Do you also struggle to translate data into understandable insights, let alone a story? I do. Some weeks, I sit on a goldmine of numbers and still walk out of the room knowing nobody remembered a single one of them.

Here is the uncomfortable truth about reporting data as a PM. Nobody cares about your numbers. They care about what your numbers mean for them. The dashboard is yours. The decision is theirs. Your whole job in that meeting or that email is to close the gap between the two, and most of us are quietly terrible at it.

The good news is that data storytelling is a craft, not a talent. You can get better at it on purpose. Below are 8 ways I have learned to do exactly that, mostly by getting it wrong first.

I will be the first to admit this is easier said than done. Half the time the meaning and the metric are technically the same thing. The difference is entirely in how you frame it.

You need to sound like a storyteller, not an accountant.

Compare these two:

"20% more people enjoyed our shows this year compared to last."

versus

"MAU is up 20% YoY."

Same fact. Completely different reception. The first one lands in a live presentation because a human can picture it. The second one is perfect for a tight status email where the reader already knows what MAU means and just wants the number.

So this is not about always picking the warm, human phrasing. It is about matching the language to the room. A board deck and a Slack update to your engineers should not read the same way. Know who is on the receiving end, then translate accordingly.

But even when you nail the language, you can still bury people under a mountain of updates. Which brings me to the next one.

Whether it is a slide or an email, try to focus on only 3 key figures. If you have multiple metric categories, give yourself 3 per category at most.

Why three? Because people can hold roughly three things in their heads when they leave the room. Give them five, and they remember zero. I have watched genuinely great quarters get forgotten because the PM proudly listed eleven wins, and the audience checked out at number four.

If you want to take it up a notch, try this: pick one single most important indicator and build the whole story around it. One number, supported by a few others. That is a memorable update. A wall of fifteen metrics is just noise wearing a suit.

And once you have your three, help the reader out by showing them which one matters most.

This sounds obvious, and almost nobody does it. If something is huge and growing, make it HUGE and green. If something is shrinking and that is bad, make it red and let it sit there uncomfortably.

Visual weight is a shortcut to meaning. Before anyone reads a word, the size and color of your numbers have already told them where to look and how to feel. Use that on purpose instead of letting your charting tool decide for you with its default gray.

A quick warning, though. Color carries meaning, so be consistent. If green is good on slide two, green cannot suddenly mean "raw revenue" on slide six. And go easy on the palette. Three colors with intent beat a rainbow every time.

You may think a packed slide full of charts looks impressive. It does not. It looks confusing, and confusion is the enemy of a clean decision.

Show a single graph at a time. If you need people to compare information that lives across two different views, do not cram both onto one slide and hope they connect the dots. Overlay it slide by slide instead. Show the first chart, then build the second one on top of it, so the comparison happens in front of their eyes rather than in their imagination.

The goal is that the audience never has to ask "Wait, which line am I looking at?" The moment they do, you have lost them, and you spend the next two minutes doing cleanup instead of telling your story.

No one wants to report something that turns out to be wrong.

Before any of the storytelling matters, know your certainty level, know your standard deviation, and make absolutely sure your calculations are stellar. Sanity check the query. Confirm the date ranges. Make sure you are not double-counting because someone changed an event definition three weeks ago and forgot to tell you.

You never want to be in the position of walking back good news. I have done it. You announce a lovely 12% lift, everyone is thrilled, and then a week later, you discover a tracking bug inflated the number, and the real figure is 4%. The 4% was actually fine. But now it feels like a loss, and worse, people quietly trust your next report a little less. Credibility is the one thing you cannot rebuild with a slide.

Focus on the gist. Resist the urge to pre-load every caveat, every methodology note, and every edge case into the main story.

By all means, have the details ready. Keep the appendix slide, the linked doc, and the full breakdown sitting one click away. But do not preemptively answer questions that may never get asked. When you front-load detail, you do two things at once: you bury the headline, and you signal that you are nervous about it.

Tell the clean version. Then, when someone leans in and asks, "How confident are you in that number?", you calmly pull up the detail you already prepared and look like the most organized person in the building. Anticipate the questions, just do not perform them.

A metric improved 0.5%? Cool, cool, cool. But there is a potential 2% margin of error on it, right?

If the change is smaller than your noise, it is not a result. It is a coin flip you happened to like. Reporting it as a win does two things, and both are bad. It trains people to celebrate randomness, and it dilutes the wins that are actually real.

This connects directly back to tip 5. Once you genuinely know your certainty level, you earn the right to be confident about the numbers that clear the bar, and the discipline to stay quiet about the ones that do not. "We are not sure yet, this is within the margin of error" is a perfectly strong thing to say. It is a lot stronger than claiming a victory you may have to retract.

Data should be available to anyone who wants it. Share your dashboards. Make them easy to find. Do not turn yourself into the single human bottleneck that every number has to pass through.

There is a selfish reason to do this on top of the obvious one. When more people can see the data, more people can spot something you missed. A curious engineer or a sharp-eyed designer poking around your dashboard might catch the pattern that turns into your next big bet. You cannot see everything. Open access is how you borrow other people's eyes.

The only thing to get right here is context. A raw dashboard with forty unlabeled metrics is technically accessible and practically useless. Add a short description of what each view means and what "good" looks like, and suddenly, self-serve actually works.

If you zoom out, every tip here points to the same idea. Respect the person on the other side of the data. Respect their attention with the rule of three. Respect their intelligence with solid, honest numbers. Respect their time by skipping the details until they want them. And respect their curiosity by giving them access.

Telling a story with data is not about making numbers prettier. It is about making them mean something to someone who is not you. Get that right, and people will not just understand your update. They will act on it, which is the entire point.

Do you usually struggle to tell a story with data? What is the one habit that finally helped it click for you? Sound off in the comments.