SEO froze brand strategy in 1961
SEO inherited advertising's obsession with measurement and its language of differentiation — and then stopped reading. Brand science spent fifty years proving how brands actually grow. AI search didn't invent new rules; it made the old ones impossible to ignore.
We thought we knew the game. The shortcuts. The way to Google’s heart. And then ChatGPT entered stage right and showed us we had all been wrong, for a long time, faster than we could blink. It stripped us of the fallacy of our metrics. Pulled back the curtain of the biases behind — likely — every model and algorithm trained. There is no neutral; there’s only opinion.
So we’re polar bears, losing our solid ground beneath our feet and scrambling to find purchase on terrain shifting beneath our feet faster, sometimes, than we can catch it. What do we hold onto? How do we move forward?
The answer isn’t new. It isn’t even ours.
We go backwards.
Because here’s the thing. We thought, as SEO’s, we invented something new. The Internet was new, so of course, finding ways to manipulate the Internet was new. Wrong. All the way wrong.
Here’s the interesting part: our obsession with metrics kept us stagnating. We inherited Claude Hopkins’ 1923 demand that every claim be measurable: coupon tracking, split testing, the birth of empirical advertising. And we inherited Rosser Reeves’ 1961 Unique Selling Proposition (USP), the “what makes you different” question that opens every brief we’ve ever written.
Then we stopped reading.
And brand science moved on without us. To metrics and measurability that looked, at the surface, more fluffy. Levitt, Ogilvy, Ehrenberg. All pointing to the concept of brand and how to best go about building it.
When we were stuck on meta keywords and bidding to buy the best exact-match-domain to build our affiliate website on, brand science had evolved to concepts we weren’t even considering because it was outside of our controllable slice of the Internet: penetration, mental availability, distinctive brand assets.
We couldn’t balance our metrics-led thinking with the truth that brand will always be an intangible, because brand sits deep in us — there is no rational man, in economics or marketing. And that is a contradiction this numbers-forward discipline of SEO is not used to sitting in. Humans make decisions emotionally, and justify rationally.
Because to quantify an emotion into any number is messy at best. The measurements will never be perfect, but just because something doesn’t fit neatly into a box doesn’t mean it’s not worth pursuing.
Because emotion sticks.
Some of us knew and dismissed it. Most of us didn’t bother to look. We were data-driven. The brand team was the opposite, all hand-waving and smoke and mirrors. The brand team was a barrier in the way of publishing something on the website quickly. They were an obstacle to overcome, in the 2010’s, for most SEOs. It was not a team to pay attention to. We were above it.
We told our clients they didn’t need to be brands as long as they ranked. Visible but not superior. Oh, how we’re eating crow now.
Because that worked when Google couldn’t tell the difference. Google can finally deliver on the promise it made twenty years ago. Plus, AI surfaces have been trained on data already shaped by companies who never stopped being brands on the Internet. Then add a Google ecosystem that could, in 2022, disseminate a brand from an exact match domain masquerading, and every colour in between.
So now we’re going back to the future to learn a discipline that as recently as three years ago was below most of our notice, proudly so.
Brand, Built Into Our Thinking
Pay attention to these four brand concepts and rationales. They shouldn’t hurt too much to integrate into our thinking while building back a practice that should help you keep a job in an industry quickly shifting. We ignored a few things in favour of easy data, and we should roll that thinking back quick smart:
- Penetration over loyalty: Ehrenberg in ‘69 found brands grow by reaching more customers rather than getting existing customers to buy more. Focusing on the existing funnel and ‘qualified traffic’ as the primary or only tactic is a red herring. You can’t loyalty your way out of small.
- 95-5 and mental availability: Only 5% of buyers are in-market at a given time. We win the rest by reliably being present in their ecosystem before they make a decision. We need to do the work to be present in people’s lives; their media, their influences, their analogue life, something last-click attribution could never measure.
- Category entry points: Purchase triggers are situations, not keywords. We need to go back to the 7 w’s, grounded in our qualitative data: sales, support, subject matter experts, customer reviews, focus groups. See the whole messy journey, rather than the long-tail keyword we’re trying to optimise for. Why, when, where, with whom, with what, feeling what, while.
What all of these ladder into is this: none of anything you do online matters if people can’t remember you. We need to do the work to actually be a sticky, emotion-generating brand. And that isn’t something we can shortcut.
Distinctive Brand Assets
Every competitor analysis you’ve ever run probably has a unique selling proposition, a USP. And once you work for enough clients, you might notice how similar those USPs all sound. We’re simple. We’re fast. We’re reliable.
It’s Reeves and his research on it, straight down the line. That inherited, data-led framework. The one brand science has long since moved beyond, and we should too.
