:26 Oh, in the next six months, it’s going to happen.
0:08:30 But we might say like 80% chance it’s going to happen in the next six months because that
0:08:34 timing scale, there’s so many factors that are impossible for us to know unless we’re inside
0:08:37 the brain of the founder to know if they’re going to fundraise.
0:08:42 But that signal, getting back to a question you asked earlier, helps the use case of I’m
0:08:42 an investor.
0:08:46 I’m not looking at this company, but maybe I should be because it looks like Crunchbase
0:08:47 thinks they’re going to be fundraising soon.
0:08:50 Or I’m a large public company.
0:08:51 I want to acquire this company.
0:08:56 They might be going to fundraise soon and I want to get them before they raise that money
0:08:57 or increase their valuation.
0:09:02 There’s a lot of different uses for even just that one fundraising prediction among the almost
0:09:04 dozen different insights and predictions that we have.
0:09:05 Gotcha.
0:09:05 That makes sense.
0:09:10 I mean, I think it would be great to just jump in and sort of get a little demo, a little
0:09:11 tour of what it’s capable of.
0:09:14 And, you know, this is kind of a YouTube first podcast.
0:09:18 We like to be really visual and show what we can, so love to jump in and just sort of
0:09:21 get a sneak peek and give people a little demo of what it can do.
0:09:22 Yeah.
0:09:25 So let me see if I can work the internets here.
0:09:27 So this is the new homepage.
0:09:30 That’s very different than the old homepage of yesteryear.
0:09:34 And you’ll notice that we’re right up front saying sort of the new data is coming in.
0:09:39 I think we’ve got 12 and a half million new predictions in just the last 30 days.
0:09:44 So this engine is constantly running, constantly updating, trying to find not just the next
0:09:48 funding round, but also, you know, what is the next acquisition?
0:09:49 What is the next IPO?
0:09:51 Which companies do we predict to grow?
0:09:55 So we’re really looking at like a lot of different aspects of what a company is.
0:09:58 We’ve got this new sort of AI agent that can help you sort of navigate crunch space, but
0:10:01 you can just type in, you know, blue sky as an example.
0:10:06 And now we’re looking at the blue sky profile and you’re going to see right up front, this
0:10:07 didn’t used to be there.
0:10:10 Now up front, we put some of the biggest predictions up at the top.
0:10:14 So in blue sky’s case, we think it’s probable that they’re going to raise another round of
0:10:14 funding.
0:10:18 We think that it’s likely that they’re going to get acquired at some point, which not a
0:10:19 lot of people are talking about.
0:10:21 And we don’t think blue sky is going to go public.
0:10:23 Certainly there’s no signals that they are.
0:10:27 And then as we scroll down, we sort of took this approach to the profile page of, you know,
0:10:31 what is the stock ticker equivalent of a profile?
0:10:35 So there’s no way we can put valuation day by day over time, which is what a stock ticker
0:10:36 does.
0:10:38 But what is the private company equivalent of that?
0:10:41 So we’ve got these things called heat score and growth score.
0:10:45 And there’s a little definition of what these things are, but like looking at how the company
0:10:50 is interplaying with the public web, how it’s interacting with us, how does it rank among
0:10:55 all the other companies within our corpus of companies that we track that gives us these axes of
0:10:57 data like this heat score and growth score.
0:11:01 And that, again, drives other pieces of the application and even some of our predictions.
0:11:05 So we understand what’s happening and what’s going to happen next to the company.
0:11:09 You can play around with this, you know, and sort of make it do different whiz bang things.
0:11:13 So if you’re a data nerd, you can kind of get into this and all the raw data is available
0:11:14 in the API, of course.
0:11:18 And then as I scroll down, you’re going to find predictions and insights.
0:11:21 So here’s where I can see we predict they’re going to be growing.
0:11:22 We don’t just say it’s going to be growing.
0:11:28 We actually explain with our own words why we think that this company is currently growing.
0:11:31 And if there’s a growth prediction, why we think it’s going to grow in the future.
0:11:34 And then here’s some of these predictions like we were talking about.
0:11:38 Like here’s we think there’s a 37 percent chance that they’re going to go and fundraise in the
0:11:40 next six to 18 months.
0:11:42 There’s a good chance they don’t fundraise, right?
0:11:48 So we’re kind of transparent that these numbers don’t necessarily lead up to 100 percent because
0:11:51 there’s still some percent chance that they’re not going to raise at all.
0:11:53 So we’ll go and put that in there.
0:11:54 Are they going to get acquired?
