Why Does Topical Authority Outperform Keyword Targeting at Scale?
This is the argument the rest of our work rests on, so it is worth making properly rather than asserting. It also has conditions attached. There are businesses for which this is the wrong approach entirely. Those are set out here rather than left for you to discover after six months.
What Topical Authority Actually Means
Being recognised as a genuine source on a subject rather than having a page that mentions a phrase. The important part of that sentence is that it describes a body of work rather than a page. Almost every mistake in this area comes from ignoring the distinction.
Why the unit matters. You cannot audit it page by page.
A single page can be assessed on its own terms. Authority cannot, because it is a judgement about what a site collectively demonstrates. Reviewing individual pages will never reveal whether a site has it.
What it is not. A score, a metric or a badge.
Nobody publishes a figure for it and no tool measures it, whatever the label on the dial says. It is a description of a relationship between a site and a subject rather than a number attached to a domain.
The everyday version of the idea. Familiar outside this industry.
You already treat some sources as authoritative on some subjects and not on others. That judgement is formed from having encountered them repeatedly on that subject, rather than from one impressive article. It works the same way here.
Why this page exists separately. Argument before method.
The sequence for building one is in how to build topical authority from scratch. This page is the case for doing it, because a method nobody understands the reason for gets abandoned at the first difficult month.
Why One Page Cannot Do It
A single excellent page on a subject sits alongside sites that have covered that subject exhaustively. The comparison being made is not between your page and their page. It is between what you have demonstrated and what they have.
What that changes about quality. Quality stops being sufficient.
This is the uncomfortable part. A genuinely better page can lose to a worse page belonging to a site that has addressed the whole subject. That is not a flaw in the assessment. It is the assessment working on the unit it actually operates on.
Why businesses find this counterintuitive. Effort feels page-shaped.
Somebody spends a fortnight on one definitive page, publishes it and expects the result to match the effort. The effort went into a unit that is not the one being weighed.
The observable symptom. Good pages that underperform.
Most established sites contain at least one page everybody agrees is excellent and which has never performed. That page is usually isolated. The isolation rather than the quality is the explanation.
What follows for planning. Fewer subjects, done completely.
If the unit is a body of work, resources have to be concentrated rather than distributed. That is the practical consequence and it is the decision most businesses find hardest.
The Coverage Argument
Covering the questions around a subject demonstrates knowledge in a way no single page can. That is the claim. Here is why it holds rather than simply sounding sensible.
Knowledge shows in the awkward questions. Not the obvious ones.
Anybody can write the introductory page on a subject. What separates somebody who has done the work is that they can answer the edge cases, the exceptions, the things that go wrong and the questions customers only ask after something has failed.
Which is why coverage is evidence rather than volume. The hard questions are the proof.
A site addressing only the popular questions has demonstrated that it read the same research everybody else did. A site addressing the awkward ones has demonstrated something that cannot be faked from a list.
How this connects to the mechanism. Same argument, other direction.
Relevance is assessed on what a page and a site are about rather than on which phrases appear, which is set out in what is semantic SEO. Coverage is what makes aboutness legible at the level of a site.
The test for whether coverage is real. Could a customer get stuck.
Take a genuine customer question, follow it through your site and see whether they reach an answer or a gap. Gaps are where coverage claims fail. They are visible to anybody who looks.
It Compounds
Each page makes the others more credible. That is not a slogan, it is the reason the return accelerates rather than accumulating evenly. It is the single most important property of this approach.
Why one page helps another. Two mechanisms.
Each addition strengthens what the site collectively demonstrates about the subject. And each page gives the others somewhere relevant to link to and from, which is how a set of pages becomes a structure rather than a pile.
What acceleration means practically. The later pages work harder.
The tenth page on a subject arrives on a site already demonstrating knowledge of it. The first page arrived on a site demonstrating nothing. Identical effort, different starting position, different result.
Why this is unusual among marketing activities. Most do not behave this way.
Advertising restarts every time the money stops. Most channels produce roughly linear returns for linear effort. Something where the earlier work makes the later work more effective is rare and is worth structuring a budget around.
What we will not do with this. Draw the curve.
No figures, no percentages and no period, because nothing supports them. The shape is real and any specific number attached to it would be invented, which is the subject of the next block.
Which Is Why It Looks Like It Is Not Working
A cluster half built performs worse than its individual pages suggest. That is the direct consequence of everything in the previous block. It is also the point at which businesses abandon the approach.
Why the middle is the worst part. The mechanism is not yet running.
Halfway through, you have paid for half the work and none of the compounding. The pages exist without the body of work that makes them credible, so they perform like isolated pages, which is exactly what block two says they will.
Why that is the moment of maximum doubt. Spend is real, result is not.
Several months of invoices, a visible pile of published pages and reporting that does not justify either. Anybody would ask the question. It sounds like a reasonable challenge rather than a predictable stage.
