The first data strategy decision I made at Broadcom didn’t involve writing any code. I noticed that the finance team used a different format for profit and loss (P&L) reporting than the accounting department, and individual financial analysts even used different formats from each other. Anyone can decide which details to show within those financial statement line items, and the analysts were deciding what to include on the fly.
Nobody used the wrong format, but nobody used the same format.
My job was to help everyone with their data. When I pointed out the disconnect in the report formats, the managers immediately agreed to make them match. Everyone at Broadcom – now the world's second-largest semiconductor company – expected the data to align, making this an easy decision for them.
Recently, an insurance company asked me to help streamline their analytics platform. I got the business and IT teams together to review potential solutions, and we focused mainly on the technical side. Then, at the end of our meeting, I asked the business teams the same simple question I had asked Broadcom managers a few decades earlier:
How many different P&L formats do you use?
The business managers looked surprised. I even heard a muffled laugh. They had dozens of different formats between them. They realized they’d been pushing IT to write more code to synchronize their reports, but they could have avoided that by synchronizing their business requirements.
That epiphany simplified the technical work for my data team.
I know it’s much easier to “solve” every problem by writing more logic into your data platform, but I want to discourage you from using that approach. Try asking business questions instead.
I have an unfair advantage in this situation because I spent the first ten years of my career as a financial analyst, and I understand how they think. I also know how data engineers think (having managed data teams for 20 years), and I can tell you with confidence that data specialists often overlook this blind spot.
If you’re humble enough to ask questions of your business partners and bold enough to keep asking them, you’re more likely to find the best solution for people’s data problems. No amount of technical training will help you identify a simple disconnect like the one in my example.
Sometimes, the solutions that business teams request from you mask the disconnected data hiding behind them. Bring your data insights to the discussion, but don’t assume you’ll fix every problem with technical solutions.
Don’t miss the data problems hiding in plain sight.
