Overcoming the Catch-22 of Data Aggregation Businesses
Building a data business or a marketplace based on user data often feels like a Sisyphean task. Until there is enough data, no one wants to be a user. And until there are enough users, there is not enough data.
How can data aggregators go from nothing to something and get on the path to liquidity and scale?
1. Simple Data-to-Value Exchange
It is important to start with a simple and clear proposition, where the user provides personal data in exchange for a specific benefit. The first step needs to be almost invisible and with time, you gradually expand the quid pro quo.
For example, Foursquare in its initial days, asked users for access to their locations — through explicit checkin or implicit access — in exchange for the ability to share where they were at any point in time. As a next step, they asked users to contribute tips about the businesses they were visiting.
As the network grew larger and the data-to-value exchange became stronger, the data started to build up and eventually reached critical mass. And now, Foursquare is primarily a data service business. And, they even predicted Chipotle’s results for Q1 2016 with a creepy level of precision.
2. Single User Mode
The death knell of data businesses built on user data is the inability to overcome the catch-22 or the chicken-egg cycle.
Addressing this demands a single-user mode from the very start — implying that a single user can get value from the service without the need for anyone else to be using it.
For example, Onavo asked users to provide access to the list of apps on their device in exchange for providing insights on battery consumption by app. This was a value proposition targeted at each individual user, without pre-supposing the existence of other users. Over time, as more users participated, the insights grew deeper. Facebook eventually acquired Onavo for over $200 M.
3. Bootstrap Data
You cannot start with an empty hall and expect people to come in droves. At the very least, you need to provide pre-assembled data and services that make it easier for initial users to get value.
For example, Yelp started out with a simple MVP — a business directory. Initial users found value in searching for a business category, or a specific business in a particular zip code, even though there were no reviews, fancy features, or participating businesses.
Datalogix struck early partnerships with key retail partners to build a trove of offline purchase data. Datalogix then used this data to partner with Facebook and helped them prove value to an advertiser by allowing them to target people who buy specific categories of products, and helped track if they bought products from the advertiser offline, after being exposed to Facebook ads.
4. Bootstrap Users
It is incumbent upon founders to find initial traction through constant experimentation. Some startups have used special channels and incentives to jumpstart acquisition; and in doing so, ended up building bypass tracks to liquidity a lot faster and cheaper.
For example, Quettra integrated their mobile intelligence SDK into the apps of a handful of developers with large user bases.
The quid pro quo for users was that they share the list of apps on their phone with a one-click consent, in exchange for more relevant products and offers.
The quid pro quo for the developer was that in return for providing app graphs of their users, Quettra will provide them specific insights that will lead to more targeted advertising at higher CPM, and a more optimized user experience.
In a few months from launch, Quettra had assembled the profiles and app graphs of over 100 M users.
Jigsaw controversially paid initial users for uploading their contacts, but quickly pivoted off the strategy when they reached a critical mass of data. They were eventually acquired by Salesforce.com for $142 M.
5. treat data with respect and ask for consent
A lot of supply-side data aggregator businesses — particularly DMPs or data management platforms — have become highly controversial, attracting the attention of consumer rights activists and regulators.
All data that is sold needs to be anonymized and de-personalized. There are serious consequences to real people when companies are fast and loose with data.
User consent (opt-in vs. opt-out) is a key consideration. None of us want our browsing and purchase data to be tracked, packaged, sold and resold without consent. But this happens every second on the internet with hundreds of third party beacons and cookies constantly tracking your browsing data and speculating on what you may want to buy.
The harsh truth is there are too many bad actors in the data ecosystem today. It is hard to police such behavior with antiquated laws designed for a different age and time. We are in desperate need of data related laws and regulations that reflect the world we live in, circa 2016.
In conclusion,
Data businesses have tremendous potential for growth and value creation — both for the companies that create and sell the data service, and the enterprises and professionals that use the data service. The economics and characteristics of data businesses are akin to platforms with high gross margins, strong network effects and high levels of customer loyalty.
Every company that has proprietary data about its customers, needs to consider the opportunity to provide customer insights and data-as-a-service to its ecosystem. But, data security, privacy and user consent always come first and they need to be handled with great care.
Social, Search and Commerce platforms are data businesses by design. They extensively use data and machine learning to deliver more targeted advertising and a more optimized customer experience. As they continue to amass rich stores of user data, expect to see more platforms deliver data-as-a-service offerings to enterprises and professionals.
Supply-side data aggregators will increasingly proliferate across multiple industries, geographies and domains. The key to success is to carefully manage the quid pro quos on the path to liquidity, while finding ways to jumpstart users and data through new channels. And most importantly, treat user permission to personal data as a privilege, not a right.
Customer-side data aggregators represent significant promise for the long-term. As more customers seek agency and respect for their privacy, and as more suppliers offer APIs to customer data, we will see a new generation of startups that help customers aggregate their own profiles and use it to engage with existing and new suppliers. I wrote more about this in the rise of intelligent agents.