How Does SEO Strategy Change as Websites Grow?
A larger site is not a small site with more pages. At each stage of growth something different is the thing holding you back. Work aimed at the wrong one is expensive whichever direction the mistake runs in.
The Work Changes, Not Just The Amount
Growth does not multiply the work. It replaces it. What limits a site at one size is not what limits it at the next, so a plan that scaled correctly would still be the wrong plan.
The useful idea underneath this. One thing binds at a time.
At any moment a site has a single constraint doing most of the limiting. Effort spent anywhere else produces very little, however well executed, because the binding constraint is still binding.
Why that is easy to miss. Everything looks improvable.
A site always has things that could be better. Being able to identify improvements is not the same as identifying the one that would change the outcome. Most audits produce the first while claiming the second.
What changes as a site grows. Which constraint is binding.
Four stages, each with a different one. Existence, then organisation, then allocation, then permission. Those need genuinely different work rather than more of the previous kind.
What this page is for. Working out where you are.
The four stages follow, each pointing at the material that serves it. Read the one describing your situation and treat the rest as context.
The Small Site Problem Is Existence
A small site's problem is not having enough of the right content at all. Not organisation, not technical quality, not measurement. There is simply not enough there to answer what people are asking.
What that looks like in practice. Gaps rather than faults.
A handful of service pages, a contact page and possibly a blog nobody has updated. Everything present may be perfectly good. The problem is what is absent.
Why technical work disappoints here. It is not the constraint.
A small site can be made faster, cleaner and better structured. None of that produces enquiries if there is nothing addressing what people search. Technical work at this stage improves something that was not limiting you.
What actually moves the needle. Publishing the missing answers.
Working out what customers ask before buying and answering it properly. Unglamorous, slow and the only thing that changes the position at this stage.
Why this stage is underserved. It is hard to sell.
An audit produces a list of findings and looks like value. Telling somebody they need to write twelve pages over six months looks like less. It is worth considerably more.
The Middle Stage Problem Is Structure
Once there is enough content, the problem becomes organisation. The pages exist, they are individually reasonable and collectively they are a pile rather than a structure.
The symptom that announces this stage. Pages competing.
Several of your own pages addressing the same question, none clearly the answer. That is a structural problem rather than a content one. Writing more makes it worse.
What the work becomes. Grouping and connecting.
Deciding which pages belong together, which subject each section covers and how they refer to each other. That is information architecture and it is the defining work of this stage.
Why businesses keep publishing instead. It is what worked before.
Publishing produced results at the previous stage, so the response to slowing results is to publish harder. That is applying the small site remedy to a middle stage problem and it accelerates the accumulation.
The moment this stage begins. Not a page count.
When somebody asks where a new page should go and there is no obvious answer. That question arriving is the signal. It usually arrives long before anybody acts on it.
The Large Site Problem Is Allocation
At scale the problem becomes which pages get attention and which do not. There is not enough to go everywhere, so the distribution decides what happens. Nobody designed that distribution.
What changes about the question. From what to publish, to what gets seen.
Earlier stages assume that a published page will be assessed. At this scale that stops being safe. Pages can exist for years without anything ever requesting them.
Where the answers are. Two pages in this section.
Establishing where attention actually goes is log file analysis. Changing it is crawl optimisation at scale. Both only become relevant at this stage.
The counterintuitive answer that appears here. Fewer pages.
For the first time, removing content improves the site. That is a reversal of everything the previous two stages taught. It is covered in what is content pruning in SEO.
What makes a site reach this stage. Not always size.
Generated addresses arrive here far earlier than a large body of written pages. A modest catalogue with filters can be dealing with allocation while a substantial editorial site is not.
And The Very Large Site Problem Is Governance
At the largest scale the constraints become organisational. Who is allowed to change what, how long agreement takes and how many teams have to be persuaded. None of that is a search problem and all of it decides search outcomes.
What the work becomes. Persuasion and sequencing.
