If you’re a data analyst right now, or anyone working in business intelligence, you might be eager to get your hands on the latest and greatest machine learning algorithms. You might be trying to sell your business on the benefits of predictive analytics and patiently waiting for the go ahead.

If this is you, you’re in a privileged position.

Not only do you have the whole world of machine learning ahead of you, but all your experience delivering insights to stakeholders means that you have good karma in the bank.

There are so many ML people out there who get angry whenever they’re asked to produce a report or work on a spreadsheet. Sometimes these requests are misguided or driven by politics, but most of the time it’s an opportunity to do your real work.

Machine learning projects can take a long time. They’re inherently experimental. The features you spend weeks engineering from hard-to-come-by data can end up being statistically useless. After a month, or two, or three, it could turn out that your problem isn’t linearly separable at all. There are a lot of failures.

If you work for a company your job is not to fail. Researchers fail, entrepreneurs fail. They call failing learning and they’re right to do so. But when you have an active hand in which decisions are made, a simple model, a basic analysis, even a pie chart are always better than a failed attempt at a complex solution.

You might think that the users of your work are your stakeholders. And maybe they are. But there are other users, too. As a data scientist, analyst, ML engineer, whatever, the decisions we help to make will decide where budget money goes. We help decide who gets laid off. We have a say in which customers will get spammed by our marketing team.

No one understands what happens inside a neural network. But only people with the right training can interpret a regression. Just because it’s simple to you, doesn’t mean it’s simple to anybody else.

If you’re a data analyst excited by all the opportunities in machine learning, don’t leave behind all the skills you’re using right now to turn the basics into actionable information. Remember the users.