Byron Sharp argued differentiation matters less than distinctiveness. Instant recognition beats unique positioning. Jenni Romaniuk wrote the seminal book on building distinctive brand assets in 2018. Visual, verbal, sonic, more. The elements which make a brand identifiable before you actively recognise you’re doing it, without it needing to be named.
The look of Coca-Cola. The sound of Netflix. The McDonald’s arches. You don’t need the brand name to imagine any of it.
That’s the point.
Even in 1963, Ogilvy got there. Every advertisement, he noted, should contribute to “the complex symbol which is the brand image.” Coherence, Jono Alderson’s word when he talks about how brands survive in AI search, is a sixty year old idea in a new medium.
And. We’ve actually been doing distinctive brand asset work for a while. We didn’t call it that, so we did it badly, because we thought what we were doing was solving for the USP.
Organisation schema. NAP. Product markup. Author bios. Wikipedia entries.
Look at consensus scoring, where Google is looking for the baseline agreement on facts. It’s another version of this: all your brand asset infrastructure translated for and ingested by machines, in one place, and judged. Knowledge Graphs are what happen when you consistently signal your brand, have those recognisable, consistent assets in the first place, and leave no room for ambiguity.
Think about what most sites look like during an audit. The About page says one thing, the schema another, the Twitter logo left un-updated from three rebrands ago, the top press release mentions a product that got sunset five or twelve or twenty years ago, all with inconsistent category descriptors and a slightly different reference to the company each time a new PR firm was brought on board.
Humans figure it out. Machines need to work to understand the fuzz. The more inconsistency, the more the work compounds as more vertices need to be cross-referenced. It quickly becomes a hobble to clear brand identification.
We’ve been asking the finite question: ‘what makes us different from our competition?’. We should be asking the infinite question: ‘what makes us unmistakably us?‘
Where this leaves us
Four findings. Read them together, and the pattern is embarrassing.
We’ve chased the easy audience: high-intent, in-market, ready-to-buy.
We’ve measured the easy outcome: last click today, rather than presence in tomorrow.
We’ve worked to the easy measure: the query, rather than the moment.
We’ve sought the easy definition: different, rather than entirely ourselves.
Each choice, in the moment, felt like a good SEO best practice. And each was the short-term, reliable, money-earning choice for the state of search. But put them together and they missed the point. The long-term, generous view of what the company could become; of what a good, lasting, profitable business should be.
Idealistic, perhaps, but I fundamentally believe a good business is like any good relationship: entered into with the intention to build rather than demolish, go for the long haul, and to contribute well to the whole. A good brand does all those things. Many times, they do them explicitly, when you look at corporate social responsibility, supply chain transparency and ecological initiatives. We lose that intangible, ephemeral whole of a brand when we reduce what we measure ourselves on to a handful of bottom-of-funnel numbers. The human in it disappears.
And yet, the human in it has fundamentally shaped the Internet over the last twenty years. Google has built and rebuilt its search engine in part as a response to our needs as we mature in our Internet usage, and it has never stopped in its goal to, essentially, become the digital version of word-of-mouth, with all of your neighbors’ latent understanding of the good brands baked in.
And now, we can’t hide behind the algorithms because the algorithms are evolved enough to pattern-match to those real world distinguishers of brand. The things we thought didn’t matter for fifteen plus years definitively do.
Brand is the Base, Not A Ranking Factor
Brand isn’t a signal you optimise for. It’s what the signal has always been measuring. Keyword ranking has always been a lagging indicator for brand salience, whether brand or non-brand.
We got confused for twenty years because the surface of the Internet was big enough to feel like the whole world. It wasn’t. Google has always been a proxy. A very good proxy, yes, but a proxy — and now one of many. Pulling from the cultural zeitgeist it can read and disseminate and return back to us, ranked, returning brands they think are a part of the conversation because they’re already in the consideration set of the analogue world and have been translated, consistently, widely, accurately, to the digital.
And to make it more brutal — the AI-visibility vendors are already selling the wrong answer. Every dashboard is peddling brand mentions as the measure, with an aggregate visibility score to report on. Same trick, old horse. Take something opaque, simplify it to its nearest accessible metric, sell the metric.
And just like SEO drifting to keyword rankings — it’s easy, and fundamentally misses the point. Brand salience and mental availability are already baked into the results they’ve served, so measuring brand mentions is like asking for the recipe after the cookie has been put in the oven. It’s playing the finite game against everyone else.
So take a page from Apple’s book, perhaps. Measure yourself instead on progress over time. Am I doing better than I was before? Maybe measure with sales. Or with excess share of voice. Maybe with traffic, still. But within an entirely different frame, and clearly communicated assumptions. The simpler, the better. Your CEO doesn’t care about the detailed modelling you made that takes you 45 minutes to explain. You lost them after 60 seconds.
