The Demand Perspective in Data

How to find the right JTBD for data work

As data professionals we’re taught to look at data from a single perspective. We see messy data, scattered across multiple systems, begging to be organized and governed.

We want to integrate it into a warehouse and build a strong foundation for future use. For us this is the holy grail. I call this the supply perspective. We’re building capacity anticipating some future demand before any has surfaced. But at the same time this is exactly like the entrepreneur who builds a product before validating the market.

What does demand look like? Aren’t questions from executives a form of demand? No, not all of them. Very few such questions constitute true demand. Most of them are simply curiosities. Making the distinction between them is a skill you must master. It just takes some time.

Now we can finally understand the source of that annoying question “what’s the ROI of data?” when the CFO sees the costs of our infrastructure build. They see the investment but can’t see the return. What happened is we got duped by the “Big Data” vendors of the 2010s who created the narrative of “store all your data in the data lake, then create the value.”

There’s another perspective I’ve been exploring and advocating recently, called of course the demand perspective. It’s not that hard to understand. The tricky part is learning to separate it from everyday questions. The demand perspective stipulates that you start by looking for jobs-to-be-done (JTBD) for data across all departments, divisions and business units.

Data is the lifeblood for many departments. Modern marketing wouldn’t function without data, especially attribution. The right piece of information can make the difference between emailing the right offer to the right customer and making a sale or emailing the wrong offer to the wrong customer and looking amateurish.

Another example I have recently discovered that the field of Revenue Operations (RevOps) seems to have subsumed many of the tasks the data team would normally do (such as integrate data across multiple systems, align sales, marketing and finance on key metrics, etc.).

When companies hire for RevOps, they know exactly what they want: predictable revenue. This is a true JTBD. The value of data is crystal clear. You can ask any CFO how much they would be willing to invest to get it and they will know for sure.

When you start from demand, you start with a problem that absolutely must to be solved in order to unlock value. How you solve it is up to you but the simpler the better. If you have to copy data from three separate systems to Excel and then run some formulas once a week, then that’s what you do. You don’t go and pitch your CEO a quarter million data infrastructure build.

That’s it for this issue.

Until next time.