0:11:55 We give reasons as to why.
0:11:59 In the API, we give all of the detailed reasons, right?
0:12:04 So we actually give percentages and we go and say, here are the drivers that we believe lead
0:12:04 up to this thing.
0:12:09 So as an API user, you can discount things that we think are true that maybe you don’t
0:12:13 want to sort of incorporate into the prediction, or you can just use the prediction score on top
0:12:15 of your own prediction algorithms, which a lot of VCs do.
0:12:19 They sort of use as an input into their own proprietary algorithms.
0:12:21 So this is some of the stuff that we’re doing nowadays.
0:12:26 There’s a lot around like we rock all the news that is happening on a company and sort of summarize
0:12:27 it for you.
0:12:30 So you don’t have to read 30 news articles to figure it out.
0:12:31 We sort of bring it all together.
0:12:34 So there’s just a lot of different pieces that help you understand.
0:12:40 And of course, we still have the funding data, but that’s all sort of just drivers now into
0:12:43 these sort of bigger, meatier questions that we’re trying to answer.
0:12:43 Yeah.
0:12:45 There’s a lot around this.
0:12:49 There’s a lot around, even on that homepage, you know, sort of seeing what’s important,
0:12:52 what’s trending, what’s happening in all the data we’re tracking.
0:12:56 And you can decide, I want to look at these particular predictions for these types of industries
0:13:00 and sort of get a daily feed of all the stuff that’s happening and what we think is going
0:13:03 to happen next in these companies, which is pretty exciting as well.
0:13:05 So lots of interesting use cases.
0:13:10 And again, you can always go and have a conversation with a scout, which is our little sort of dog
0:13:14 fetching thing that will go and sort of do some of the logic stuff that you couldn’t do in
0:13:15 Crunchbase before, right?
0:13:21 Before you couldn’t figure out sort of what’s the business model of this company, or how does
0:13:26 this compare to another company or how do public events impact these particular private
0:13:26 companies?
0:13:30 Now you have a way to have that conversation with Crunchbase programmatically, which is kind
0:13:30 of cool.
0:13:35 Have you found any like questions that have been like really, really valuable, like any sort
0:13:38 of like best practice questions where you’re like, if you ask this, you’re going to get some
0:13:39 really good stuff out of it?
0:13:40 Yeah.
0:13:42 I mean, there’s meaty like policy questions, right?
0:13:46 Like, so we see whenever something happens in the government, you know, people come to Crunchbase
0:13:49 and go type in, how are tariffs going to affect this company?
0:13:53 And it will do a pretty good job of sort of speculating and sort of figuring out what’s
0:13:53 going to happen next.
0:13:55 I think those are some of the interesting ones.
0:13:57 And then just the analysis, right?
0:14:00 It can be hard, like if we go over to AI Search Builder.
0:14:05 So now like you can just natural language in your query and it’s going to go and build sort
0:14:10 of these very complex searches because, you know, you think about multi-joint searches and
0:14:11 how to build those.
0:14:12 It’s always been sort of cumbersome.
0:14:18 Now you can say, show me all the CEOs at companies where they used to work at Salesforce and
0:14:23 then they went to Stanford, you know, like you could type that all in, in a huge run-on
0:14:28 sentence and it will go and show you exactly which companies do it because it builds a joint
0:14:28 for you.
0:14:31 And just that alone is a huge time saver for our users.
0:14:33 I wonder what this is going to do to like startups.
0:14:37 It feels like it’s going to like really increase like the velocity of rounds, like how fast rounds
0:14:40 will close because, you know, you guys are kind of creating like the ultimate like FOMO
0:14:41 machine, right?
0:14:44 People are like, oh my God, look, it’s a hot company.
0:14:45 Crunchbase just told me it’s hot.
0:14:47 I got to get in.
0:14:47 Yeah.
0:14:49 I’m curious if you guys have seen that or like what your thoughts are.
0:14:50 Yeah.
0:14:54 I mean, it’s hard for us to have data on this, but anecdotally we’ve heard that when we go
0:14:59 and signal that is very imminent that a company is fundraising, that they get a lot of inbound
0:15:03 interest from investors because there’s now awareness that it’s happening.
0:15:05 Oh, you notify people or?
0:15:07 We don’t, but people set up their own alerts, right?
0:15:14 I can just say, you know, show me biotech companies that are very likely to fundraise,
0:15:15 right?
0:15:16 I can just make that search.
0:15:19 It’s going to go and, you know, this is a live demo.
0:15:20 We’ll see what actually happens.
