The case for branding on record

Brand work should leave an account of itself, so that anyone who paid for it can see what it changed; almost none of it does, and this is the case for changing that, with my name and the date on it.

The scene, from both chairs

The scene below is a composite. If you've bought brand work, you'll recognise it.

The work is finished. The agency presents it: the brand strategy, the new mark, the palette, the type, the voice, and a closing slide about what all of this will do for your business. The work is paid for and the files arrive. Six months later someone asks, in a board meeting or alone with the accounts, what it changed. The honest answer is that nobody knows, because nobody wrote down how your buyer saw your brand before, so there's nothing to set the after beside. The work may have been excellent. That's the whole problem: there's no way to tell, and excellent work that can't be told from ordinary work gets paid for as ordinary work.

From the other chair the scene looks different and ends the same way. The agency is judged on the day of the presentation, by how the work looks. Everything that happens to your buyers afterwards happens out of the agency's sight, and nobody sends the agency a note when a deal is won. So the agency measures the one thing it can see, which is your reaction in the room. In more than 20 years of selling brand work I have watched this happen from both chairs, and I have done it myself. Nobody in it is acting in bad faith. The habit is what the incentives do when nobody resists them.

Why the trade works off the record

There's a name in economics for what brand work is. In 1973 two economists, Michael Darby and Edi Karni, described a class of goods and services whose quality the buyer can't judge even after buying them, and called them credence goods; their examples were car repairs and medical treatment. You take it on trust that the work was needed and done well, because checking would cost more than you know how to spend. Half a century of research since, summed up in a 2020 review of the field by Loukas Balafoutas and Rudolf Kerschbamer, points the same way: sellers in such markets are judged on what can be seen, and what can be seen is the finish, the credentials and the manner. Brand work is a credence good: how it looks is clear the moment it's presented, and what it does takes months to see.

An agency judged on how the work looks has every reason to make it look better and no reason to be measured. Measuring costs money before the fee is earned, and it hands the buyer a way to be disappointed. Nobody in the trade decided this; it's simply where the incentives settle, and the trade's own clients describe the result. In 2026 the European agencies' association, EACA, and Kantar asked 141 senior marketers across 22 European markets, technology among the sectors, how their agencies perform. On creative campaigns, 18 per cent said very well. On measuring what the money returned, 11 per cent said very well, the lowest rating given to any capability in the study. The same marketers named trust as the first thing they choose an agency on, and the report calls measurement a hygiene factor, expected but rarely the thing an agency is chosen for. That closes the loop. Agencies are chosen for trust and creative work, so trust and creative work are what they get good at; measuring what the money changed is the skill nobody is hired for, so few agencies ever build it.

The clients, meanwhile, are asking. In the UK, What Clients Think 2026 interviewed 700 people who buy creative services, on behalf of the agencies they work with, and asked what a new agency could say in 30 seconds to get their interest. The first answer was evidence of commercial effectiveness, which the report calls more persuasive than anything else. Among the things those same clients most want their agencies to understand are the arguments they have inside their own companies to justify the spend. They are asking for the record. The trade, by its own account, isn't keeping one.

What working on the record means

The alternative is easy to describe. Brand work done on the record has three parts. Before the work begins, the practitioner measures how your buyer sees your brand and documents the result. While the work runs, every decision is logged with its date and its reason. When the work has been out in the world long enough to be judged, how your buyer sees your brand is measured again in the same way, set beside the first, and read back to you: what changed in your buyer's view of your brand, what didn't, and what the practitioner will and won't take credit for.

Each part costs the practitioner something, which is why the discipline is easy to describe and rare to find in the trade. The first reading of how your buyer sees your brand happens before any of the fee has been earned, and it could show that your brand isn't the problem, in which case the honest move is to say so and to sell less brand work, or none at all. Logging the decisions and the progress as they happen removes the practitioner's freedom to change course without saying so or to take shortcuts. The second reading hands you the power to judge the work on something other than how it looks. No agency gives that up lightly. Those that do would rather be judged on what the work changed than on how it looked in the room.

The record is about the work, never about you, your team or your customers. It holds the practitioner to account, and no one else. It contains the first reading of how your buyer saw your brand, every decision with its date and reason, the second reading, and the agency's own account of what the work did and what it didn't.

What the record can't do

There are two limits, and they belong here, where the promise is. A record can't prove that the brand work was the only cause of what changed in your business afterwards. Deals are won by product, price, timing and people as well as by how your brand is seen, and a practitioner who claims sole credit for a pipeline has left the record and gone back to storytelling. What a record shows is whether the way your buyer sees your brand changed, by how much, against the KPIs agreed before the work began, and what else was going on at the same time that could also explain it: a new product, a price change. That's a narrower claim than the trade usually makes, and it has the advantage of being one you can check.

The second limit is time. Buyers change their minds about a brand slowly. Les Binet and Peter Field's work on the IPA Databank, The Long and the Short of It (2013), found that brand building takes over from short-term activation as the main driver of growth after six months. In 2019 they re-ran the analysis on the Databank's B2B cases for LinkedIn's B2B Institute, working to the same six-month line, and found business brands responding to brand building much as consumer brands do. They call the B2B findings tentative: under 50 cases, most of them British and most with large budgets. Regardless, my own experience says the same. The commercial effects of brand work start to show from around month six. Measure how your buyer sees your brand any earlier and you'll record the attention a launch attracts, which fades and says nothing about whether the work changed anything that lasts. Your buyer reassesses your brand without noticing they're doing it, which is why it takes them months, and anyone offering to show you what brand work did within the quarter is showing you something else.

What a record is worth

A record exists only from the day someone starts keeping it. That sounds obvious, but it carries a consequence people miss: the value of brand work done this way compounds, because every later reading is set beside all the earlier ones, and how your buyer sees your brand becomes something you can follow over years. The loss falls on you, the owner of the brand: how your buyer saw your brand last year can't be measured this year, so every year without a reading is a year of your brand's history lost, and every decision that could have used it is made without it.

The record is yours. It's your brand's file, and it goes with you to whichever practitioner you use next, where it makes their first job easier and their claims checkable. A brand with three years of readings behind it is worth more than one without them: to you, because you know what has changed and why; to anyone considering acquiring your business, because they can check it; and to the next branding practitioner, because they start from evidence.

I sell brand work on the record and no other way, and it costs me exactly what I've described. The first reading could show a founder that the brand isn't where the losses live, and then there is less for me to sell. It costs me revenue. The alternative is the trade's habit: sell the project first, and never find out whether the brand was the problem. And the last reading carries its own risk: it could show a client that the work changed less than we hoped, and I'm the one who reads that back to them. I'd rather that than being paid as if I knew.

The same scene, with the record

Go back to the scene, six months after the work, and ask the same question: what did it change? This time the answer is a document. How your buyer saw the brand before, how they see it now, every decision in between with its date and reason, and the practitioner's own account of what moved, what didn't, and what they won't take credit for. It's a plain document, and the most useful one your brand will ever own.

I call working this way On-Record Branding, and the name is given so that anyone can use it. A discipline with a name can be asked for by name, and a firm asked for it has to accept the costs that come with it or say no. The costs fall on the practitioner. The benefit falls on you: you see what the brand work changed, or you're spared paying to have no idea whether it did.