You Can’t Advertise Your Way Out of This
The Disconnect: It’s not you…it’s the budget black hole
When a measurement becomes the target, it fails to be valuable any more. Goodhart’s Law
Show me the money! Rod Tidwell, 1996
There is something peculiar happening in modern retail. We have become extraordinarily good at measuring marketing while remaining surprisingly uncertain about whether it is making businesses meaningfully better off.
A retailer can now tell you, with remarkable precision, how many people saw an advert, clicked it, visited a product page and eventually bought something. The sale can be attributed to a channel, a campaign, sometimes even an individual creative. Dashboards make the whole thing look satisfying…even scientific.
Then somebody asks what should be a simple question: did we actually make more money?
Can you hear the silence? People looking at each other? This is often where things become less certain.
How far did each pound of spend go? Or did it vanish down the black hole…with an inexplicable shrug?
It’s not because digital marketing does not work. Quite the opposite. We have gradually allowed the measurement of marketing activity to become confused with the measurement of business growth. As Goodhart’s Law says, “When a measurement becomes the target, it fails to be valuable any more.”
Traffic can rise. Orders can rise. Revenue can rise. ROAS can improve. Yet by the time discounts, acquisition costs, returns, fulfilment and customer service have all taken their share, the commercial improvement can be rather less impressive than the marketing dashboard suggested. The ‘M’ word, margin. Less so an elephant in the room, more a hyena and leaner than you imagined.
The business has become busier. But better?
This distinction becomes more important as retailers spend more money trying to generate demand. UK digital advertising expenditure reached £40.5 billion in 2025 according to IAB UK, while retail media continued to grow rapidly. We now have more inventory, more automation, more customer data and more sophisticated attribution than retailers could have imagined twenty years ago.
The Dark
And yet, the abundance of measurement can create a strange side effect. The things that are easiest to measure begin to receive the most attention.
This is a fairly common human tendency. If you lose your keys in a dark street but search for them beneath a lamp because the light is better, the search is very measurable. It is also unlikely to help. This is called The Street Light Effect. We need to fumble around in the dark!
ROAS can sometimes become retail's lamp post.
There is nothing inherently wrong with Return On Advertising Spend. It answers a useful question: how much revenue was attributed to a given amount of advertising expenditure?
But there are several other questions it cannot answer on its own:
- Would the customer have purchased anyway?
- Were they genuinely ‘new’?
- Was the order profitable after the discount?
- Was it subsequently returned?
- What did fulfilment cost?
- Did another channel create the original demand?
- Will that customer come back without needing to be acquired all over again?
- Will they tell their friends and family?
- Will she ever call me back?
Attribution tells us who received the credit for a sale.
Incrementality asks whether the sale would have happened without the activity.
Profitability asks whether the sale was worth creating in the first place.
These sound like subtly different versions of the same question. Commercially, they are not.
The Illusion
It is perfectly possible for a campaign to have excellent attribution and terrible incrementality. It is equally possible to generate genuinely incremental sales which are barely profitable.
This is where the retail growth illusion begins. Don’t be fooled!
Imagine that marketing successfully creates demand and sends a customer to a retailer's website. From the media team's perspective, the job is working. But the customer does not experience a retailer. They experience a brand.
They arrive on a site where the navigation is confusing. They search for a product and get poor results. The product page does not answer an obvious question. Their preferred size is unavailable. Perhaps they eventually purchase because a discount code persuades them to overlook some of the friction.
The campaign has technically succeeded, but wouldn’t you rather it hadn’t?
The commercial system has [sort of] worked but the feelings that are stirred aren’t valuable.
The order may then cost more to fulfil than expected, generate a customer-service enquiry or eventually need to be refunded. Even when the transaction goes perfectly well, the retailer may do (likely) nothing useful with the relationship afterwards. Six months later the same customer clicks another advert, and the business pays to acquire them again (so the customer is rented not owned!)
None of these events is dramatic enough to set off an alarm. They are not under the “Street Light”.
This is what makes them expensive.
The Sum, the Parts, and Lots of 1%
Retailers rarely have one enormous commercial failure into which millions disappear. They have hundreds of small inefficiencies spread across marketing, commerce, CRM, operations, technology and customer experience. This is why connecting systems is SO important.
The money has not really vanished.
It has been converted into invisible friction.
And much of that friction exists because of the way businesses are organised.