0:15:21 But there we go.
0:15:22 So it did.
0:15:25 It said industry is biotechnology, funding predictions here is very likely.
0:15:28 So these are all the companies that are very likely to fundraise soon.
0:15:28 Yeah.
0:15:30 And we can go and create an alert off of this, right?
0:15:35 So you’ve got investors who are subscribers of ours say, when a new company shows up on
0:15:37 this list, shoot me an email, right?
0:15:39 Go and let me know that that’s happened.
0:15:41 And so we don’t need to send them emails.
0:15:44 They’ll get their emails themselves because they’ve set up their alerts the right way.
0:15:45 Yeah, that’s awesome.
0:15:46 So I can do like AI coding or something like that.
0:15:50 And I can just like, as soon as you guys have a new prediction in that category, you guys
0:15:51 will email me or something.
0:15:52 That’s amazing.
0:15:52 Yeah.
0:15:54 And that’s just one of the predictions.
0:15:57 Another very common one is I’m looking for these sorts of companies to acquire.
0:16:02 Let me know when a company of this size, no bigger than Series C, I don’t want them who
0:16:05 raised more than a hundred million dollars, whatever it is, when a new one shows up is
0:16:07 very likely to get acquired.
0:16:10 Let me know because I’m in the space of acquiring those companies.
0:16:13 So a lot of corp dev departments get excited about that.
0:16:16 Hey, we’ll be right back to the show.
0:16:19 But first I want to tell you about another podcast I know you’re going to love.
0:16:21 It’s called Marketing Against the Grain.
0:16:24 It’s hosted by Kip Bodner and Kieran Flanagan.
0:16:26 And it’s brought to you by the HubSpot Podcast Network.
0:16:29 The Audio Destination for Business Professionals.
0:16:33 If you want to know what’s happening now in marketing, especially how to use AI marketing,
0:16:35 this is the podcast for you.
0:16:41 Kip and Kieran share their marketing expertise, unfiltered in the details, the truth, and like
0:16:42 nobody else will tell it to you.
0:16:49 They recently had a great episode called Using Chat TVT 03 to Plan Our 2025 Marketing Campaign.
0:16:56 It was full of like actual insights as well as just things I had not thought of about how
0:16:57 to apply AI to marketing.
0:16:59 I highly suggest you check it out.
0:17:02 Listen to Marketing Against the Grain wherever you get your podcasts.
0:17:10 Now, are all the companies that are in here, are they all like self-added or is all the data
0:17:11 sort of pulled by Crunchbase?
0:17:15 Sort of absolutely is the short answer.
0:17:20 We get data from our users, but in 2014, it was 100% from our users.
0:17:22 Today, it’s about 5%.
0:17:23 So we’ve sort of transitioned.
0:17:28 But the brand belief is that still, like if you ask our users, where do we get our data from?
0:17:29 They’re mostly, we’ll say it’s user-generated content.
0:17:34 But really, that’s only for the smallest companies, ones that haven’t had a news article yet.
0:17:35 No one knows they exist.
0:17:37 They go and put themselves in.
0:17:39 They sort of announce themselves on Crunchbase.
0:17:43 But the reality is we invest tens of millions of dollars now every year into getting data from
0:17:44 a lot of different sources.
0:17:47 We have 5,000 partnerships of data coming in.
0:17:49 We, of course, go look at government filings.
0:17:52 We’ve got lots of data partnerships to go and give us seed data.
0:17:55 And then we have our own AI systems that go out and find the data.
0:18:00 So if we don’t hear about a company that isn’t in our data set, we very actively go and fill
0:18:03 out the profile as best as we can, assuming it fits a certain set of criteria.
0:18:08 And that plus the engagement data, plus the generated data, right?
0:18:12 Like the biggest source of data now is Crunchbase generating its own data on the data that we have.
0:18:17 That is that huge, huge, huge data set that we’ve collected over the years now.
0:18:19 But not to devalue the user-generated data.
0:18:20 That’s still very important stuff.
0:18:21 Right.
0:18:26 It just seems like it could be such a really good discovery engine for small new startups.
0:18:29 You want to make sure you’re in Crunchbase because then you’re sort of in that algorithm.
0:18:33 You’re in that system where now people might discover you if they’re looking for,
0:18:36 you know, small startups in X niche, right?
0:18:38 So that’s why I was curious.
0:18:43 Can companies just go and sort of input their data in there to make sure that they get discovered
0:18:44 when people are making those sort of queries?
0:18:46 Yeah, they absolutely can.
