The Value of Data, Part 3: Data Business Models
Data is incredibly valuable. It helps create superior products, it forms a barrier to entry, and it can be directly monetized. This post is the third in a 3-part series about making data a core part of a startup’s business plan.
3-Part Series
Part 1: An explanation of why data is becoming increasingly valuable + a catalog of ways in which data creates competitive advantages.
Part 2: How to accumulate data + pitfalls to be aware of.
Part 3: A catalog of business models where data plays a key role. [this post]
In this post, I’m going to talk about the key ways in which data can be monetized. I will then provide a list of concrete business models whose foundations lie in datasets. The list will be expanded over time so that it can serve as a comprehensive resource.
What is a Data Business Model?
According to Wikipedia, a business model “describes the rationale of how an organization creates, delivers, and captures value.” A Data Business Model is a business model where data is an indispensable component. If you remove the data, the business fails (or at least suffers greatly).
To take one example, Amazon’s data is core to their business. Their historical transaction data helps them figure out how much inventory to hold and how to price products. Additionally, data about product views and purchases powers the recommendation engine, which drives a large portion of sales. Furthermore, product reviews drive traffic and SEO. As icing on the cake, all of this is a virtuous cycle: recommendations drive purchases, which result in more reviews, which lead to better SEO and more traffic, which results in more visitors and better recommendations. If Amazon wasn’t so effective as using data, it would be a much smaller company.
The best part of data business models is that they often have the same kind of positive feedback loop as Amazon. In each business model, the more you use data to make money, the more data you get as a result, which helps you make more money in the future. It’s a beautiful system.
How can Data be Monetized?
Data is the product or service. If you have a valuable dataset that others would pay to access, then you can sell it directly (e.g. Foursquare, FullContact, Yodlee) or build the only access point and sell it indirectly (e.g. DataFox, Mattermark, Bloomberg, and LoopNet).
Data helps drive revenue. As you learn about customer preferences, you can improve product recommendations and greatly increase each customer’s lifetime value. This is huge for Amazon and most eCommerce companies. If you’re building a repository of content (which is a form of data), you can use that to drive ads. You can also use data about your users to better monetize ad targeting, like Facebook and Twitter do.
Data helps improve profit margins. Possible ways to improve margins with data include conversion funnel optimization, price optimization, and accurate supply/demand prediction. Examples: Amazon, online marketplaces, and every company that uses A/B testing.
For example, improving margins may not seem like a huge deal, but being able to cut acquisition costs by 40% while increasing revenue by 50% through product recs can turn a business from unviable ($30 customer acquisition cost, $26 customer lifetime value) to very profitable ($18 CAC, $39 CLV).
Specific Monetization Game Plans
The business models above are great, but somewhat abstract. How can you turn them into actual businesses? Here are some specific recipes for different types of products:
Content Companies
Build a content site, use engagement data to decide what content to produce (e.g. BuzzFeed, Bleacher Report)
Build a user-generated content site, display relevant ads/affiliate links/product recs next to content (e.g. Yelp, Pinterest, eventually Quora)
Use behavioral data to create better content recommendations and higher engagement, then charge for usage (e.g. Pandora, Netflix)
eCommerce
Use purchase and conversion data to implement profit-maximizing pricing (e.g. Amazon, eBay, most eCommerce companies)
Use data to create better product recommendations and increase basket size (e.g. Warby Parker, Lumoid, True&Co)
(These two recipes can also be applied to other companies, like SaaS startups, but they have a deeper impact on eCommerce companies because of the lower margins. Taking a SaaS company from a 50% margin to a 75% margin is great, but taking an eCommerce company from -5% margin to 20% margin is what turns it into a real business.)
Data providers
Sell access to premium data (LinkedIn subscriptions, IMDB Pro, DataFox, LoopNet).
Help customers augment their datasets with external data (e.g. Foursquare for location data, Anju for medical and health data, Socrata for government data). This is different than selling data because that model is more about selling an entire self-contained datasets to customers; this model is more about helping customers who already have some data enrich their data with other attributes. This business model is often much more reliant on integration and deduping algorithms than on data acquisition.
B2B and B2C tools
Build models from product usage data (e.g. LendUp for credit scoring, Sift Science for fraud detection, Framed Data for churn prediction, Metromile for car insurance). Increased product usage leads to better models, which are both more valuable to customers and more difficult for competitors to replicate.
Build a consumer app that saves time for customers and collects data as a result (e.g. inbox organization tools like Unroll.Me, shopping related tools like Honey, and smart launchers and homescreens like Bento). This data can be used for better recommendations or ad targeting, and can often be monetized by affiliate fees.
Build a SaaS product that makes some industry more efficient, usually through replacing faxes/voicemails/emails with online forms. Use form data to build killer features (e.g. Flexport, SimpleLegal, Sourcery)
Of all of these business models, the last one is my favorite. It’s building a tool that can be sold as an efficiency play, using that tool as a Trojan Horse to collect data, then turning the data into a huge competitive moat. This business model makes it easy to build a valuable dataset because you don’t need data to start – you just need to streamline data entry for customers. Then, once you have that dataset, you’re unstoppable.
Conclusion
There are many ways to use data to make money. I’d love for this post to become a definite catalog for data business models, so I plan to add to it over time. If you have suggestions for ways to make money off of data, or specific “game plans,” please let me know on Twitter.
Thanks to Sean Byrnes, Eva Ho, Niv Dror, Elliott Hauser, Nagarjun Palavalli, Yann Person, and Jeremy Baker for feedback on this post.