The Truth Behind the Dashboard: Are We Measuring What Actually Matters?

We have access to more marketing data than ever before.

• Paid media.
• Content.
• SEO.
• Social.
• Email.
• PR.
• Websites.
• CRM systems.
• Ecommerce platforms.
• Brand tracking.
• Dashboards pulling various bits of it together.

Almost everything we do leaves some kind of measurable footprint... and that's both a blessing and a curse. Because having more data doesn't necessarily mean having more insight.

At the 2026 CMA Summit, I gave a talk called The Truth Behind the Dashboard, looking at something I think we could all afford to question a little more:

The most convincing reports aren't always the most truthful.

Not because somebody is deliberately trying to pull the wool over our eyes. In fact, quite the opposite.

A report can be professionally built, full of accurate-looking numbers, positive trends and encouraging conclusions, and still leave us with completely misplaced confidence.

The data underneath it might be unhealthy.

We might be measuring the wrong things.

We might be giving a metric more meaning than it deserves.



Or, somewhere between the business objective and the final dashboard, we might simply have lost sight of what marketing was supposed to achieve in the first place.

That was really what my talk was about. And it started with the best marketing report ever made...

The best report ever

Well, according to me, anyway.

I put a fictional ecommerce dashboard on screen and gave the room a few seconds to take it in, and it looked pretty cool.

Revenue was climbing while advertising spend remained relatively flat. Google Ads was delivering 7.2x ROAS. Meta was at 5.8x. Purchases were up. Conversion rate was up. Average order value was up.

Green arrows everywhere.

The executive summary confidently announced: Another record month.

And the recommendation? Increase paid media investment by 20%.

I asked the room who would be relatively happy receiving that report. Quite a few people would.

And why wouldn't you be? Everything appears to be going in the right direction. 



Except I wouldn't have made a single business decision from it.

Start pulling at the threads

Revenue was up 24%. Great!!

But what if the business wasn't trying to maximise revenue? What if it was trying to increase profit?

Conversions were up 18%. Also great..

Except I'd deliberately broken the tracking behind my fictional report. Some purchase events were firing twice.

Now CPA was falling too.

Of course it was... we'd inflated the number of conversions it was being calculated against.

Google Ads was reporting a 7.2x ROAS, but how much of that revenue did Google genuinely create? How were the other platforms attributing those same customer journeys?

And then there was the recommendation to increase advertising spend by 20%... but why?

Nothing in the report had established that another 20% of spend would generate incremental profit.

Suddenly: Another record month

Becomes: Another record reported month?

There's an important difference, but this isn't really a story about ecommerce. And it certainly isn't a story about ROAS.

Every marketing discipline can fall into the same trap

The ecommerce example just makes the problem particularly easy to see.



A content report could tell us:

A social report might celebrate reach, followers and interactions.

An SEO report might show rankings, impressions, clicks and organic sessions.

A PR report might show coverage, mentions and share of voice.

And all of those numbers could be completely accurate.

But the question is: What were we actually trying to achieve?

If content generated another 50,000 visits, what did we want those visits to contribute towards?

If engagement doubled, why did engagement matter? If our share of voice increased, what was the wider marketing objective that made that important? If we've reached another million people, were they the right people and what were we hoping would happen as a result?

This isn't an argument against those metrics. But It is an argument against giving a metric meaning simply because we can measure it.

As I said during the talk: "There is no universally good metric. There is only the right metric for the question."

And before we even get to that question, there's something more fundamental to deal with.

Can we actually trust the data?

Marketing dashboards naturally create authority.

A figure like £186,245 somehow feels much more trustworthy than saying “about £186k”.

It looks precise.

It must have come from somewhere.

There must be some logic behind it.

But precision isn't the same thing as accuracy. Marketing measurement ecosystems are complicated.

Over the years I've encountered events firing twice, conversions disappearing altogether, thank-you pages recording page loads rather than actual enquiries, GA4 and advertising platforms configured differently, ecommerce revenue not matching the ecommerce platform, old conversion actions still being used for bidding and CRM outcomes never finding their way back into marketing reporting.

And broken tracking rarely has the decency to announce itself.

There's normally no enormous warning saying: THIS NUMBER IS WRONG.

It just gives you a number.

That number gets fed into a dashboard.

The dashboard gets included in a report.

The report gets discussed in a meeting.

And eventually somebody makes a decision from it. And that's why data health has to come before data insight!

Or, as I rather reluctantly put it at the Summit: "Get your data ducks in a row."

Do the boring stuff first. Make sure the foundations work before asking the numbers to tell you a story.

Then deal with the sea of data

Even with perfectly healthy data, we still have another problem.

There's bloody loads of it!

At the Summit, all of those were floating around a ship I called HMS Just Because We Can Measure It.   

Because that's sometimes what reporting becomes. We open the platforms, look at what's available and decide which numbers should go into the report. And that is the wrong way round.

I'm not saying impressions don't matter. I'm not saying CTR doesn't matter. I'm definitely not saying content engagement doesn't matter. They matter when they help us understand something we're actually trying to achieve.

The problem comes when the availability of a metric determines what we report... which leads to probably the most important part of the whole talk.

Build downwards, not upwards

If I could get marketers to change one thing about how they approach reporting, it would be this.

Don't start with the dashboard. Start here:

BUSINESS OBJECTIVES

then...

MARKETING GOALS

then...

CHANNEL GOALS

then...

KPIs

then...

DATA & TRACKING HEALTH

then...

REPORTING

The report is the end point and not the starting point. 

If a business wants to grow market share, increase profitability, launch into a new market or defend its position in an existing one, that's where the conversation begins.

Marketing then has to establish what it needs to achieve to contribute towards that.