0:18:50 And we do a bad job of this, like giving reasons why you should, because they ask,
0:18:53 you get discovered by VCs who wouldn’t normally have found you.
0:18:54 And if you’re like, I’m a bootstrap company.
0:18:55 I never want venture funding.
0:18:56 I don’t need to be in Crunchbase.
0:19:02 Job seekers is like a good solid 10% of our users are going and researching your company
0:19:03 to see if it’s a company they want to work at.
0:19:07 If you’re not in there and it’s not up to date, they’re like, this isn’t something I want
0:19:09 to go and participate in because they couldn’t grok your website or they couldn’t find your
0:19:11 website or whatever the case happens to be.
0:19:16 There’s a good chance that Crunchbase’s profile comes up higher than the website of the smaller
0:19:16 company.
0:19:21 So it’s usually a good idea to have that data correct in Crunchbase.
0:19:26 Now, it used to be traditionally like VC-backed companies that were mostly on Crunchbase, right?
0:19:27 Is that still the case?
0:19:29 Are there companies that are private companies that are non-VC?
0:19:31 Like you just said, you suggested that they should do that.
0:19:34 But also like, I can see this working for even like public companies, like just like a
0:19:39 general tool to help me guide my investments in companies in general.
0:19:42 Yeah, another brand challenge that we have.
0:19:44 There’s about 300,000 companies that have ever received funding.
0:19:46 We have about 4 million companies in Crunchbase.
0:19:51 So that math is very different because there’s a lot of companies that never get to funding.
0:19:52 There are a lot of bootstrap companies.
0:19:56 There are a lot of companies that will do acquisitions that we had not seen before that will go and
0:19:57 put in Crunchbase.
0:19:59 And then every public company is in Crunchbase.
0:20:01 So we do track all those things.
0:20:05 Now, are you going to get all the information you want to about a public company?
0:20:05 Probably not.
0:20:09 Like you should go to Yahoo Finance or Google Finance or whatever the latest degree this is,
0:20:12 because there’s so much public data out there that we just aren’t going to track because
0:20:13 it doesn’t apply to private companies.
0:20:16 But for the data, it does overlap for sure.
0:20:22 So for instance, when like Virgin America back in the day, that was a hot startup airline.
0:20:24 So it was definitely tracked in Crunchbase.
0:20:27 Alaska Airlines was not tracked in Crunchbase.
0:20:31 But when Alaska Airlines bought Virgin, then we started tracking Alaska Airlines.
0:20:36 So like this ecosystem and sort of the spider web of how companies are interconnected with one
0:20:38 another expands, expands, expands, and we’ll just keep adding the companies.
0:20:43 But we’re not going to put in, you know, Joe’s Pizza Shop on the corner because it probably
0:20:46 isn’t relevant to our community unless it’s got some cool tech.
0:20:50 Let’s say you’re not a VC or an angel investor or something like that.
0:20:54 Do you see value in using Crunchbase for just general people that are interested in investing?
0:20:59 What’s the value of Crunchbase to the people that can’t write big checks and get into like
0:20:59 early startups?
0:21:04 Actually, it’s a minority part of our users are actually VCs who have funding that want
0:21:04 to go invest.
0:21:08 So that is an important use case to us, but it’s certainly not the biggest, you know, you’ll
0:21:09 see use cases across sales.
0:21:13 Let me go and find the companies that are going to have money soon.
0:21:15 I’m going to go and start a sales cycle with them.
0:21:16 It’s a good time.
0:21:19 And it’s better than waiting for the fundraise to happen too, right?
0:21:20 Because everyone knows when the fundraise happens.
0:21:25 If you can know it six months, a year in advance, maybe you can sort of get entrenched earlier
0:21:26 than that.
0:21:28 So that’s a pretty big use case for us.
0:21:30 I mentioned CorpDev, right?
0:21:31 Anyone who’s buying companies, that’s important.
0:21:35 On the self-service side, we do have job seekers who go and are paying us and say, I want to
0:21:40 find hot-level companies in my area because I want to be there at their early stage, right?
0:21:42 So you’ve got that use case pretty heavy.
0:21:48 A lot of researchers and analysts from the largest consulting firms all the way down to students
0:21:50 who are trying to figure out sort of some interesting trends.
0:21:53 So you’ve got a lot of that sort of use case lurking in Crunchbase.
0:21:58 Those are some of the big ones, but it always is surprising to me to see that pie chart of
0:22:02 all the use cases of Crunchbase and saying, first, how the heck are we going to build a product
0:22:04 for all these different use cases?