Establishing what should change is frequently the quick part. Getting four teams with different objectives to agree, then finding a release in which it can happen, is the project.
The characteristic failure. Correct recommendations, unimplemented.
Organisations at this stage accumulate excellent unactioned advice. The audits were right and nothing about them addressed the actual constraint, which was never analytical.
What that means for who you need. A different profile.
Somebody able to build agreement across an organisation achieves more here than a stronger specialist who cannot. That is set out in the enterprise material in what is holistic SEO.
The other governance problem. Things coming back.
At this scale enough people can change the site that fixed things get unfixed by somebody with no reason to know. Maintaining a state becomes as much work as reaching it.
Doing The Wrong Stage's Work
Small sites are routinely sold technical work they do not need. Large sites are routinely sold content when their problem is structural. Both are the same error and it runs in both directions.
Why small sites get sold technical work. It is packageable.
A technical audit is definable, deliverable and produces a document. Twelve pages written over six months is harder to sell and it is what would actually help. The incentive favours the wrong answer.
Why large sites get sold content. It is what they bought before.
An organisation that grew through publishing keeps commissioning publishing. Adding more to a site whose problem is that too much of it goes unseen makes the situation worse while feeling like progress.
Why nobody catches it. Both parties are satisfied.
The work is delivered, the invoice is paid and something was produced. The absence of a result gets attributed to search being slow rather than to the work having addressed the wrong constraint.
What to ask a supplier. One question.
Which of these four things is currently limiting us, then why that one. Somebody who has diagnosed the stage answers immediately. Somebody selling a package changes the subject to what they do.
Why diagnosis comes first. It decides everything else.
Every remaining decision depends on the answer. It is the step most commonly skipped, because it is not itself billable.
What Stays The Same Throughout
The fundamentals do not change with scale. A large site failing at them fails larger. Reading this page as a sequence in which later stages replace earlier ones would be an expensive misunderstanding.
Content still has to be worth reading. At every stage.
Allocation work on a site whose pages nobody wants moves attention onto pages that will not convert. The later constraints assume the earlier work was done rather than replacing it.
The site still has to be reachable. Always a precondition.
Nothing at any stage compensates for pages that cannot be found. That is a floor rather than a stage. It applies to a five page site and a fifty thousand page one identically.
The subject still has to be one you can own. Scale does not confer authority.
A very large site covering everything shallowly is a large version of a problem rather than a solution to it.
What actually happens with growth. Additions, not replacements.
Each stage adds a constraint rather than retiring the previous one. A very large site has all four, which is why the work becomes harder rather than merely different.
The practical implication. Do not abandon the basics.
Businesses reaching a new stage sometimes stop doing what got them there. The new constraint is additional rather than a substitute.
Growth Creates Its Own Problems
Every stage transition introduces something that was not a problem before. That is normal rather than evidence anything was done badly. Anticipating it is most of what a strategy is for.
Why success produces the next problem. Directly.
Publishing enough to escape the first stage is what creates the structural problem of the second. Growing enough to matter is what creates the allocation problem of the third. The remedy at each stage is the cause of the next.
Why that feels like failure. Results plateau after work.
A business does the right thing, sees it work, then sees the returns flatten. The natural reading is that the approach stopped working. The accurate reading is that the constraint moved.
What a strategy is actually for. Seeing the next one coming.
Not a list of tasks. Knowing which constraint is binding now, then which one will bind once this is resolved, so the transition is planned rather than discovered.
The signal that you have moved. Diminishing returns on what worked.
When the thing that reliably produced results stops producing them, the first question is whether the constraint has moved rather than whether the work got worse.
What to do with that. Re-diagnose rather than intensify.
The instinct is to do more of what worked. The useful response is to establish what is limiting you now. The full series is on the advanced SEO guide.
Which of the four
is limiting
you now?
Existence, structure, allocation or permission. Every other decision follows from the answer. It is the step most commonly skipped, because it is not itself billable. We diagnose the stage first and will tell you plainly if what you need is not what we sell.
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.
