The Data Businesses of the Future

Data-as-a-service that can deliver relevant insights to enterprise (both profit and non-profit) users is a fast growing market opportunity that can reshape traditional industries. Data that is generated and collected at end-user touchpoints points holds the potential to deliver dramatic improvements to operational and financial metrics of multiple industry eco-systems. It can reduce the distance between supply and demand, manufacturer and end-customer, donor and recipient, and intent and fulfillment.

Some of these industry ecosystems include:

• Consumer Products: CPG, Wholesale, Retail, Consumers.
• Health: Life Sciences, Pharmacies, Healthcare, Medical Insurance Providers, Enterprises, Consumers.
• Automotive: OEM, Distributors, Dealers, Auto-Insurance, Accessories, Body shops, Consumers.
• TV Advertising: Advertisers, Broadcasters, Cable TV, Satellite TV, Telecom, Consumers, Rating Services.
• Digital and Mobile Advertising: Advertisers, Publishers, Browser Companies, Ad-Tech, Mobile OS, App Developers, Consumers.
• Location Intelligence: Locations, Businesses, Consumers and Out-of-Home Advertisers.
• B2B Intelligence: Private Investors, Retail Investors, Venture Capitalists, Startups, Private and Public companies, Sales Executives, Marketing Executives, Strategic Planners
• Donor Funding: Trusts, Funders, Individual Donors, Non-Profit Organizations running healthcare & education initiatives, Refugee Programs,

The party closest to the final user, or a data aggregator who cuts across multiple layers, is in the best position to create, package and sell the data-as-a-service offering.

A few examples of how Data-as-a-service can create value:

• B2B sales and marketing users can discover new leads and gain intelligence on current leads by leveraging an industry-specific data service that provides deep insights and real-time alerts on potential buyers.
• Brand managers in CPG companies can understand who buys their products from retail stores and their behavior. These insights will help them make better decisions related to merchandising, packaging, branding, and store allocation.
• App developers and online websites who access a real-time consumer insights API from a data service, can a) gain access to the broader interests and device profiles of a specific user, b) use the insight to deliver the most relevant ads or products, and c) help advertisers track if the ads influenced the user’s behavior offline.
• Retail planners can find the best locations to open new stores based on the traffic patterns of consumers in their desired demographic segments.
• Scientists in pharmaceutical companies can formulate new drugs based on insights derived from anonymized healthcare data at scale.
• Traders in a hedge fund may get insights into yet-to-be-announced revenues and earnings of brick & mortar retail companies based on location intelligence.
• Donors are better placed to allocate capital towards economic development projects that need the most help and/or deliver maximum impact.

Arguably, the data-as-a-service movement is already under way, with the success of numerous data startups in the last few years.

Nevertheless, we may just be at the very beginning of a multi-decade trend.

With digital technologies pervading all industries, data is indeed the new gold. More relevant data leads to better insights and drives better outcomes for all parties in an inter-dependent value chain leading to the end-customer.

We are likely to see a wide range of new data-as-a-service businesses: vertical data services focused on specific industries or lines-of-business, and horizontal data services that aggregate across physical locations, devices or groups of customers.

But.

To build a data business, you need to deliver non-obvious insights and patterns that create value from the buyer’s perspective. The fuel for that is unique and proprietary data, typically not easy to obtain.

What are the Sources of Unique and Proprietary Data?

Broadly speaking, there are six sourcing strategies. Most data businesses will use some combination of these.

1. You have a core business that involves selling products and services, that gives you access to proprietary data.
2. You get end-users to contribute data — willingly — on a quid pro quo basis. (More on this later.)
3. You buy access to 3rd party aggregator data: audience segments, demographics, maps, points-of-interest, social data fire hoses and offline purchase history — from companies like Bluekai, Here, Four Square,Gnip, Data Sift, Axciom, Data Logix, etc.
4. Publicly available data — scraped from web pages or accessed through APIs. Proprietary techniques and algorithms can help correlate across these sources to find new insights
5. Panels of paid users who represent a statistically significant sample.
6. Surveillance and tracking mechanisms that collect user data without consent, or bind the user to onerous terms with lots of fine-print — the most morally bankrupt method of all, but widely prevalent in the digital advertising ecosystem.

Four Types of Players Are Building Data Businesses. Last-Mile Suppliers, Platforms, Supply-Side Data Aggregators and Customer-Side Data Aggregators.

1. Last Mile Suppliers

The main business of last-mile suppliers is to sell products and services directly to end-users — like Retailers, Healthcare Providers, Telcos, Cable and Satellite TV companies etc. Some of them also deliver content and advertising impressions to end-users.

A Simplified View of Industry Eco-Systems. Products, Content, Services and Advertising Impressions Flow Top-Down.

