Making Sense of the Crowded Customer Data Market
In the wake of Facebook’s Cambridge Analytica scandal, Europe’s General Data Production Regulation, and the California Consumer Privacy Act, the massive market for consumer data no longer operates unbeknownst to most Americans. But for digital marketing practitioners and the average consumer alike, making heads or tails of the industry is no easy task.
To break down the different kinds of customer data in the market, the impact of data sharing and selling on consumers, and the potential of privacy regulations to shape the industry going forward, Anindya Datta, founder, CEO, and chairman of Mobilewalla, recently checked in with Street Fight.
Can you help readers make sense of the different ways companies make use of consumer data? What are the distinctions among companies interested in raw data, simple derived data, and complex derived data and attributes?
Absolutely. Companies use consumer data in a variety of ways. Many B2C brands use consumer data to better understand their existing customers, helping to retain them and grow the relationship. Brands also use data to help them identify the best prospects and to make more effective offers, increasing acceptance and acquisition rates. Research companies and research teams leverage data to understand consumer trends and then use this information to inform business decisions. In many organizations, consumer data is used as part of AI and machine learning efforts to build models that predict specific behaviors or outcomes.
Anyone who wants to use consumer data to drive marketing decisions needs to determine what type of data is going to be most effective depending on their specific use case. Raw consumer data is most widely available and requires little additional processing beyond aggregating and cleansing. Raw data sources include: bidstream signals, SDK signals, geospatial and location data (POI), and beacon data, just to name a few. Raw data is usually location-centric and when overlaid with geo-spatial data can provide insight into the places where people go.
Simple derived data uses raw data as an input, applying relatively light processing to derive secondary artifacts, like consumer demographics and behavior. Many vendors use these artifacts to create audience segments that can be used to target specific customer segments through programmatic advertising. Other companies perform analytics against this derived data to provide market research and insights.
Complex derived data takes raw data to the next level by applying sophisticated AI and ML techniques to understand complex patterns like mobility (how far someone travels from home to work or how many places they visit during the day), or householding (people who live in the same household and how they may influence brand propensity or purchase behavior) and to create attributes that combine multiple data points to create features that are used to build more predictive data models.
How does demand differ for the kinds of data above? Do companies need different levels of technical expertise to instrumentalize raw data, or do most data providers ensure the information they furnish is comprehensible for a wide array of buyers?
Demand for the different types of data differs by use case and the type of data processing resources at a company. Raw data is very “noisy” — there are a lot of anomalies present that would need to be addressed through aggregation and cleansing. Moreover, raw data volumes are very high, which results in economic issues around storing and manipulating data at this level of scale. These challenges will require organizations to have teams with specific technical skill sets and the financial resources to support an infrastructure that enables storage and processing at very high scale. Most businesses who use raw data are building products where the data is used as an input or an ingredient in another product or service – effectively, data is their business. Most brands use some form of derived data as they don’t have the internal skill sets to deal with cleansing, aggregating, and manipulating a large raw data set and they want the specific insights that derived data provides to them.
Data science skill sets are in high demand and not every organization can staff to meet all of their resource needs. One way brand marketers are addressing resource challenges is to partner closely with a consumer intelligence company that does more than offer data products, but also offers data science as a service to help them apply their data.
How have recent privacy regulations including GDPR and CCPA as well as private moves such as Google’s phasing out of third-party cookies affected the data selling/sharing business?
Privacy and consent have always been a concern for legitimate companies in the business of consumer data. Making sure we provide high-quality information that doesn’t compromise consumer privacy is table stakes.
Today’s concerns focus on both regulatory compliance and public sentiment. For companies like Mobilewalla, a values-based business invested in maintaining excellent relationships with consumer brands, it’s imperative that we do our part to ensure we are fully compliant with all regulations and that we inspire public confidence through transparency and honoring direct opt-out requests as well as those passed to us by our upstream and downstream partners. We are focused on continuing to provide high-quality data that meets all privacy and consent guidelines.
Using third-party cookies is just one way to do targeted online advertising, but not the only way. Mobilewalla does not use cookies as a source of digital identity, so we are not impacted by these changes. While third-party cookies are going away, first-party cookies remain. We are able to build linkages between first-party cookies and the mobile advertiser ID (MAID) and can use this capability to expand a brand’s identity graph by adding additional data.
What kinds of third-party data are most often sold and shared? How does this data benefit the companies who buy it? How many companies can survive on first-party data alone?
Third-party consumer data has been around in various forms (such as address-based lists, voter and postal files) and has been used by businesses for a very long time. As technology has evolved, digital third-party data, which includes consumer attributes tied to digital IDs (such as emails, MAIDs, or cookies) and can include information such as demographics, behavior, etc. are a newer entrant to marketing but potentially even more powerful when used with customer acquisition, growth, and retention efforts. From my experience with Mobilewalla’s customers, companies want access to a vast array of consumer data in order to understand their customers and their prospects and determine what is predictive of success in their marketing programs. The types of digital data available ranges from behavioral and demographic data to location and point-of-interest visitation data.
More and more, companies realize that they need a broad data set that is persistent over a long period of time (cookies are not persistent, emails are often one too many, MAIDs are the one persistent ID that gives you scale and reach) to understand trends and see where consumer interest is headed. Predicting if a particular consumer might shift from a world-traveler phase into a nesting-at-home phase is extremely valuable if you’re a real estate company, bank, or home goods retailer. Likewise, being able to predict whether a consumer who has been home raising children might move into their second or third career is marketing gold for companies that sell to professionals.
Companies are also realizing third-party data is essential for insights that will inform brand strategy and campaigns. What defines a good customer based on first-party data simply doesn’t translate to an understanding of which consumers are most likely to be a good prospect. This one can be tricky to grasp, but marketers are seeing it — the attributes that define a loyal customer, based on their first-party data, are not necessarily the same attributes that can be used in the market to help them find who is likely to be interested in the latest product or service.
Does third-party data selling and sharing benefit consumers? How can it be shared in privacy-compliant ways?
Here’s a good way to think about third-party data in terms of its benefit to consumers: When you see a product or service that fits your needs, is delivered in a format and at a price that appeals to you — that’s thanks to third-party data. A company’s ability to understand consumers, their changing needs and interests, is at the crux of third-party data. Otherwise, consumers would be inundated with offers that don’t matter to them or make sense — brands will just try to sell everything to everyone.
Understanding a consumer does not mean sacrificing consumer privacy. There are plenty of publicly available permission-based data sources as well as data that consumers are comfortable sharing. It’s a matter of using data wisely, in compliance with regulations and with a sophisticated understanding of data science to see what the data indicates — and then acting on it with a keen sense of ethics as well as social, emotional, and cultural intelligence.
Anindya Datta, PhD, is a leading technologist and innovator with core contributions in best-in-class large-scale data management solutions, artificial intelligence, and internet technologies. He is founder, CEO, and chairman of Mobilewalla.