Briefly

TRANSCRIPT

00:10
Adam Stofsky
Today we're going to talk about monetization of data. Data monetization. So why don't we start off with a pretty simple question. What does that mean? What does it mean to monetize a company's data?

00:21
Daniel Goldberg
Monetization is a really broad term. It can encompass a, a lot of practices. And I'd say one of the first questions before we even get to that is what do we mean by a company's data? Right. So that could be anything that you have internally, it could be data that you have about your customers, it could be general analytics data, it could be personal data, which is data that identifies a consumer and that is regulated by privacy law. So I think as a starting point, we even have to think about like, what do you have? Like what's the data? And then once you have that subset of data that you're looking at, you can think about what the value is of that data.

01:01
Daniel Goldberg
And that's essentially what data monetization would be, is saying, let's dig into like what we have and how can we identify the value of that data and use that for the benefit of our business in some way? So some of the ways, right, that we see companies use their data in different valuable ways would be just to understand things about who they're working with, the businesses, the consumers. Analytics. That's one very big way that we see. We also can see users using or being, basically getting their data being used to figure out ways to target them better, right? So for advertising purposes, that's a really big one that we see. And that's the way that the free web is driven. When you go to a website, a social media platform, you're not getting a product for free.

01:50
Daniel Goldberg
They're taking your data and they're using it for some value to that company. And so it may be to show you advertisements that other companies buy on that platform in order to get your eyeballs. And there's a huge value to that because these companies have the ability to understand a lot about who their users are and try to identify them not necessarily on a one to one basis, but to identify the segment of users, the type of users to be able to get them the right ads. And, and that's very valuable to advertisers. And then the really big one that we're seeing now is around AI training how data can be very valuable to actually train models and build those models, especially for the large foundation LLMs, because the idea is that you need a lot of data to build these sophisticated models.

02:41
Daniel Goldberg
And so you can use it that way. And many companies also use. Use data for insights. They can use insights. You know, there's some companies out there that I'm aware of that can purchase. They purchase data they don't care about, again at the personal level, but they're purchasing it from companies to say, what can we glean about models of certain types of practices or businesses that you know may benefit our company? So that's very broad. I know that's a starting point, Adam, but we can dig into things.

03:11
Adam Stofsky
Can you give me a few. Just an example from each one. So, so like, and the first one you talked about was kind of analysis. What does that look like? What's an example?

03:20
Daniel Goldberg
I'm going to give you a really basic one. I'm a SaaS platform and I make my platform available to a business customer. And I want to understand like how that business customer is using my platform in order to benefit my product. Because I want to improve my product. So I'm going to understand how their clicks are moving around, what they're doing with it, and then I can actually take that there's a real value to that data to my business. Because, you know, it's one thing to say, oh, I'm going to try to figure out and anticipate what users are going to do. This is another thing to actually say I'm figuring out like what's breaking, what's not working. This has been around for a very long time, right? We're talking about some stuff like this is not a new thing with Genai at all.

04:10
Daniel Goldberg
Every company does this. But that to me is even a very basic way of monetizing the data that you have because there's value there in it, right?

04:19
Adam Stofsky
Like we do a lot of webinars. When do people tend to come, right?

04:24
Daniel Goldberg
Who's who, what type of people? I mean, I think the primary example we have here is like a Google Analytics. They provide you insight about your website, how many people came to the site, and that's a value, right? And if you think about it, you know, you're paying Google for that, so they're making money off of that.

04:41
Adam Stofsky
Right, interesting. So you're not like selling data, but you're deriving value from it.

04:46
Daniel Goldberg
Because other things, I don't think data monetization necessarily means what like is. Oh, here's a bulk sale of data. We are giving this big amount of data. And like, that's the misnomer is that you're ultimately, you're always thinking about here is, you know, there's these data brokers, they're buying huge amounts of consumer data. They're intending to, I don't know, impact elections or something like that on a one to one basis or how you're thinking. That's not actually what data monetization means. Monetization is something that every business should be thinking about and can be involved in.

05:24
Adam Stofsky
Okay, let's get an example of the. This one might be the most obvious, but you mentioned kind of advertising.

05:28
Daniel Goldberg
Yeah.

05:28
Adam Stofsky
So what's going to be an example of what that might look like?

