There’s a conversation we’ve had with marketers many times.

The marketing reports look good.

Return on ad spend is improving. Cost per acquisition is falling. Campaign managers are hitting their targets. Individual platforms appear to be performing well.

But when you step outside the marketing platforms and look at the business, the picture can be quite different.

Revenue might be flat. Customer growth may have slowed. Margins may be under pressure. Or management simply can’t see the improvement in business performance that the marketing reports suggest should be there.

It’s a good illustration of a problem we see surprisingly often: the numbers inside the marketing platforms can be improving while the thing the business actually cares about isn’t.

And we’ve also seen the opposite.

When worse numbers can accompany better marketing

Some businesses deliberately accept higher acquisition costs or lower reported return on ad spend because they’re trying to acquire customers they wouldn’t otherwise have reached.

Paid search provides a simple example.

A campaign dominated by people searching for your business by name can produce exceptional results. Those people already know who you are. You’re already in their consideration set and, in some cases, they may already have decided to buy from you.

A business competing more heavily for generic searches, where the customer hasn’t yet chosen a brand, may report a much higher acquisition cost.

Look only at the advertising platform and the first campaign appears considerably better.

But what if the second business is acquiring more genuinely new customers and producing measurable growth period on period?

Which marketing is actually performing better?

The answer is: we don’t know yet.

And that’s the point.

Sometimes branded search can be a tax on a sale

We’ve had this discussion many times over the years: do you really need to pay for someone who is already searching specifically for your business?

Sometimes the answer is no.

If that customer was always going to find you and buy from you, paying for the branded search click may simply have added another cost to a sale you were already going to make.

But sometimes the answer is absolutely yes.

Competitors may be able to advertise against your brand. Paid search may allow you to protect valuable search territory, communicate a particular offer or direct customers towards a more useful destination.

There is no universal rule that says businesses should or shouldn’t buy their own brand terms.

What problem are we asking branded search to solve?

Start with the business outcome

The same principle applies beyond search.

A campaign with a 10x reported ROAS isn’t automatically better than one returning 4x.

An activity producing an immediate conversion isn’t automatically more valuable than activity introducing new customers to the business.

And a marketing metric becoming less efficient doesn’t automatically mean the marketing is getting worse.

The job is to understand what the marketing was supposed to achieve and then determine what the available evidence tells us about whether it did.

Platform performance is evidence. It isn’t the business objective.

That’s why we believe measurement should reflect the job the marketing was asked to do.

Because ultimately, the purpose of measurement isn’t to make the dashboard look better.

It’s to make the next decision better.