Data. It’s everywhere these days: in newspaper articles, books, TV shows, and endless whitepapers. And for good reason, since there really has been a revolution in how companies create and use data. The tools we now use produce a data exhaust that’s easily captured. Storage is cheap. Computers are powerful. And everyone has access to affordable data science and analysis capabilities that would have made spies and codebreakers blush 20 years ago. Data is the lifeblood of modern business, so companies do need to make it a high-level priority. Hence the rise of the Chief Data Officer.
A Chief Data Officer to get a handle on your data?
One of the most important ways that data informs business decisions is by joining disparate datasets. The value of data is in the gaps, or more accurately, in the hidden connections between different parts of the business. For example, combining the information in engineering usage logs with data in a marketing CRM database lets you understand which customer cares about which features. This can be incredibly useful in deciding what products and features to build next, and how much those things are worth.
The problem is, often times these databases are all over the place in your organization. Even if the various departments work together well (which isn’t always the case), their respective databases and datasets sure don’t. Different systems, different data formats, different names and terms for similar elements. It’s a mess.
Modern companies recognize this problem. They also understand that in order to cut through the barriers and get all the data together, they need a C-level person in charge of the effort. Only by empowering someone with the organizational (and political) capital to pull things together can they reap the benefits of all that data they have stored all over the place. Problem is, CTOs and CIOs usually have their hands full running the tech operation, and keeping the IT lights on day-to-day. That’s why it makes sense to create a new position that’s more forward-looking and strategic with respect to these issues. Enter the Chief Data Officer (CDO).
Why is data so important?
This seems like an easy question to answer, right? Businesses need to make good decisions, and you can’t make good decisions without information. And data contains the information we need. But data isn’t information, even though data contains it. Think of data as the raw ore that’s in the mine. But to get the information contained therein, you need to do a lot of refining. After all, it’s called the mining industry, not the “ore” industry. The raw ore is just a component: an input to the process. Sure it’s an important component, but it’s not everything.
But notice something else. In resource mining, the process itself is pretty straightforward. By now, it’s well-known how to turn raw bauxite ore into finished aluminum. One you have the ore, the rest is maybe costly but easy to do. In data science, though, it’s not so easy. The process of transforming data into information is pretty complex, and worse still, it changes depending on the decisions you need to be making. There’s no procedure or process you can follow, as there is in aluminum smelting, which guarantees the result you want. Going from data to information to decisions is difficult, as well it should be. That’s where the value is. But given that’s the case, why are we do focused on the “data” part of the equation?
Don’t put the cart before the horse
In fact, one of the common failure modes when companies do decided to get serious about data is, paradoxically, to focus too much on the data, and not enough on the rest of the process. Who remembers the Underpants Gnomes from South Park? They Underpants Gnomes were a secret society of gnomes bent on profit. Step 1 was to collect the underpants, Step 2 they weren’t too clear on, and Step 3 was profit.
Every time I hear of a company making a big push on data and data science, I think of the gnomes. Because a lot of the time, that’s exactly what’s happening. “We need to collect the data” they say. And once they do, profit is somehow going to come out of the other end. But there’s rarely a real plan for the important hard bit in the middle: going from data to dollars.
In fact, starting with Step 1 is a full-blown mistake. You need to start with Step 3 first: how are you going to improve your profitability? Specifically, what decisions do you need to improve on, and how are you going to improve on them? By starting at the goal and working backwards, you avoid the “if you build it they will come” mindset that poisons so many corporate data efforts. Don’t be like the Underpants Gnomes. Figure out Step 3, and Step 2. Then maybe start thinking about Step 1.
Chief Decisions Officer, not Chief Data Officer
Finally we come to the crux of the matter: making “data” a C-level responsibility is pretty much exactly backward. Data is only valuable if you can get information out of it, and that information is only useful if it leads to better decisions. It’s the decisions that are the important part. So, your CDO shouldn’t be a Chief Data Officer. If “data” is in your job title, that’s the lens through which you’re going to look at every problem. As the saying goes:
To a hammer, everything looks like a nail.
What you want is a Chief Decisions Officer. That person’s job is to study and improve the company’s decision-making processes, no matter the type and domain of decision. Sure, often you’re going to need to dig into some data to provide better information for those decisions. But the important thing is the decision not the data. In the end, the difference between successful and unsuccessful companies isn’t whether they have a handle on their data or not. It’s whether they’re making the best possible business decisions. And that, without question, is something with C-level importance.
So, the next time you hear of a company (maybe even your own company) announce a splashy new CDO, ask yourself what the D stands for. Then ask yourself what it should stand for. And if you need help sorting all of this out, give us call.