Data and Value
As with other economic systems, at the heart of the data economy is value. As we will discuss, organizations have made it clear that they see tremendous value in consumer data. Some have explicitly said as much. Others have revealed it through their actions. Depending on the organization we’re discussing, there are a number of ways in which detailed data have been turned into value.
Collecting and storing data should be viewed as an investment. As with other investments, the question that organizations should be able to answer is, what is the payoff associated with compiling databases? Police forces, for example, have turned to predictive analytics. In addition to the hardware and software investment, there’s a cost associated with hiring the analytic talent to conduct the necessary analyses. Here, the payoff can be viewed as increased public safety, as manifested through a reduction in crime.
Google demonstrated that certain search terms are correlated with flu activity.11 From a public health perspective, such information could be useful in determining when it is most essential to ramp up efforts to encourage individuals to get vaccinated. Health insurance providers may take such efforts upon themselves, looking at the potential savings associated with reducing the number of hospitalizations. Employers may also promote vaccinations, hoping to curb the amount of worker productivity that is lost due to workers taking time off to recuperate. Viewing this problem from another perspective, pharmaceutical companies could identify the value of such data if it would enable them to make more efficient use of their marketing budgets. As you can see from this one example, the potential value associated with a particular piece of data depends on the organization’s goals.
Although these two illustrations demonstrate what can be gained by organizations turning to data that is generally available, consider briefly data that may not be available for public consumption: a consumer’s purchasing habits. Consider simply the question of how strongly you prefer Coca-Cola to Pepsi. If Coca-Cola knew which consumers were only interested in its products, which consumers were only interested in Pepsi’s product, and which consumers did not have a strong preference, it might change the way it approached marketing to each of these different consumers.
It might decide, for example, to spend just enough on marketing to loyal Coca-Cola consumers to encourage them to purchase more frequently. For these consumers, though, the company is not worried about them switching over to their competitor. For those consumers loyal to Pepsi, it may not make sense for Coca-Cola to exert any effort marketing toward these individuals. If their brand preferences are so strongly in favor of Pepsi over Coca-Cola, there’s little that Coca-Cola would be able to do to sway them. For the consumers in the middle, perhaps that’s where Coca-Cola’s (and Pepsi’s) marketing efforts have the potential to have the biggest impact.
Coca-Cola and Pepsi, as well as all other publically traded companies, are in the businesses of what’s best for their shareholders, but we could also apply the same thinking to the presidential election. Across the country, there are states that are deep blue and there are states that are deep red. Although a candidate could pour money into advertising in those states where his party has not fared well historically, barring a huge shift in the demographics of the state, such advertising expenditures are not expected to yield much of a payoff. Instead, what we are left with is a deluge of advertising concentrated in battleground states, specifically in counties where the advertising is expected to yield the biggest impact.
As we’ll discuss in more depth, not all data are equally valuable. Detailed data about the television programs viewed by a voter who lives in Wisconsin’s Dane county are likely to be of less value compared to the same data about a voter who lives in Ohio’s Hamilton county. Some pundits considered both Ohio and Wisconsin to be battleground states, so why the difference in the likely value of data from voters in these two counties? Dane county leans heavily to the left, as reflected by the 71.1% of the vote received by President Obama. In contrast, President Obama received only 51.8% of the votes coming out of Hamilton county. Political advertising can exert some sway on voters, but there’s a limit to its effectiveness. Given the strong leaning of Dane county, not much could have been done there by either party to sway voters. Hamilton County, in contrast, was identified as one of the seven most important counties in the 2012 election by The Washington Post.12 If campaigns knew the programs that different types of voters in Hamilton county were watching, such data could be used to ensure that advertising occurred in the programs viewed by the voters they were most interested in reaching. Regardless of context, the determining factor in how much data are worth to an organization is based on what the organization can do with the data and whether having the data can potentially further the organization’s goals.
Although the ability to take actions that affect consumer behavior is necessary for data to be of value, a few other conditions will affect just how valuable an individual’s particular data are to an organization. First, there needs to be a sufficient number of consumers who are “like you” with regard to your preferences and attitudes. If your outlook is so idiosyncratic that an organization can’t identify other consumers who are similar to you, it’s simply too inefficient for that organization to acquire your data and take actions tailored to you. There just isn’t the scale for this to be viable. Fortunately, it turns out that consumers are not as different from one another as they might think. Or, more precisely, they are similar enough that they can be grouped together into consumer segments, enabling organizations to pick and choose the segments on which they want to focus their efforts.
In addition to there being enough consumers like you, how much an organization is willing to spend on your data hinges on how valuable you (or the segment to which you belong) are to the organization. Frequent travelers are of interest to the airline and hotel industries because of the volume of business they generate, so they may provide these individuals with a separate telephone number for customer service, express check-in, or other perks. Casinos pay particular attention to their high rollers and provide them with a number of complimentary offerings to attract and retain their business because the amount that the casino stands to gain from the individual’s gambling activities can be quite substantial. Undecided voters in swing states are targeted because they can decide which presidential candidate receives the state’s electoral votes. In short, the segment has to matter to the organization.
With all these factors in place, ultimately the value of your data to organizations depends on how readily available such data are from other consumers. Although all consumers may be distinct from each other, the intent in forming a small number of market segments is to identify groups of consumers who are similar enough to each other and sufficiently different from other consumers. Claritas put this into practice with its Potential Rating Index for Zip Markets (more commonly referred to as PRIZM) segmentation scheme.
Members of the Executive Suites segment, for example, tend to place orders at barnesandnoble.com, play golf, and watch Saturday Night Live. This segment consists of upper-middle class singles and couples who typically work white collar jobs. In contrast, although the Bohemian Mix segment falls into the same age range, they are more inclined than the Executive Suites to live in cities and more likely to have children at home. Their media and lifestyle habits also differ—they express an interest in foreign films, are more likely to rent rather than own their home, and are more likely to read GQ.
Organizations must determine which of the segments are of interest to them so that they may focus their efforts on those segments. Once that has been determined, though, it doesn’t matter which particular individuals from the segment provide data to the organization. If other consumers like you are willing to share their data with organizations at no cost, then that’s how much the organization should be willing to spend on data acquisition. However, if each individual in a segment has determined that there is a minimum the organization must offer for someone to be willing to share the data, then it may be in the organization’s interests to invest in acquiring this data.