Last week I attended a symposium celebrating the change of the name of the Web Analytics Association to the Digital Analytics Association. This might seem like a minor semantic exercise, but it is the outcome of significant, meaningful changes in the way data is being generated, analyzed and consumed.
The event was, for me, pretty awesome. Hanging out in New York with a bunch of Web data geeks — err, digital data geeks, I mean — is pretty much my definition of a good time. One of the main topics that was hammered throughout the day was the increase in data sets, and how they are being mashed together to create meaning for decision-makers.
Here’s a quick recap of where we are in terms of data analytics — not just for real estate, but for business in general. While real estate has some unique challenges (more than 800 ways to describe a house, for example), many of the important aspects of data are shared across many businesses.
More and better data sources are becoming available daily, it seems. Many of the technical-implementation, drudge-work tasks are being handled by "extensible software" chunks, like WordPress plug-ins, or taken on by developers who have been made aware of the importance of data by their clients or venture capitalists.
Data from a variety of sources, not just the Web, are being combined and analyzed in ways that simple Web analytics dashboards, with their limited data sources, cannot match. The opportunity for finding useful, meaningful stuff in data is increasing in relation to the available and accessible data sources.
Data geeks now spend their time trying to cobble together various data sources. There’s a lot of learning going on in regards to how different sources and types of data can be used to generate meaning about people, places and things.
But the bottleneck is still people. Sure, we’d like to have even more data sources, and the big data vendors would love for us to have exponentially more data. Yes, we’d like to have more accurate data that is gathered in a cleaner fashion. But even if we had all of that, the bottleneck would be people.
To make use of data in some way other than voyeurism, someone has to explain it to someone. And the person receiving this explanation needs to know what’s being explained.
The explanation bottleneck
There are two problems associated with the "explanation bottleneck." This bottleneck prevents data, in whatever quantity or consistency, from being fully used.
The person doing the explaining is the first problem. This is typically the analyst or Web geek or quant or whatever. Even they aren’t wearing pocket protectors, these people know data. They know how to read numbers. They can make symphonies with a spreadsheet.
This group of people produces reports that get consumed by the people who are, themselves, the second problem: the data consumers. You might think of them as managers, or CEOs, or other decision-makers.
Decision-makers wander around all day talking about "operationalizing" things or making something "actionable." If you have a really old-school decision-maker, he’s often talking about making decisions from his "gut." Their new-age (but old-school) counterparts will talk about "intuition" instead.
These two groups of people — the Geeks and the Guts — are the bottleneck that prevents data from being used effectively.
The reason is simple: The same traits that make them successful within their respective cultures make it challenging for them to relate to each other, even though that would make them significantly more effective. I could make a cheesy left brain versus right brain analogy here, but I’ll spare us.
The focal point of the bottleneck can be found in the way these groups communicate with one another: through reports. In some instances, the Geeks work for the Guts. In others, the Guts work for the Geeks. But in all cases, it’s the Geeks that are producing the communication object: the report.
The report fails. It always fails. No matter how painstakingly produced, or how intricate the charting — or, in some cases, because of the intricate charting — it fails. It fails to result in the Guts really getting a clear picture and understanding of what is in the Geeks’ minds.
This happens because of literacy. Not reading literacy, usually. But data literacy. People like the Guts who have a lot of experience making decisions based on feelings or vision sometimes don’t have as much experience reading charts and graphs.
People like the Geeks don’t have a lot of experience making decisions based on their feelings or emotions. In fact, they prefer their information to contain a veneer of objective thought. The Geeks perhaps have a lower emotional literacy than the Guts. Emotional or persuasive content can make them a little anxious.
The end result is that the Geeks aren’t presenting information in a way that is accessible to the Guts. As a result, the data isn’t being used to inform decisions or change things.
It’s about data literacy. And let me tell you right now, as much as the Guts may pay lip service to gaining data literacy, there are precious few who actually will go through with it. Not because they’re lazy or stupid. They just aren’t wired that way.
The effort to widen this bottleneck will, ultimately, reside with the Geeks learning to do the following:
- Make real recommendations based on their data.
- Stop trying to force pictures and charts to tell the story. Instead, tell the story with a story.
- Understand when a story can be told with data, and when there is no story, don’t try to make one up with a pretty chart.
The Guts aren’t off the hook though. They can help widen the bottleneck by:
- Applying intuition to the gray area between "not enough data" and "too much data."
- Understanding that data is an ever-changing thing, use gut instinct to follow where data is going instead of disassembling where data has been.
- Understand that your Geek may be reluctant to choose a side or direction from the data. Make her give you her opinion anyway.