4 awkward stakeholder conversations you have to master if you want more money and respect

Learn how to talk like a leader, even if you’d rather hide in your code

This mini-course is part of the Level-up data engineering playlist. Click here to explore the full series.

Hi fellow data nerd, Yordan here,

There’s a moment every data person secretly dreads.

You’re buried in code, trying to get this damn model to stop exploding, when you get a ping:

Hey, can you send a quick update for leadership?

Your heart sinks.

Because sure, the update is “nothing’s working yet.” But saying that feels like admitting failure. And saying too much will confuse them. You stare at the cursor. Dread creeping in. One Slack message feels harder than writing 300 lines of Spark logic.

Been there?

I have. More times than I care to admit.

This has prevented me from selling ambitious projects more than once. I wasn’t missing a technical skill. I was missing communication muscles.

That’s what this month is about

It’s not a fluffy “soft skills” course. This is a field guide for any data person who’s sick of being overlooked, undervalued, or misunderstood.

You’ll learn how to:

• Summarize what someone wants so they feel heard

• Explain trade-offs in plain English

• Say no without pissing people off

• Write updates that build trust (even when nothing moved)

Each week, we’ll focus on one real-life conversation. The kind that happens in Slack. Or Zoom. Or hallway run-ins you didn’t plan for.

By the end of the month, you'll “know” how to manage stakeholders and feel it in your fingers.

Like muscle memory.

Because here’s the truth:

The people who get promoted don’t always do the best work. They’re the ones who make their work easy to see.

This is how you do that without selling out, faking it, or becoming someone you’re not.

---

How to work with this mini-course

Reading the article alone would take you over 20 minutes. So:

• Bookmark this guide and set a reminder to revisit it weekly.

• Skim the entire article once to understand the big picture.

• Each week, complete the exercises before applying them to your own projects.

• Share your progress on LinkedIn to reinforce learning and expand your network.

Take your time. Don’t rush to apply everything at once. Master each step before moving to the next.

Also, you will need about an hour to read the whole thing and write the code at once. It’s much easier to spend 15 minutes per week!

---

Week 1: the “here’s what I heard” recap

There’s this moment that happens all the time on data projects:

You leave a stakeholder meeting, and no one’s sure what got agreed on.

One person thinks you’re building a dashboard. Another thinks you’re doing deep attribution analysis. Someone else thinks you’re fixing tags.

You’re trying to get out of the meeting without looking dumb.

So you nod.

And when you get back to your desk, your brain feels like a browser with 78 tabs open. No clear ask. No owner. No priority. Only chaos.

It doesn’t have to be like that.

You don’t need to be a master facilitator. You just need to build one habit: summarize what you heard. Out loud. Back to them. In plain English.

This one habit will:

• Cut your rework in half

• Make people trust you more

• Make your own work feel less like a guessing game

And it’s so simple most people skip it.

Why this works

Stakeholders aren’t trying to test you. They’re under pressure. Their boss wants answers. The deadline’s too close. The launch feels risky. They ramble. You misinterpret. Shit breaks.

But when you take 30 seconds to reflect back what they said, it does two things:

1. It makes them feel heard: people trust people who “get it”

2. It makes sure you’re not solving the wrong problem

That second one is a game-changer. Because there’s nothing worse than doing 2 weeks of work only to hear:

Oh, I thought you were going to pull this number instead…

What a good summary sounds like

Let’s say a stakeholder drops this on you:

We need to clean up this marketing report. The metrics are all over the place. I think the new tags broke something. Leadership needs this by Friday for the ops review.

Here’s what most people do:

Okay I’ll check the tags.

That’s not wrong, but lazy. You’re not showing understanding. You’re jumping to fix mode.

Now here’s a better version:

Got it. Let' confirm this is for Friday’s ops review, and you need accurate metrics more than speed. Sounds like the issue started with some new tags, so I’ll start there. I’ll update you by end of day.

Why this works: