Over roughly two decades, marketing budgets shifted substantially towards performance channels — activity with directly attributable outcomes — and away from brand building.
The reason was straightforward. One was measurable and the other wasn't, and in any organisation the measurable thing wins budget arguments.
The evidence assembled since suggests this was an overcorrection with real costs, and the argument is worth understanding.
The measurability asymmetry
Performance marketing produces a click, a visit and a conversion in a traceable sequence. However flawed the attribution, there's a number.
Brand advertising produces an impression on someone who may buy something in eighteen months for reasons they won't be able to articulate. There's no traceable path.
Faced with a choice between a channel reporting a specific return and one reporting nothing, the rational-seeming decision is obvious. Repeated across an industry over two decades, it produced a substantial reallocation.
The problem is that measurability and effectiveness are different properties, and confusing them is a category error.
What the research argues
The most cited work in this area analysed a large body of marketing effectiveness case studies and found consistent patterns.
Brand-building activity works slowly, produces effects that accumulate over years, and generates the largest long-term business results. Activation activity works quickly, produces short-lived effects, and is more efficiently measured.
The proposed conclusion was a budget split weighted towards brand building for most businesses — the frequently quoted figure being around 60 percent brand and 40 percent activation, varying by category.
The underlying argument is that activation converts existing demand while brand building creates it. A business that only does activation is harvesting a stock of demand it isn't replenishing, and that works until it doesn't.
The findings have been debated, as they should be, and the broad direction has held up across multiple independent analyses.
The mental availability idea
The mechanism most commonly proposed comes from a body of empirical marketing research, and it's more concrete than "brand equity".
Most purchases in most categories are made infrequently and with limited deliberation. When somebody needs something, they consider a small number of options that come to mind readily.
Brand building works by increasing the probability of coming to mind in a buying situation. Not by persuading anybody of anything — by being retrievable at the relevant moment.
That reframes what advertising is doing. It isn't argument, it's memory structure, and memory structure is built through repeated distinctive exposure over long periods.
It also explains why the effects are slow and why they decay when spending stops.
What can be measured
Brand effects aren't unmeasurable, they're just measurable differently and less conveniently.
Prompted and unprompted awareness. Survey measures, tracked over time. Blunt and directionally informative when measured consistently.
Share of search. The proportion of category search volume that includes your brand name. Available from public tools, free, and it correlates reasonably with market share in a number of categories.
Price elasticity. Strong brands sustain higher prices. Tracking your ability to hold price against competitors is a real measure of brand strength.
Marketing mix modelling. Statistical attribution at the aggregate level, which can capture long-term effects that user-level tracking cannot.
None of these produces the clean return figure a performance dashboard offers. They produce trends over quarters and years, which is the timescale on which the effect operates.
The organisational problem
Here's why this is difficult regardless of the evidence.
Brand investment costs money now and pays back over years. Marketing leaders have shorter tenures than that in many organisations, and the results of their brand investment arrive after they've left.
Performance marketing pays back this quarter and is attributable to the person who commissioned it.
So the incentives facing an individual marketing leader point towards performance regardless of what's optimal for the business. That's an agency problem, not an analytical failure, and it explains the persistence of a pattern that most practitioners privately agree is wrong.
A practical position
For most businesses, some allocation to brand building is defensible even without clean measurement, on the reasoning that a business generating no new demand has a ceiling.
The proportion should reflect your situation. Long purchase cycles and infrequent buying favour brand. Immediate need and high switching favour activation.
Measure what you can, over the timescales the effects operate on. Quarterly review of a brand investment is measuring noise.
And be honest about the uncertainty. The strongest argument against brand spending has always been that nobody can prove it works. The honest answer is that the evidence is aggregate rather than attributable, which is a real limitation and is not the same as an absence of evidence.
Distinctiveness over differentiation
One practical implication of the mental availability framing that runs against conventional advice. Much marketing thinking emphasises differentiation — being meaningfully different from competitors on some attribute customers care about.
The empirical work suggests distinctiveness matters more: being immediately recognisable as you, regardless of whether the message is unique. Customers in most categories cannot articulate meaningful differences between competing brands and buy them anyway.
Which shifts the priority towards consistent use of the assets that identify you — colour, shape, sound, character, format — and away from searching for a unique positioning claim that customers will not remember.
It also explains why changing creative approach frequently is expensive. Every change resets recognition, and recognition is the thing doing the work.