Only then can individual disciplines establish their role.

• Paid media might need to acquire more profitable customers.
• Content might need to build authority within a strategically important category.
• SEO might need to increase visibility amongst audiences demonstrating particular needs or intent.
• PR might need to strengthen awareness or credibility amongst a defined group.
• Social might have a role in reach, consideration, community or demand.

The point isn't that every piece of marketing needs a direct line to a transaction. It just needs a reason for existing. And once we know that reason, we can choose the measurements that help us understand whether it's doing its job.

That's a very different process from opening GA4, Meta, Search Console or whatever reporting platform happens to be available and asking: “What numbers have we got?”

Core KPIs and diagnostic metrics aren't the same thing

There's another distinction I think helps enormously here. It's that not every useful metric needs to be elevated to the same level. KPIs that refer directly to business and marketing objectives should be 'core' KPIs, where metrics that are more helpful to a channel specialist should be considered a 'diagnostic' KPI.

Take paid search, for example...

A CEO might care about profitable customer acquisition.

The marketing team might therefore use profit, POAS, new-customer CPA and contribution margin as core measures of success.

The PPC specialist may also care deeply about CPC, CTR, conversion rate, impression share and ROAS.

Those diagnostic metrics help them understand why something is happening and what they might do about it. 

The same principle applies elsewhere...

A content specialist might absolutely need to understand rankings, impressions, engagement or individual page performance. However, that doesn't automatically mean every one of those numbers belongs in the board report.

Different people need different levels of information because they're making different decisions.

That's not hiding data. Instead, it's making reporting useful to all stakeholders.

Sometimes we need more commercial context

One of the examples I used at the Summit was ROAS.

Imagine two campaigns...

Both spend £10,000. Both generate £60,000 in revenue. Both have a 6x ROAS.

If that's all we know, they look identical!

But Campaign A sells products at a 50% gross margin, while Campaign B sells products at 20%.

Campaign A generates £30,000 gross profit before advertising cost. After £10,000 media spend, £20,000 remains.

Campaign B generates £12,000 gross profit before advertising cost. After the same £10,000 media spend, just £2,000 remains.

It's the same spend... the same revenue, even the same ROAS. However, campaign A is ten times the contribution.   

ROAS hasn't done anything wrong. We've just asked it to answer a question it can't answer. And that's the bit that matters beyond performance marketing.

We do this everywhere:

• We ask traffic to demonstrate business impact.
• We ask engagement to demonstrate effectiveness.
• We ask reach to demonstrate brand growth.
• We ask rankings to demonstrate SEO success.

Sometimes those are useful proxies or diagnostic measures. Sometimes they're exactly the right measure, but sometimes they're not.

Context determines their usefulness!

This matters on both sides of the agency/client relationship

There's another dynamic worth recognising. Which is that agencies naturally need to demonstrate the value of their work. It's fact, and there's no escaping it.

Client-side marketing teams naturally need to demonstrate the value of marketing to the wider organisation. Also fact, but often slightly less cringe.

These facts create a lot of reporting.

And without anybody doing anything deliberately misleading, it's easy for reporting to gradually become about proving that activity worked rather than understanding what happened and deciding what to do next.

These, however, are not quite the same thing.

The agency reports that organic traffic increased. The marketing team reports that up the chain in their business. Everyone agrees that's positive... cool!

But somewhere along the chain we need somebody to ask: "Why did we want organic traffic to increase?"

That's not cynicism. It's curiosity. And I think good reporting needs much more of it.

Less data. Better questions.

Towards the end of my Summit talk, I changed the starting question.

Instead of: “What should we put in this report?”

Ask: “What decisions does this dashboard need to help us make?” 

A CEO might need to know:

Is marketing contributing towards the commercial objectives of the business?

A Marketing Director might ask:

Where should I invest more or less?

A channel or discipline lead might ask:

Why has performance changed, and what should I do about it?

It's the same marketing ecosystem, but these are different questions, that require different levels of reporting.

Your reports and dashboards should serve the people using them, and absolutely NOT the other way around.

Four questions to take back to your next marketing report

If you were at the Summit, these will hopefully be familiar. If you weren't, this is probably the bit I'd steal.

The next time you're reviewing an existing dashboard, receiving an agency report or building something new, ask:

1. Can I trust the data?

Am I confident the measurement ecosystem underneath it is healthy?

2. Does it connect to a business objective?

Why are we measuring this? What are we ultimately trying to understand?

3. Is this the right metric for the question?

Is it genuinely a measure of success, a useful diagnostic, or simply something we've included because it's available?

4. Can I make a decision from it?

What changes because I now know this? 

That last question is perhaps my favourite. Because insight should always lead somewhere.

Back to the best report ever

At the end of the talk, I put my original dashboard back on screen.

Nothing had changed.

• Still 640% ROAS.
• Still £186,245 revenue.
• Still 1,284 conversions.
• Still lots of lovely green arrows.

But hopefully the way everyone in the room looked at it had changed.

Instead of asking: “Is this a good report?”

We could ask:

• Do I trust what's underneath it?
• Is it measuring what the business actually cares about?
• Does it give me enough information to make a better decision? 

And I think that's ultimately the point.

This isn't about throwing away dashboards, declaring war on vanity metrics or pretending everything marketing does can be neatly tied to a pound sign. It's about being more deliberate.

Get the data foundations right.

Start with what the business and marketing are trying to achieve.

Understand the role each discipline plays in getting there.

Choose metrics capable of helping you understand that role.

Keep diagnostic information where it's useful.

And build reporting around the questions and decisions that actually matter.

Because the most convincing reports aren't always the most truthful. Don't measure more. Understand better!!