Everyone was a data analyst

I once inherited a team where almost everyone was titled Data Analyst or Data Scientist, regardless of what they actually did all day. Some were building pipelines. Some were shipping models. Some were making dashboards that the business genuinely ran on. All of them had the same two words on their badge.

That is not a naming problem. It is an accountability problem with a naming symptom.

If the title does not describe the outcome, three things follow. Nobody can be held to a standard, because there is no agreed standard for a job that is really four jobs. Nobody can grow, because the ladder has one rung. And nobody outside the team can work out who to ask, so they ask everyone, which is how you get reactive order-taking.

So I rewrote every job description to the outcome I actually needed: BI specialist, data engineer, analytics engineer, ML engineer, AI engineer. Real roles with real definitions of what good looks like, including the things each role should stop doing.

It is a slow, unglamorous piece of work and it does not demo well. It also changed more about how the function performed than any platform decision I made in the same period.

If you are about to buy a tool to fix a delivery problem, check the job descriptions first. It is cheaper and it is usually the actual answer.