We want to keep them engaged and investing in digital, in whatever form that may take, for the next fifty years instead.
What changes Monday morning
Brief for the moment, not the query. Start with the person, the qualitative. The sales calls and customer interviews.
Audit for distinction and consistency, rather than USP or schema. Watch the drift across surfaces. Flag it, track it, chase it.
Move the investment. Binet and Field’s IPA data, drawn from 996 case studies across 700 brands, found the optimal split for consumer brands is roughly 60% long-term brand building, 40% short-term activation, though this is fuzzy and depends on what study you look at, who you talk to, and what industry you’re in. Their B2B follow-up with the LinkedIn B2B Institute in 2019, for example, put B2B closer to 46:54, still nearly half on brand.
Most SEO budgets run 100% activation. So honestly: if you can shift 10% of your budget, it’s a start. The hardest part of this ask isn’t measurement. It’s convincing whoever holds the budget to reallocate a quarter or a half of it to work that won’t show ROI for eighteen months or longer. That conversation is harder than any technical audit you’ll ever run. It’s also the one that actually matters.
None of this is new. And we need to integrate it like it’s not.
The measurement bridge we already have
Share of search, developed by Les Binet and James Hankins in 2020.
Hankins followed up in 2021 with cross-industry IPA think-tank data. Across categories, countries, and languages, Share of Search represents about 83% of Share of Market on average. It’s a leading indicator with a 6-to-24-month lead time. When Share of Search goes up, market share follows. When it goes down, market share follows.
There’s a parallel finding from Binet and Field’s earlier work on Share of Voice. Their IPA data in The Long and the Short of It found ten points of Excess Share of Voice (your SOV minus your Share of Market) correlates with roughly one point of market share growth per year. The ratio is directional and varies by category, brand size, and whether you’re a challenger or an incumbent, but the finding has held across replication.
Two different metrics, same underlying pattern. The gap between visibility and market presence is the strongest growth signal you have.
Fair warning on the methodology. Google Trends is the data source Binet and Hankins used, but the metric isn’t a one-click pull, and is not reliable. It’s not something I’d advise trying to replicate today.
None of this is a reason not to run it. It’s a reason to run it carefully. Paid tools (SimilarWeb, Branquo, Kantar’s validation methodology) handle the normalisation and small-brand data problems, and are really the only way to do it in the current environment.
Even with that, measurement is broken, and at best relative. I could go on about why it’s broken, and GDPR is only a part of it. Google Trends, specifically, is also free. Always be wary of free data.
But the point isn’t that this metric is new. The point is that we’ve been sitting on top of it the whole time and reporting on it as a vanity metric called “branded search volume.”
Brand science isn’t monolithic. Sharp and Binet disagree on details. Ritson has spent years defending things Sharp dismisses. The IPA Awards have surfaced case studies where segmentation worked despite Sharp saying it shouldn’t. The empirical core holds across the disputes, but this piece shouldn’t be read as evangelical for any single school.
I would throw out reporting and measurement, if I could: I haven’t reported on keyword rankings to clients for nearly a decade. I would do the same here, if I thought it was possible. But for SEO this is a new stance. Many folks will need hand-holding. So we need to give them something to hold onto.
Measurement often makes people look in the wrong place for impact, corrupts the importance of hygiene work or incremental shifts and collapses the desire and capacity for new. So we need to do it carefully, clearly, and with buffers.
Alongside the measurement protocol, we need strategic, soft conversations to buy-in our stakeholders, with whatever secondary justification we can tie to it: customer service complaints, accessibility requirements, the 5-year plan — whatever we can slip the less-measureable-still important work up against and validate in at least one other way important to the corporate machine.
Because measurement should not dictate all decisions, and we need to hedge against it now, because measurement will always be in our vocabulary, and the work we will be doing is more difficult to measure.
Start Reading
SEO is a job of translation: of the real-world, analogue brand to the digital surface. The surface may change, but the foundation underneath doesn’t. In the same way, the companies that’ll survive the AI rewrite of search are the ones who are building a brand. Who have been working on mental availability through distinctive brand asset management, and have been throughout the digital era. Who are building emotional and psychological throughlines in their market. Who care.
The businesses that’ll get buried are the ones who have substituted ‘brand mentions’ for recognition.
We have to own we’re a part of that. We told our clients they didn’t need to be a brand as long as they ranked. That’s no longer true, and hasn’t been for a while. (RankBrain in 2015, if you’re counting; that’s my guess.)
We need to start being brands now.