0:22:07 But more generally, it’s really exciting to sort of see how many different people have
0:22:10 different uses for that private company data.
0:22:13 I was curious, I know there’s Crunch Fund, which has no connection.
0:22:13 No connection.
0:22:17 But it feels like you guys, all this data, you know which companies are going to raise money.
0:22:20 Somebody should be like piggybacking off of that and like making a lot of money off this.
0:22:23 Like you guys have all the data, find out the right companies, get into them, even like
0:22:24 small allocations.
0:22:25 Yeah.
0:22:28 I mean, we’ve toyed with the idea of doing it ourselves, honestly.
0:22:32 Like one of the ideas that we have lurking out there, it’s on the roadmap for not this
0:22:37 year, but maybe some future year, is I showed you a profile that has our equivalent of a stock
0:22:37 ticker, right?
0:22:39 Which is this growth and heat score.
0:22:41 What if we aggregate those, right?
0:22:43 What if we do that across entire industries?
0:22:47 So now you’ve got the AI heat score and growth score over time.
0:22:48 We’re using our stuff.
0:22:51 We’re predicting, is this moving up or down in the right direction?
0:22:58 What if we worked with maybe a secondary provider and create a little index for those companies?
0:23:01 So you could invest in some subset, right?
0:23:04 And letting the retail investor maybe start playing around with this.
0:23:06 It’s still an idea.
0:23:10 We’ve sort of had some early chats about it, but you sort of stumble into a lot of regulatory
0:23:13 issues pretty quickly, but you know, you’ll never know.
0:23:15 It could be distracting for you guys.
0:23:16 It’s not part of your core mission.
0:23:18 And the ROI is way out there, right?
0:23:20 So like, yeah, we could start a fund.
0:23:21 It’s a 10-year thing.
0:23:24 We’re a little bit more focused on the present than that far out.
0:23:24 Right.
0:23:25 But so intriguing.
0:23:27 Yeah.
0:23:31 I’m kind of curious about like what the sort of future of investing looks like.
0:23:35 And I don’t necessarily know the exact question to ask because I don’t know what I don’t
0:23:37 know when it comes to like AI and investing.
0:23:43 But I’m trying to figure out like if the general population has access to the information and
0:23:45 like what’s likely to sell next.
0:23:49 And, you know, this information is, for lack of a better term, democratized, right?
0:23:53 I’m curious about what the world looks like as we move closer and closer to that reality.
0:23:54 And I’m just curious if you have thoughts on that.
0:23:55 Yeah.
0:23:58 I think there’s a lot of potential disruption across a lot of different industries.
0:23:59 I think data is included in that.
0:24:02 And that’s honestly why we moved the way we did.
0:24:06 Like I would argue funding data is already commoditized, right?
0:24:08 So like, yes, we think we have the best.
0:24:09 Yes, there’s a lot.
0:24:14 But if we just kept resting on those laurels, like that company goes out of business when
0:24:19 all of the data gets absorbed into our LLM masters, you know, what else is there?
0:24:20 Once it goes in, it’s not going to come back out.
0:24:23 So facts are a dangerous business to be in.
0:24:27 So speculation and predictions is at least dynamic and changing.
0:24:29 And I think a lot of data companies are going to be thinking like that.
0:24:34 If you deal, even if it’s hard to get facts or you lose it all, if someone takes it all
0:24:36 and uploads it, is your business in trouble or not?
0:24:38 That’s the question I think everyone should be asking.
0:24:43 Now to the broader question of how does it affect the entire industry?
0:24:49 You know, I think generic tools that do not have proprietary pieces of the story are going
0:24:50 to be very, very hard.
0:24:52 The first mover to manage isn’t going to be a thing.
0:24:54 It’s always going to be a challenge.
0:24:58 So how do you go and build a thing that is uniquely yours?
0:25:02 I don’t know of how to do that unless you are building the foundational models, right?
0:25:07 Like you are the open AI and everyone’s on top of you, or you’ve got something that truly
0:25:10 changes all the time and is only available to you.
0:25:13 And it’s critical to people’s business workflows.
0:25:15 I don’t know how else you survive.
0:25:20 You know, I go to a lot of these AI conferences and I see a lot of people like building AI on top
0:25:21 of their product.
0:25:25 But I really think there’s a day, I mean, Replit’s almost there and Cursor’s getting
0:25:30 there as well, where you can just describe the thing and it’s going to build as good as the
0:25:30 thing that you have.
