The Executive Dashboard Test: Can You Decide in 90 Seconds?

An executive dashboard has one job. It should help a busy person make a decision quickly. If you open yours, stare at it for a few minutes and close it without changing anything you were going to do, it has failed. It doesn't matter how polished the charts look or how many data sources feed into it.

That's the core truth, and it's uncomfortable because most dashboards fail it. We see this constantly, from large organisations in Brisbane to owner-operated businesses in Launceston. The reporting exists. People log in. Yet the real decisions still happen in spreadsheets, phone calls and gut feel. This article gives you a simple way to test your own dashboard, explains why most fail, and walks through a worked example of fixing one.

Why Ninety Seconds Is the Right Yardstick

Ninety seconds isn't a magic number. It's a realistic one. Think about how a general manager actually uses reporting. They open it between meetings, on a phone in the car park, or while a supplier waits on hold. They don't have twenty minutes to explore filters. They need to know whether something needs their attention, and if so, what.

If your dashboard can't deliver that in ninety seconds, the person will go back to whatever got them answers before. Usually that's ringing the production supervisor or asking the bookkeeper to send through last week's numbers. The dashboard then becomes something you show the bank or the board, not something you run the business with.

What happens in the first ten seconds

The first ten seconds decide whether anyone trusts the screen. In that window, a reader should be able to answer one question. Is today normal, or not? If they have to read axis labels, decode colours or scroll to find the headline number, you've already lost them. The top of the screen needs to say, in plain terms, how things are tracking against what you expected.

The remaining eighty seconds are for the follow-up. Where is the problem? How big is it? Who owns it? A good layout guides the eye from the headline down to the cause without the reader needing to think about where to look next.

Why owner-operators feel this more

In a big corporate, a CFO might have an analyst who interprets the numbers for them. Most Launceston businesses don't have that luxury. The owner is often the analyst, the decision-maker and the person dealing with the forklift that just broke down. Reporting that needs interpretation is reporting that won't get used. That's why a practical, low-overhead dashboard matters more in a market like northern Tasmania than in a head office with a dedicated data team.

How Good Dashboards Quietly Go Bad

Very few dashboards start out cluttered. They get that way one reasonable request at a time. Someone asks for a new chart, it gets added. Someone else wants a breakdown by region, that gets added too. Two years later you have fourteen visuals on one page and nobody can remember which three actually matter. Sound familiar?

Here are the patterns we see most often when we review existing reporting:

• Every metric is given equal visual weight, so nothing stands out as important.
• Numbers appear without context. Revenue of $412,000 means nothing unless you know whether you expected $380,000 or $500,000.
• The dashboard shows what happened but not whether it's good or bad, leaving the reader to do the maths.
• Filters and slicers hide the answer behind three or four clicks.
• Data refreshes weekly or monthly, so the decision has already been made by the time the numbers arrive.
• Different pages define the same metric differently, so margin on one screen doesn't match margin on another.

None of these are technical failures. The data connections work and the visuals render correctly. They're design and thinking failures. Someone built a report around the data that was available rather than the decisions that needed making.

Three Questions Every Screen Has to Answer

So what does a dashboard that passes the test actually look like? Strip it back and every effective screen answers three questions in order. If yours answers all three within ninety seconds, you're in good shape. If it stalls on the first one, no amount of design polish will save it.

Are we on track?

This is the headline. It should compare actual performance against a target, budget, forecast or last year. The comparison is the point. A single number in isolation is trivia. The same number shown as 8% below budget is information you can act on. Use plain status indicators sparingly, and only where the thresholds have been agreed in advance.

If not, where is the gap?

Once the reader knows something's off, they need to locate it fast. Is it one product line, one site, one customer, one shift? This section should break the headline down by the two or three dimensions that genuinely explain variance in your business. For a manufacturer that might be line and product. For a tourism operator it might be booking channel and tour type. Pick the dimensions that explain the most, not every dimension you happen to have.

What should happen next?

This is where most reporting stops short. The best dashboards point toward an action, even if indirectly. That might be naming the owner of a metric, showing the trend so you can see whether a problem is worsening, or flagging which orders or jobs are at risk this week. You're not trying to automate the decision. You're trying to shorten the distance between noticing a problem and doing something about it.

A Worked Example From a Launceston Food Processor

Let's make this concrete. The following is an anonymised composite based on the kind of engagement we see regularly with food and beverage businesses in northern Tasmania. The details are simplified but the shape of the problem is real.

Where they started

A mid-sized food processor in the Launceston area was running three separate reports. Production had a spreadsheet tracking yield by batch. Quality had another tracking rejects and customer complaints. Finance had monthly margin reports from the accounting system that arrived about three weeks after month end. The owner had an executive dashboard that pulled headline revenue and volume, but it didn't connect any of this together.

The result was predictable. When margin dropped in a given month, nobody could say why until weeks later. Was it lower yield? Higher rejects? A shift in product mix toward lower-margin lines? Rising input costs from growers? Each department had part of the answer and spent the first half of every management meeting reconciling numbers instead of discussing them.

What changed in the rebuild

The first step wasn't building anything. It was agreeing on the handful of decisions the owner actually made each week. It came down to four. Should we adjust the production schedule? Do we need to talk to a grower about raw material quality? Is a particular product line still worth running? Do we need to chase pricing with a key customer?