05:31
Daniel Goldberg
Yeah, so in advertising, let's just say you are a blog, right. You want to make money in some way on your blog. How do you make money? Okay, well, I'm going to have ads. I'm going to put display ads on my blog. You're going to get paid for those display ads based on the number of users and the value of those users that are coming to your site. So if the number of users in the Valley, you know of those users, if you can actually get advertising that is more relevant to those users so that those users will click on those ads that are on your website, then you get paid more money. That's just generally how the advertising ecosystem works. And so that's what drives so much of the ad tech, targeted advertising, personalization.

06:16
Daniel Goldberg
It's all about saying we want to be able to get the most value for our ads on a website. And, and if you can be able to provide that value and have individuals on your site that are potentially high net worth to be able to capitalize on those ads, that can be very valuable. And it's your data ultimately from your site. Because in that context we're not necessarily needing to provide emails, phone numbers, none of that. We're providing IP addresses, we're providing device user data as part of that process. Process. And so again, there's a huge element of monetization there.

06:53
Adam Stofsky
So the next category you mentioned was monetizing data through large language models and selling data to someone who runs an AI model or a company that runs an AI model. Can you give us an example of.

07:06
Daniel Goldberg
What that might look. Yeah, this has been a really hot topic over the last few years and we're still seeing the legality of some of this play out. But the way that these AI models are built is they require a large amount of data. And so what has happened has been these AI models for years have been scraping the Internet to try to figure out everything that's publicly available. But a lot of companies have caught up and have basically said we're going to shut our doors unless you pay us for that data. So depending on the company that you are, that data has a lot of value. So I'm just going to give you an example, like let's say it's a Reddit or something, because I believe they've licensed, they have an API and they license some of their data.

07:51
Daniel Goldberg
Your Reddit, you're generating all this content, right? You have all these users, just not even at the individual level, but like, what's happening? What's going on in these spheres, what are people talking about? All that is so valuable and the LLMs want to be able to pull from that and to understand what's happening, to train their models to provide news. So this has been a really big area recently and in this circumstance we've seen a number of deals that actually translate to very large amounts of money, especially for the big players with respect to licensing their data for these data sets. And I've seen it interesting places.

08:30
Daniel Goldberg
I've seen it also in the entertainment industry where licensing of video, like let's say old back footage that you have just years of footage that was never used on some reality show like that actually has value all of a sudden. Well then there's all these questions about contract of. Oh, well, there's people in these. How does that work? Like, we never thought about that when were creating these videos on this reality show 20 years ago. So again, it cut all these really interesting questions, come up with the licensing in that regard.

09:00
Adam Stofsky
The final category is insights, you said. I think this is sort of the most interesting in some ways. So, so what do you mean by this and what's.

09:06
Daniel Goldberg
Well, like I have, you know, I work with some companies. I'm thinking about one in particular that is able to glean certain insights just generally based on how like for example, app stores are doing. Right? And so it's not based on an individual level, but they take aggregate level data and then they translate that and there are buyers out there that want to understand performance metrics of app stores, etc. And they could be hedge funds, they could be, you know, investment companies that are trying to do their own predictive models. And so in these cases it's really fascinating because we're talking about selling data here, but it's not personal really in any way. It's more gleaning around how are certain areas of the ecosystem doing? Or maybe again it may even be relating to like how is your own service doing?

10:02
Daniel Goldberg
And you could glean elements about how users are using it and sell it. So that's what I would say is, you know, is a very big area and there's a lot of value for companies. And the last one, Adam, before we move on to the next subject, is that there is the quintessential data sale.

10:21
Adam Stofsky
That we didn't cover that, which.

10:23
Daniel Goldberg
Is just like, hey, we have a bunch of consumer data. We're going to straight up take this and we're going to sell it to the highest bidder. That's the riskiest one. Like, that's the category you don't want to be in if possible, because that's where the regulators are looking at it and trying to shut that down. And in some areas, if you're dealing with certain high risk data sets, like around health data, some of that may not even be permissible in the next five to 10 years.

10:54
Adam Stofsky
All right, Daniel, thank you. That was a really helpful summary of different business models for monetization of data. We talked about analytics, kind of improving your own products or understanding your own products and services better. We talked about advertising. We talked about selling data to large language models and other kind of AI model builders. We talked about insight, the interesting category of insight, and then direct sales of data to data process brokers and other buyers. Did I get that right?

11:21
Daniel Goldberg
That's right. And that last one is really specifically consumer data.

11:25
Adam Stofsky
Consumer data.

11:26
Daniel Goldberg
Right.

11:26
Adam Stofsky
Great. All right, Daniel, thanks so much.