0:25:36 As long as you’ve got a good product manager with a good set of ideas, that tech is not
0:25:36 that far away.
0:25:38 And then you just fast forward five years.
0:25:41 It’s going to suggest things to do to beat the competition, right?
0:25:43 It’s going to code it for you and build a feature.
0:25:45 It’s going to raise the money for you too, or it’s going to…
0:25:46 Yeah, I mean, maybe.
0:25:49 So all of that becomes commoditized, essentially.
0:25:50 So there is…
0:25:54 Now it’s just like companies are going to go back to building their own internal tools
0:25:59 because their customers bespoke for what they need rather than trying to fit into someone
0:26:00 else’s package.
0:26:05 So in those scenarios, you’ve got to bring some other value other than that into the equation.
0:26:08 And that’s why being a data company that has some stuff that no one else has feels pretty
0:26:09 good for that long-term vision.
0:26:10 I’m biased.
0:26:13 You also mentioned that your Crunchbase has an API as well.
0:26:18 So I mean, that API can sort of work into your own sort of proprietary stuff.
0:26:21 I’m sure there’ll be people out there that figure out some really good prompts and really
0:26:25 good data points to look at to sort of make their own predictions and then not want to share
0:26:29 them with the world because that’s their sort of little secret sauce that they figured out.
0:26:30 Totally.
0:26:33 Almost every major VC now has their own data science team.
0:26:37 And we have conversations with them saying, hey, how would that API feed into your team?
0:26:39 We don’t want to replace that team.
0:26:40 We’re just going to supplement them.
0:26:43 That API also is used in lots of different applications.
0:26:44 So a lot of people don’t know this.
0:26:49 Like we power most private company data you’re going to find on any site out there.
0:26:56 So when you’re using, you know, a major financial tool or a CRM tool, our data often is in there
0:26:59 because we have a partnership with them, which is code for the same.
0:27:02 They’re a customer of ours who have taken our API and putting it in their product.
0:27:06 And that was a strategic decision because we want to make sure people don’t build competing
0:27:08 databases to ours.
0:27:13 But also it’s really exciting to sort of see the innovation and like see how people incorporate
0:27:15 our data into their own tools and make them successful.
0:27:16 For sure.
0:27:21 Nathan, I’m curious, like the sort of investing, you know, venture capital world is sort of more
0:27:24 the world that you’ve played in and less the world that I’ve played in.
0:27:28 So I’m curious if there’s any ground that we haven’t covered that you want to make sure
0:27:28 we cover.
0:27:31 I mean, just still, like I said, the thing that keeps going through my mind is like, okay,
0:27:34 if everyone has access to this data, then you have to find the outliers.
0:27:38 Somebody probably should be making like a really great newsletter on top of this data
0:27:42 and like giving their own opinion about what this means and what they’re saying, you know,
0:27:43 beyond just the data.
0:27:44 Yeah, that’s true.
0:27:44 That’s my take.
0:27:45 That’s true.
0:27:50 I mean, we are thinking about how to build sort of value reports on top of the data.
0:27:54 So it’s not just like reporting on the raw data, but can we put a narrative to the data?
0:27:54 Right.
0:27:56 So that’s one angle that we’re thinking about.
0:27:59 Another angle we’re thinking about is how do we take what’s happening in public markets
0:28:02 and interlock it with private market data?
0:28:08 So for instance, if, you know, a certain set of companies, let’s say biotech companies are
0:28:14 suddenly their stock market is tanking and they’re doing really poorly, how does that affect the
0:28:15 VC market?
0:28:17 How does that affect our predictions, right?
0:28:19 Like there’s these external influences.
0:28:24 And then can we report on that and say, look, based on what we’re seeing in these sort of
0:28:29 external sort of public markets, we predict there’s a cooling happening on this side of
0:28:32 the house and those dollars are going to get redirected to, you know, whatever, whatever
0:28:34 the other hot trend is at the time.
0:28:38 Like we can start making more of a commentary on what’s happening in the world than just leave
0:28:40 it to others to interpret.
0:28:43 There’s sort of countless opportunities lurking out there, especially when people are willing
0:28:48 to pay for sort of like these very deep analyses of what’s happening in a certain industry
0:28:50 or a certain micro-league industry.
0:28:54 I do think out of the use cases you mentioned, like for me, the sales one is super interesting.
0:28:59 Like if I could talk with AI and like, okay, I’ve got a marketing agency and maybe, you know,
0:29:00 you could even get really detailed.
0:29:04 Like I went to Stanford, maybe like look up startups that they went to Stanford so I can
0:29:05 like bond over that.
