I’ve been thinking a lot about data lately.
What’s that? You have too? Well, of course. You, like me, work in the advertising business, with at least a passing interest in the technologies that continue to morph and change our industry and culture in general. In this industry, data has risen to some pretty lofty heights. We seem to have turned our attention toward the vast storehouses of data that we know (or believe) exist or can be captured. Inside this deep vein of data, we believe there is truth and insight. We simply need the right tool to extract that truth and insight.
Perhaps this is a natural outcome of the ever-decreasing cost of storage. In 1992, a gigabyte of storage cost $1,000. This year, the same amount costs $0.08. In a few years, the cost of a gigabyte will be too small to even measure. Couple that with fast networks and you have no real reason not to capture everything.
But of course, capturing everything just means that you have everything. A significant number of tools have begun to appear on the horizon that help to find real trends – if not those actual truths – inside that data. These tools are generally called data management platforms and they are beginning to be seen as important parts of a publisher’s toolset, which means that the benefit they provide is beginning to become available to advertisers.
My feeling, though, is a tool is only good if you have some concept of what you want to ultimately do with it – what you want to achieve. You can buy the world’s greatest power saw, but big deal. You need to not only know that you want to build a cabinet, but what you’re going to put in that cabinet, where it’s going to go, and whether you want it to have a modern or rustic look. Plus, you have to know why you even want a cabinet and what benefit it’s going to bring you. There’s a big difference between “using a tool” and “building a thing,” and there’s an enormous difference between “building a thing” and “solving a problem.”
I’m going to guess that it isn’t hard for you to think of a few problems you have that are worth solving. But I’m willing to bet that the connection between the ability to use data (in a new and more powerful way) and that problem isn’t totally clear to you. To help move you a bit closer to being able to intelligently use that data, I suggest that you first think about the various types of data that exist in the world.
Think about which ones you have and which ones you can get access to. This is critical, because the next generation of interactive experiences will come from the ability to dynamically access and blend two or more of these data types together.
Here are the three types of data:
Personal data is the one that we are all most familiar with, mostly because of concerns about its misuse. This is the data that’s connected to one individual person about her. This can range from the extremely sensitive, like bank information, to the more casual, like favorite band. For a brand, personal information can be very powerful since it can involve shopping preferences, past purchases, or even current desires (such as the record of someone comparing cellphone plans). Personal data is powerful because it’s the pathway to providing highly relevant experiences to one person. We can think about personal data as the catalyst for many experiences, since it’s the cloud of information that follows someone through the day. When we can tap into that cloud, we may have the opportunity to use that to spark an experience or provide an offer.
Public data is any freely available, structured storehouse of information. This could be Wikipedia, the Yelp reviews database, or weather forecasts. Public data must be structured, though, for it to be of any use. The text of “Moby Dick” may be up on the web somewhere, but unless there’s the ability to dynamically call all references to waves (and only references to waves), it doesn’t do us much good as a data source. Public data must be freely available, but it need not necessarily be free. Many data sources charge a fee for access, which is fine. Sometimes this is done as a revenue source, sometimes the fee is just a way to limit access (if you have to pay for something, you’ll be more judicious in how often you use it). The thing that has exploded the use of public data is the growth of application programming interfaces (APIs). Without getting too technical (because I’m not 100 percent sure I know what I’m talking about), APIs allow anyone to pull data out of a public source, as well as put new data in. This is the technology that has enabled a million mashups, where a database of restaurants is intertwined with a map application to create an eater’s guide to a city.
Proprietary data is the information that you and you alone own. It may be something not too secret, like a list of all your recipes. Or it could be financial performance data for your company. Whatever it is, it’s data that only you have access to. This is the key because it allows you to create an experience or an interaction that’s totally different from anyone else’s. This is where you can start to offer something fully new to the world. In fact, starting with your proprietary data is a great way to begin brainstorming a new service: what do we have that we can use to create something that absolutely no one else can?
This is why I’ve been thinking so much about data. I’ve been thinking a lot about generating new interactive experiences for brands. Increasingly, I see these experiences being built upon dynamic blends of distinct data sources. That, hopefully, starts to answer the question of “what do we do with all of this data?” now that we not only have it, but can get at it. We can blend these three types of data together within a particular instance to create something that is just right.