What abandoning it produces. The worst available outcome.
Stopping halfway leaves a business with the full cost and none of the compounding, plus a set of pages that will now never earn what they were built for. That is worse than never starting.
Why we put this before the work rather than during it. The difference between an expectation and an excuse.
Said in advance, this is a shape to plan around. Said in month five, it is indistinguishable from a supplier explaining away a failure, however true it is. Our measuring performance material covers reading the interim period properly.
Where It Does Not Work
This approach is not universal and presenting it as universal would misrepresent our own method. Three situations where it is the wrong choice.
Where there is no genuine expertise. The absolute limit.
A business with no real knowledge of a subject cannot manufacture authority in it. The awkward questions in block three are precisely what cannot be produced without having done the work. A cluster written from research alone reads exactly as what it is.
Where the subject is too broad to finish. A scoping failure.
Choosing a subject nobody could cover completely guarantees a permanently half built cluster, with all the consequences in block five and no end point. That is the commonest way this fails and it is decided at the start.
Where the business needs revenue this quarter. A timing mismatch.
The compounding is real and it is not fast. A business that needs enquiries within weeks needs something that produces enquiries within weeks. This is not that. Saying otherwise is how agencies lose clients they should never have taken.
What thin coverage of many subjects produces. Worse than nothing.
Not slower results. A site demonstrating shallow knowledge of eight things has demonstrated something about itself. That impression sits alongside everything else it publishes.
Authority In What
A business has to be able to state the subject it wants to be known for in a sentence. A firm unable to do so is not ready for this approach. No amount of production makes up for it.
Why a sentence is the test. It forces a decision.
Anything requiring a paragraph is several subjects. Anything phrased as a list of services is a catalogue rather than a subject. The constraint of one sentence does the scoping work that a workshop usually fails to.
What a good answer sounds like. Narrow and commercial.
Specific enough that complete coverage is achievable, valuable enough that the people searching it are people you want. Both conditions, not one.
The two ways it goes wrong. Opposite directions.
Too broad and it can never be finished. Too narrow and it is finished quickly and nobody is searching for it. Most businesses err broad, because narrowing feels like turning away work.
Why narrowing does not turn away work. The point most resisted.
Being known for one thing brings enquiries about adjacent things, because people who trust a source on one subject ask it about neighbouring ones. Breadth is an outcome of depth rather than an alternative to it.
What to do if you cannot answer it. Not a content problem.
If nobody in the business can state the subject in a sentence, that is a positioning question rather than a search one. It should be resolved before anything is published.
The Line Against Volume
Coverage means answering the questions that exist rather than generating pages until a number is reached. Those two things look identical on a content plan and they are entirely different exercises.
The difference in one test. Was each page worth writing.
A topical cluster and scaled content abuse can produce the same number of pages on the same subject. The distinction is whether somebody decided each page answered a question that genuinely exists. That decision is the whole of it.
Where the discipline breaks. When the number becomes the target.
A plan specifying a quantity of pages per month has replaced the question with a quota. Once that happens, pages get invented to fill it. The ones invented are by definition the ones nobody was asking for.
Why we are strict about this. It is our own risk.
We produce pages at volume, which puts us one careless decision away from the thing published policy describes as scaled content abuse. The production side of that argument is in what is programmatic SEO and the policy side is in our risky practices material.
How the three arguments agree. Deliberately.
Strategy, production and policy reach the same test from three directions. A reader arriving from any of them should end up at the same question, which is whether the page was worth making.
Keywords Still Matter
None of this is an argument against keyword research. Businesses read a page like this and conclude that research is obsolete, which produces content about subjects nobody is searching.
What research is for now. Finding the questions.
Its purpose changed from producing a list of phrases to place towards producing a map of what people genuinely want to know. That is the input to coverage rather than an alternative to it. Our keyword research material covers the mechanics.
Why language still decides matching. People type words.
A page that never uses any phrasing a real person would use is harder to match to a query, whatever it is about. Understanding meaning does not remove the need to meet somebody in their own words.
The specific risk of ignoring research. Covering the wrong subject completely.
A perfectly executed cluster on something nobody asks about is the most expensive possible failure, because everything was done well except the decision that mattered.
What the two approaches actually are. Sequence rather than alternatives.
Research establishes what to cover. Coverage is how you address it. Anybody presenting them as competing philosophies has misunderstood both. The full series is on the advanced SEO guide.
Say your subject
in one
sentence.
If you can, this approach is available to you and we can scope it. If you cannot, that is a positioning question rather than a search one. We will say so rather than starting anyway. We will also tell you plainly if you need enquiries faster than this produces them.
What the national tier covers:
The national tier is £1,550 a month. It suits businesses competing beyond one town. We will say so if a lower tier would serve you better.
Every guide.
One practice.
Semantic search, machine learning, topical authority, programmatic production, rendering, log files, crawl allocation, information architecture, content pruning and how strategy changes with scale.