Purple boxes represent makers who are responsible for making products, services, or advertising impressions.

Green Boxes represent last-mile suppliers who sell or display products, services, or impressions directly to end-users. Some of them are also makers (like Tesla), but most are resellers, insurance brokers, or retailers. 

Orange boxes represent intermediaries between suppliers and customers. 

Last-mile Suppliers (the boxes in green) collect reams of proprietary data about the behavior of their customers, across their transactional, loyalty, billing, point-of-sale, IoT and CRM systems.

The systems of green suppliers are instrumented to throw off operational data that reflect the daily events of the enterprise. Analytical tools access this data to generate insights on operational efficiency and customer experience.

The same infrastructure also provides the foundation for offering data and insights as a service to purple, orange and other green companies. But the data needs to be anonymized, stripped of all identity, and packaged up to meet the varied needs of the ecosystem. And made accessible both via granular APIs and applications.

The applications that wrap around the data need to include utilities for search, aggregation, navigation, tagging, comparison and visual exploration. More sophisticated applications include the ability to share datasets and dashboards, collaborate with other users in-context, and set up alerts or actions based on specific triggers and rules.

For example, Kroger sells anonymized customer behavioral data — packaged as an interactive application — to upstream brands, DISH TV sells set-top box data to Nielsen and Comscore, and fleet management companies sell driving data to insurance companies.

For a company whose core business is not data, you can choose to a) partner with a data science firm — like Kroger and TESCO did with Dunnhumby, b) build a fully owned data subsidiary with deep competencies — like AT&T Data Patterns, Sprint Pinsight or Verizon Precision Insights, or c) choose a hybrid model. The right strategy will depend on current capabilities and competencies in data, organizational maturity, the sales model required, and the desired levels of investment, risk and return.

Data-as-a-service is an opportunity for old and traditional businesses to reinvent themselves as digital businesses. The data ecosystem they build can complement and reinforce the supplier and partner ecosystems for the primary business.

2. Platforms

Platforms like Google, Facebook, Twitter, and Linkedin connect people to each other, and to businesses, publishers, and institutions.

Platforms apply machine learning to very large data sets to optimize the user experience and to deliver the most relevant advertising. A few platforms also sell data-as-a-service offerings — like Linkedin Recruiter and Linkedin Premium, Foursquare’s location-based insights, Twitter Gnip, and Google Trends.

Google and Amazon are super-aggregators also playing the roles of last-mile suppliers (they sell digital and physical products) and supply-side data aggregators (they may not sell data directly, but sell ads and sponsored links that are powered by the data).

3. Supply-Side Data Aggregators

These companies are in the data business. Their primary and main goal is to sell data as a service. Their sources may include any combinations of the following: user contributed data with consent, 3rd party data, public data, panel data and tracking data. They are called “Supply-Side” as they aggregate data across multiple customers, and make their money by selling to suppliers.

Examples of supply-side data aggregator categories and companies include:

• Media Ratings (ComScore, Nielsen, etc.)
• Offline Purchase Data (Acxiom, Datalogix, etc.)
• Social Data (Datasift, Gnip, etc.)
• Data Management Platforms (Blue Kai, Adobe, etc.)
• Mobile Intelligence (Sideweb, Flurry, etc.)
• Cross Device Insights (Drawbridge)
• B2B Deal Intelligence (Mattermark, CB Insights etc.)
• Sales and Competitive Intelligence (Insideview, Owler etc.)

4. Customer-Side Data Aggregators

These companies offer technologies that help the consumer aggregate her own personal data — across multiple supplier silos — into her personal cloud. An example of this is Mint, which helped consumers aggregate their financial data across multiple credit cards and banks, and generated insights and alerts from that data. Companies like theSkimm, Nuzzel etc. who aggregate and deliver content across publishers for the consumer’s benefit also fall into this category. Early moves of Apple in Health and IoT — allowing customers to aggregate data across multiple apps and things — are also similar initiatives.

While supply-side aggregators make a business out of collecting data across customers and selling the data upstream, customer-side data aggregators provide customers a platform to consolidate their own data and act with agency. They are customer-first and look out for the user’s need more than the marketer or the advertiser.

Customer-side aggregators also help the customer engage with the world by using her data as currency and exercising choice and permission to exchange data for value. This is new territory in very early stages but has the potential to transform the worlds of marketing, customer engagement, and advertising over the long term. It remains to be seen how this movement can get boosted or hampered by the players on the supply-side, and if platforms might start to play this role. I discussed this more here, under autonomous agents.

In Part 2, we will discuss the key considerations for data aggregators to get to liquidity, and the broader implications for data regulation, privacy, and user consent.