Retailers have departments and think in silos. Customers don’t think this way and just feel the ONE brand.
A customer does not know that paid media is managed by one team, ecommerce by another, CRM somewhere else and logistics by an entirely different function in a warehouse somewhere totally different. They do not care that the website agency has never met the performance agency, or that the CRM platform cannot see the same information as the customer-service team.
From their perspective there is only one experience.
This creates one of the stranger possibilities in modern commerce: every department can hit its individual target while the customer experience, and ultimately the economics of the business, becomes worse. The sum becomes LESSER than its parts.
Marketing can improve ROAS. Ecommerce can protect conversion. CRM can increase campaign frequency. Operations can reduce fulfilment costs. Finance can protect margin.
Every decision can be rational in isolation.
The combined result can be completely irrational.
We have become very good at optimising parts of the machine without always asking what the machine is even supposed to do.
This matters particularly when retailers begin to stagnate.
There is a natural tendency to assume that slower growth means the business needs more demand. So another channel is added, budgets increase, promotions become more aggressive or another agency is asked to find an audience which the previous one somehow missed.
Sometimes that is exactly the right thing to do.
Oftentimes it is the commercial equivalent of turning the radio up because the engine is making a worrying noise.
The biggest opportunity may well be somewhere else entirely:
- Improving product information can make every source of traffic more valuable.
- Better navigation can improve paid, organic, direct and email performance simultaneously.
- Better stock visibility can stop campaigns generating demand for products people cannot actually buy.
- Reducing returns can improve the profitability of sales without acquiring a single additional customer.
As James Clear talks about in Atomic Habits, these little 1% gains can together make a huge compound difference.
And better retention can have an especially powerful effect because there is a significant difference between acquiring a customer and repeatedly renting one.
Accentuate the Best, Connect them Together
This is why marketing performance should rarely be viewed as a marketing problem alone. It never is.
Marketing sits at the beginning of a much larger commercial chain, and its apparent performance is affected by almost everything that happens afterwards holistically.
AI makes this argument even more important.
There is understandably enormous excitement around the ability of AI to create more content, analyse more audiences, personalise communications and optimise campaigns more quickly.
However, AI has one characteristic worth remembering: it is very good at accelerating whatever system it is given.
A satnav is fantastic when the destination has been entered correctly. If it has not, its main contribution is helping you get lost much faster.
AI can optimise advertising towards a product with poor margin. It can produce hundreds of variations of a proposition that is fundamentally uninteresting. It can personalise discounts for customers who would happily have paid full price. It can become exceptionally efficient at driving transactions which the wider business does not particularly benefit from.
The technology is not at fault. It is doing exactly what it has been asked to do. We need to also use empathy, creativity, and gut…consistently.
The more automation enters commerce, the more important it becomes to decide what the entire system is trying to optimise for.
This is the thinking and feeling behind Pivotal's mission towards Creative, Connected Commerce.
The point is not that every retailer suddenly needs a vast transformation programme, nor that every department needs to be merged into one. It is simply that growth has to be understood as a connected system…as a brand.
Demand has to be attracted. That demand then has to convert. The transaction has to become a relationship worth retaining (and then be retained). And the technology and operations underneath the business have to allow all of this to scale profitably.
The interesting part is what happens between those things. How they best interlink.
Creative needs to understand what can profitably be sold. Marketing needs to know what happens after the click. Ecommerce needs to fulfil the promise that marketing made. CRM needs to understand what the customer bought and what happened next. Operations needs to know when demand is coming. Finance needs to help everyone understand whether the resulting growth is actually valuable.
It is less glamorous than finding the next breakthrough channel.
It is considerably more valuable.
Which brings us back to that original pound of marketing budget.
Instead of asking only which platform received credit for the resulting revenue, follow the pound through the entire business…mindful of the pesky black hole!
Did it create genuinely new demand? Did that demand convert? Did the transaction make money? Could the order be fulfilled effectively? Did the customer have an experience worth repeating? Did the relationship become more valuable afterwards?
Or did the original pound eventually come back disguised as a discount, a return, a support cost and another acquisition fee six months later?
The black hole in retail growth is rarely marketing itself.
More often, it is the space between things.
And once retailers start looking there, the answer to slower growth may not be to spend more money.
It may simply be to be mindful of the black hole, and spend budget better across endeavours connected meaningfully together.