0:29:07 Like if you could get like really detailed like that, right?
0:29:12 And then reach out like, hey, we both went to Stanford, we both went wherever, and then
0:29:12 start conversation.
0:29:15 I think that could be like super powerful for a lot of people.
0:29:15 Yeah.
0:29:18 We’re trying to change the definition of what an ICP is, right?
0:29:21 Look, here’s exactly who I always sell to successfully.
0:29:22 And we also went to school together.
0:29:24 Those are still historical facts.
0:29:29 So the thing we’re trying to change in people’s minds are, is it the right time for you to talk
0:29:29 to them?
0:29:29 Cool.
0:29:33 Here’s this list of companies, which is the right one to talk to you right now for this
0:29:35 particular accounting executive, right?
0:29:38 So this accounting executive historically has been great at selling these sorts of deals.
0:29:42 Here’s companies that match the kind of companies that they would successfully sell to.
0:29:45 But here’s ones that we are predicting are going to grow, growing quickly.
0:29:47 They’re probably fundraising in the next 18 months.
0:29:51 They’re not going public and they’re not going to get acquired because that would distract the
0:29:52 sales cycle.
0:29:54 So you can kind of like tee up.
0:29:57 This is the right time for this ICP is the one you want to talk to you right now.
0:30:02 And I think that immediacy and urgency, hopefully we’ll start changing with the definition
0:30:05 a little bit of who we talk to you next as a sales team.
0:30:05 Yeah.
0:30:06 That’s worth a lot.
0:30:07 Hope so.
0:30:08 We’ll find out.
0:30:14 Now, can you use Crunchbase to sort of discover emerging markets as a whole?
0:30:18 Like, obviously you can look at in a specific market and find the companies that are sort
0:30:21 of, you know, making moves in those industries.
0:30:27 But can you find like, you know, you mentioned biotech and, you know, had you known like what
0:30:30 AI was going to do over the last, you know, six or seven years?
0:30:35 Like, are there any ways to sort of see that stuff coming a little bit sooner?
0:30:35 Yeah.
0:30:39 It is one of the harder to find features in Crunchbase, honestly.
0:30:41 So, so it’s really possible to do.
0:30:46 There’s things called hubs and not many people know what even the hub is, but we basically
0:30:50 took every major piece of metadata that has data.
0:30:51 So for instance, like industries.
0:30:54 So we have every industry that’s one access.
0:30:56 Geo is another access.
0:31:01 There’s gender, there’s founder, there’s stage of company.
0:31:02 How much they raise.
0:31:03 And we basically intermixed up.
0:31:04 So we made these pages.
0:31:09 So somewhere on Crunchbase, there is a sort of female founder in crypto in Europe.
0:31:11 There’s a page for that.
0:31:13 And we originally did it just for SEO reasons.
0:31:15 And it collects all the data.
0:31:16 So it’s like, here are all the latest funding rounds.
0:31:18 Here are all the people.
0:31:19 Here are the companies.
0:31:22 And they’re all ranked by which ones are trending the most in Crunchbase.
0:31:24 So you take that.
0:31:26 So every single thing in Crunchbase has a rank.
0:31:28 So hubs have ranks.
0:31:34 So which is the hottest one right now, today, based on what’s happening in Crunchbase?
0:31:37 There’s a first, second, third, fourth, all the way down for every single combination of
0:31:41 the, I don’t even know, 100,000 of these different hub pages that we’ve created.
0:31:44 Then you can go and say, well, which ones are trending?
0:31:48 So if you start looking at which ones are trending and which ones had the low rank that are trending
0:31:52 upwards quickly, that’s where you get to see which of those combinations is the hottest.
0:31:55 And you’ll find some really just fascinating things lurking in there.
0:32:00 Some of it’s going to be weird, you know, but like Taiwan artificial intelligence companies
0:32:01 are like a hot thing right now.
0:32:02 I love that.
0:32:02 I used to live in Taiwan.
0:32:03 It’s great.
0:32:04 Okay.
0:32:04 Yeah.
0:32:08 So there’s these little pieces of, that you wouldn’t normally otherwise know.
0:32:12 And I think if you were a savvy investor who was really trying to figure out what is an
0:32:17 emerging trend, or even just again, an analyst or even a journalist who also, they also use
0:32:20 Crunchbase, you can find some interesting things lurking in hub pages.
0:32:23 I think it’s a sleeper feature that we have.
0:32:24 Very cool.
0:32:26 Well, this has been an absolutely fascinating conversation.
