The biggest mistake Tableau creators make


Here’s something I see happen all the time in Tableau:

An analyst spends hours building a dashboard.
They want it to look sophisticated and impressive.

And instead of just showing what the data means, they reach for something complex: loads of filters, charts packed together, color everywhere.

And they wonder why nobody actually uses it.

Most of the time, the issue isn’t their Tableau ability.

The problem is almost always that they’re optimizing for “impressive” when what viewers actually need is “clear.”

So in today’s newsletter, we’re going to cover:

  • Why trying to look impressive is actually killing your dashboard’s impact
  • 4 tips to communicate clearly in every dashboard you build
  • Concrete examples you can steal and apply today

Let’s dive in.


The #1 mistake Tableau analysts make

Most people think they understand what a great dashboard looks like.

But what they build say otherwise.

Tableau analysts are notorious for saying:

“I want my audience to immediately understand the story in this data…”

And then they cram 14 charts onto a single view.

For example, we recently saw a dashboard that had:

  • A title that said “Business Performance Overview”
  • Three different color scales used across five charts
  • Filters stacked along both sides
  • A table at the bottom with 30 columns

WHAT ARE WE SUPPOSED TO DO WITH THIS?!

It has become an epidemic. Analysts optimizing for “impressive” instead of clear.

Here’s why it happens:

  • They think complexity signals expertise.
  • They believe the viewer will explore every part of the dashboard and find the insight themselves.
  • They assume more = more valuable.

All of these are wrong.

Imagine you’re an executive opening a dashboard during a 10-minute gap between meetings.

You need one answer. Fast.

Except every dashboard in your inbox looks like a cockpit.

At first, you try to engage.

Then after clicking through three dashboards like this, you stop opening them.

“Just tell me the number. Tell me if it’s good or bad. Tell me what to do.”

This is exactly what happens every time someone opens your dashboard, glances at your viz, or tries to use the report you spent two weeks building.

So here’s a good rule of thumb:
If the viewer has to work to find the point, you’ve already lost them.

Nothing kills a dashboard’s impact faster than the trying to look smart instead of being useful.

Remember that.


Tip #1: Be clear about the question.

The very first thing your dashboard needs to communicate is: what question does this answer?

It doesn’t matter how beautiful the dashboard is. Your viewer needs to know what they’re looking at before they can rely on it.

Are you showing:

  • Revenue or revenue vs. target?
  • Customers or new customers vs. returning customers?
  • Sales volume or sales velocity?
  • Overall performance or performance by region?

Notice how this subtely completely changes how you read the data.

Before anyone can trust your dashboard, they need to know what question it was designed to answer.

One question. One dashboard. That’s the goal.


Tip #2: Be clear about what “bad” looks like.

The second thing you need to communicate is: what does a problem look like here?

  • What’s the threshold that triggers concern?
  • Is this metric trending the right direction?
  • Has anyone tried to fix this before?
  • Why does it matter if it’s off?

Viewers don’t care about numbers by themselves.

They only care once they understand the context: what’s normal, what’s not, and what’s at risk.

Don’t just show them a line going up.

Tell them whether up is good or bad. Show them the target. Add the reference line. Make “trouble” impossible to miss.


Tip #3: Be clear about what to do.

“We surface cross-functional KPIs to enable data-driven decision-making across stakeholder groups.”

Huh???

That’s dashboard-speak for: we don’t know what action this is supposed to drive.

Instead: Here’s what’s behind. Here’s why. Here’s what needs attention this week.

That's better! Phew!

A great exercise: try to write the one-sentence action your dashboard is designed to trigger.

If you can’t write it in one sentence, your dashboard doesn’t have a point yet.


Tip #4: Be clear about the transformation.

Give your viewer a complete story arc:

  • [Where they start] — what situation are they walking into?
  • [The problem] — what’s off, and why does it matter?
  • [The answer] — what does the data tell them?
  • [What’s next] — what should they do now?

Stakeholders don’t open dashboards because they love data.

They open them because they need to make a decision.

So don’t make them guess. Walk them from question to answer in as few clicks as possible. If someone can’t follow the story in 30 seconds, start over.

For example:

  • Question: Are we hitting our Q2 sales targets by region?
  • Problem: Three regions are trending 15% below goal
  • Answer: The Southeast drop started in Week 6 and correlates with rep turnover
  • Next step: Flag for sales leadership by Friday

That’s a dashboard with a purpose. That’s a dashboard people use.


That’s a wrap.

Build dashboards people actually open, not dashboards that impress other analysts.

If you want to go deeper on building dashboards that communicate at the highest level, that’s exactly what we do inside Next-Level Tableau.

Live sessions. Real feedback. Training that closes the gap between where you are and where you want to be.

Join Next-Level Tableau →

– Andy

P.S. If you want to see what "clear over impressive" looks like at the highest level, join us at NLT Live — a free, 5-hour virtual conference built by analysts, for analysts. 15+ speakers. Real work. No fluff. Save your spot → https://www.nextleveltableau.com/live

ANDY KRIEBEL

I help ambitious Tableau analysts who’ve hit a ceiling

become the experts everyone relies on.
Connect on LinkedIn

Here are a few ways I can help you:

  1. Watch My Free Webinar: Learn how to use Map Layers to build advanced, standout Tableau dashboards — no hacks, no workarounds. In just 60 minutes, you’ll discover how to layer visuals with precision and design charts stakeholders actually pay attention to. → Watch the free training here
  2. Master Tableau & Advance Your Career: Join Next-Level Tableau, the most comprehensive program designed to take analysts from skilled to industry-leading. → Learn more & enroll here
  3. Get My Tableau Core Concepts: A comprehensive, self-paced course for Tableau users ready to take their skills to the next level. Dive deeper into Tableau techniques, mastering core concepts, and create dashboards that solve real business problems → Grab yours here
  4. Access My Tableau Starter Kit: This beginner-friendly course is your no-pressure entry point, built to help you understand the basics of Tableau — fast. You’ll learn exactly how to go from blank screen to interactive dashboard (even if you’ve never opened Tableau before).. → Get started here

St Matthew House, Haugh Lane, Hexham, England NE46 3PU
Unsubscribe

Next-Level Tableau

Join 10,000+ analysts getting weekly Tableau tips, tutorials, and strategies to design better dashboards, speed up workflows, and stand out in your career. Always free, always clear, always practical.

Read more from Next-Level Tableau

I have a bad neck. Not "I sleep a little stiff sometimes" bad. Full-on, wakes-me-up-at-3am, chiropractor-on-speed-dial bad. So when I started looking at pillows and found one for $230, I did what any reasonable person does. I called Eva to ask if I was losing my mind. She didn't say yes. She said: "Andy, how much did your last pair of running shoes cost?" (You don't want to know. 😊) I told her. She said: "And how many hours do they last?" I went quiet. Then she said: "You sleep eight hours a...

It didn’t happen overnight. That would have been easier to recognize. It started slowly, the way these things usually do. The Data School was expanding. I was handing off more teaching to people who’d earned the right to do it. My role was shifting. I was more involved in leadership, less on building my skills. More management, less Tableau. I told myself that was progress. And in some ways it was. But somewhere along the way, I stopped building things I didn’t already know how to build. I...

A couple weeks ago I sat down with Claude. I had no agenda, no client deliverable, no specific question to answer, no dashboard brief. I had a folder of CSV files about the city of Essen, Germany: population counts, birth and death records, physician data, business registrations, pharmacies, registered dogs. I had a lot of curiosity about what would happen if I handed them to an Claude and said, go. What followed genuinely surprised me. It was pretty eye-opening. The first thing Claude did...