From there, the dashboard was designed backwards from those decisions. Yield, quality and margin were joined at the batch level so they could be read together. Margin was estimated weekly using standard costs, rather than waiting for the accounting close. Targets were agreed for each product line so variance could be shown, not just raw totals. The data refreshed daily from the production system and overnight from accounts.

• Top row: weekly gross margin against target, with yield and reject rate beside it.
• Middle section: margin variance broken down by product line and by raw material supplier.
• Bottom section: the five batches with the worst yield this week, with the shift and supplier attached.
• A single trend line showing twelve weeks of margin, so the owner could see direction, not just a snapshot.

The ninety-second walkthrough

Here's what a Monday morning now looks like. The owner opens the dashboard and sees weekly margin is 3.2 points under target. That's the first ten seconds. Their eye drops to the breakdown and they can see most of the gap sits in one product line. Within that line, one supplier's batches are showing noticeably lower yield. The bottom table shows three of the five worst batches came from that supplier's last delivery.

That took well under ninety seconds. The decision is obvious. Call the supplier, check the raw material specs, and consider whether to adjust the schedule while it's sorted out. Previously, that same insight would have surfaced weeks later, if at all, buried in a monthly reconciliation. The data hadn't changed. The way it was arranged had.

Trade-offs, Edge Cases and Honest Limits

The ninety-second test is a useful discipline, but it isn't a law of physics. There are situations where you'll need to bend it, and pretending otherwise would be doing you a disservice. Here are the ones we run into most often.

When one business has several business lines

This is common in Launceston. An owner might run a manufacturing operation alongside a retail outlet, or a farm alongside a cellar door and accommodation. A single screen can't do justice to businesses with completely different economics. The answer is usually a top-level summary that shows each line's status in one glance, with a dedicated page for each line underneath. The summary passes the ninety-second test. The detail pages are where you go once you know which line needs attention.

The trap is trying to force a common set of metrics across very different operations. Occupancy means something for accommodation. It means nothing for a food processing line. Let each business line have its own measures, and only roll up what genuinely compares, like revenue, margin and cash.

Seasonal and lumpy businesses

Tourism operators in the Tamar Valley and agricultural businesses across the north face heavy seasonality. Comparing this week to last week is often meaningless. A quiet week in July looks alarming next to a busy week in January. For these businesses, compare against the same period last year, or against a seasonal forecast. Without that, your dashboard will cry wolf every winter and nobody will trust it by spring.

When ninety seconds genuinely isn't enough

Some decisions shouldn't be made in ninety seconds. Annual budgeting, capital investment and restructuring all deserve proper analysis. The executive dashboard isn't the place for that work. Its role is to flag that something warrants deeper investigation, then hand off to more detailed reports or a proper analysis session. Confusing the two is how dashboards end up bloated. People try to make the summary screen do the job of an analytical workbook.

When you can't trust the data yet

This one deserves blunt honesty. If your underlying data is inconsistent, a beautiful dashboard will just help you make wrong decisions faster. Common culprits include manual data entry on the factory floor, product codes that changed halfway through the year, and accounting categories that don't match operational categories. Sometimes the right first project is cleaning up definitions and data capture, with the dashboard coming second. It's less exciting, but it's the work that makes everything after it credible.

• Agree on one definition for each core metric and write it down where everyone can see it.
• Identify which numbers are entered manually and how often they're wrong.
• Check that operational systems and accounting systems use matching codes for products, sites and customers.
• Start with fewer metrics you trust rather than many you have to caveat.

Running the Test on Your Own Reporting

You don't need a consultant to find out whether your current reporting passes. You can run the test yourself this week. It takes about half an hour and it's usually revealing.

Pick someone who makes real decisions in your business, ideally not the person who built the dashboard. Give them a specific scenario, like 'Is this week going well, and if not, what would you do about it?' Set a timer for ninety seconds and watch them use the dashboard without helping. Note where they hesitate, where they click, and what they say out loud. Then ask them what decision they'd make.

If they can't name a decision, or they name one but couldn't explain how the dashboard got them there, you've found your gap. Pay close attention to the moments of hesitation. Each one usually points to a specific fix, whether that's a missing target, an unclear label, a buried number or a chart that shouldn't be there at all.

Once you know where the friction is, the fixes tend to fall into a few categories. Add context through targets and comparisons. Reorder the layout so the headline sits at the top. Remove visuals that don't support a decision. Speed up refresh where timing matters. Join data across systems where the answer depends on more than one source. Most of these changes are modest in effort and large in impact.

If you're running Power BI already, much of this can be done within your existing setup. If you're still in spreadsheets, it's worth looking at how a tool like Power BI handles joined data and scheduled refresh. Our Power BI Brisbane team works with businesses well beyond Queensland, and the principles are the same whether you're in Fortitude Valley or on the banks of the Tamar.

Your Next Step

Run the ninety-second test on your current reporting before you change anything. It costs nothing and it'll tell you exactly where the problems are. If you'd like a second set of eyes, Roar Data works directly with owner-operators and leadership teams in Launceston and across Tasmania to design reporting around the decisions you actually make. We're hands-on, we keep the overhead low, and we're comfortable working across multiple business lines in one engagement. Take a look at our executive dashboard development service to see how we approach it, then get in touch for a straightforward conversation about what your dashboard should be telling you in ninety seconds or less.

For the wider picture of how this work is run, our Power BI consulting approach across Australia sets out the stages and what each one is meant to settle.