0:32:30 I don’t want to be like a salesperson for you, but I’m actually a subscriber of Crunchbase.
0:32:34 I actually do have a subscription and I get in there and I play around with the data
0:32:35 from time to time.
0:32:40 So like I’m actually a user and it sounds like Nathan’s fairly excited as well.
0:32:42 I used to be a user a long time ago.
0:32:43 I had no idea what you guys were doing now.
0:32:44 I checked the website.
0:32:45 I’m like, oh, this is awesome.
0:32:46 It’s different.
0:32:47 It’s definitely different than it used to be.
0:32:48 Very different.
0:32:49 Awesome.
0:32:53 But no, this has been a great conversation and I really appreciate you taking the time
0:32:55 to hang out and give us the demo and everything like that.
0:32:57 Obviously Crunchbase is the place to go.
0:32:58 Crunchbase.com.
0:33:02 If anybody listening wants to go check it out, is there anything else that they should
0:33:02 know?
0:33:05 Any other places they can maybe follow along with you?
0:33:08 Anything like that that you want to shout out before we wrap it up?
0:33:08 Yeah.
0:33:09 I mean, follow Crunchbase on LinkedIn.
0:33:12 I think that’s probably our top channel of sort of sharing stuff out.
0:33:16 And you can follow me on LinkedIn as well because I usually leak roadmap stuff.
0:33:20 So if you want to see what’s coming before it does, my product team hates it.
0:33:22 But I usually will post stuff about what’s coming soon.
0:33:23 Cool.
0:33:24 Awesome, Jagger.
0:33:24 This has been great.
0:33:27 Thank you so much for hanging out with us today.
0:33:32 And for anybody listening, if you like content like this, make sure you like this video
0:33:34 and subscribe wherever you listen to podcasts.
0:33:35 And thank you so much for tuning in.
0:33:37 Hopefully we’ll see you in the next one.
0:33:37 Thank you.
0:33:38 Awesome.
0:33:38 Thanks.
Episode 59: Can artificial intelligence accurately predict the next billion-dollar startup? Matt Wolfe (https://x.com/mreflow) and Nathan Lands (https://x.com/NathanLands) are joined by Jager McConnell (https://www.linkedin.com/in/jager/), CEO of Crunchbase and a leading product and data innovator in the tech and investment landscape.
In this episode, the hosts dive deep into how Crunchbase has evolved into an AI-powered platform for investors, sales teams, job seekers, and anyone looking to get ahead in the startup ecosystem. Jager shares details on Crunchbase’s cutting-edge prediction engine, which uses proprietary data, AI, and machine learning to forecast company fundraising, acquisitions, growth, and more. Find out how data signals can reveal when companies are preparing to raise rounds, how sales teams and investors can identify trends before the crowd, and why democratizing predictive analytics might reshape the entire investing world.
Check out The Next Wave YouTube Channel if you want to see Matt and Nathan on screen: https://lnk.to/thenextwavepd
—
Show Notes:
-
(00:00) AI-Driven Corporate Prediction Engine
-
(03:43) Predicting Company Fundraising Timelines
-
(06:50) Predictive Accuracy in Fundraising
-
(11:15) API Prediction Score Integration
-
(12:23) Programmatic Insights with Crunchbase
-
(15:42) User Data to Investment Shift
-
(20:58) Future AI Industry Heat Score
-
(23:38) AI-Driven Business Workflow Evolution
-
(27:20) AI-Driven Sales Conversations
-
(29:41) Crunchbase Trends and Rankings
-
(31:53) Podcast Appreciation and Subscription Invitation
—
Mentions:
- Want to create your own AI Agents? Get the playbook here: https://clickhubspot.com/dkc
-
Jager McConnell: https://www.crunchbase.com/person/jager-mcconnell
-
Crunchbase: https://www.crunchbase.com/
-
TensorFlow: https://www.tensorflow.org/
-
Replit: https://replit.com/
-
Cursor: https://www.cursor.com/
Get the guide to build your own Custom GPT: https://clickhubspot.com/tnw
—
Check Out Matt’s Stuff:
• Future Tools – https://futuretools.beehiiv.com/
• Blog – https://www.mattwolfe.com/
• YouTube- https://www.youtube.com/@mreflow
—
Check Out Nathan’s Stuff:
-
Newsletter: https://news.lore.com/
-
Blog – https://lore.com/
The Next Wave is a HubSpot Original Podcast // Brought to you by Hubspot Media // Production by Darren Clarke // Editing by Ezra Bakker Trupiano