How to Build a Content Strategy Around Target Keywords
The bridge between research and actual work. Most content strategies fail because they are never completed rather than because they were wrong, so the useful content here is about sequencing and finishing rather than about how much to publish.
Start From The Map, Not From Ideas
A content plan built from ideas produces a site nobody can navigate and nobody searched for. A content plan built from a keyword map produces pages people are already looking for. The difference is where the list came from. It decides everything downstream.
What a map is. Subjects assigned to pages.
A record of which searches belong together and which page will own each group. Our clustering guide covers producing one.
What an idea list is. Guesses.
Things somebody in the business thought would be interesting. Occasionally right, frequently about topics nobody searches and usually skewed towards what the business finds interesting rather than what customers ask.
Why the distinction is not pedantry. Wasted months.
Writing five well made pages about subjects with no demand costs the same as writing five about subjects with demand. The only difference is whether anybody arrives.
What the map gives a plan. Boundaries.
It tells you what is in scope, what each page is for and when a subject is complete. Without it a strategy is an open ended commitment to keep producing.
What to do if you have no map. Make one first.
An afternoon of grouping beats a year of publishing into the dark. The research method guide covers getting to that point.
Order By What You Can Win And What You Need
The first pages should be winnable enough to produce something and commercial enough to matter. Those two conditions together rarely point at the most searched subject on the list, which is why so many strategies begin in the wrong place.
Why winnable comes first. Evidence.
A business needs to see the approach working before it will keep investing. Early pages that appear and bring enquiries buy the patience for the harder ones.
Why commercial matters equally. Otherwise nobody cares.
A page that ranks quickly for something nobody buys from proves the mechanism and not the value. Both boxes have to be ticked for the first few.
What that usually means. Specific and unglamorous.
A particular service in a particular place. Or a specific problem you solve constantly. Not the broad term the business would most like to own.
The trap in ordering by volume. Nothing happens.
Starting with the biggest search produces months of work with no visible result, which is how businesses conclude the approach does not work when in fact the order did not.
What to keep for later. The ambitions.
Broader subjects are worth having on the list, positioned after the site has earned some standing. Deleting them loses the direction and starting with them loses the confidence.
Who should decide the order. The business.
Which services matter commercially is not a search question. That ordering should come from whoever knows which jobs are worth having.
Build The Anchor First
Within a subject, the main page comes before the pages that support it. Publishing three detailed pieces with nothing for them to point at leaves a set of orphans and a subject with no obvious centre.
What an anchor page does. Holds the subject.
It covers the subject broadly, answers the general version of the question and links out to the pages handling each part in depth.
Why it goes first. Everything attaches to it.
Each supporting page then has somewhere to point and somewhere to be pointed at from. Built in the other order, that structure has to be retrofitted.
Why the structure matters. Our advanced material.
The argument for why grouped, interlinked coverage outperforms scattered pages belongs to the advanced SEO material and this page will not restate it.
What an anchor is not. A summary.
It should be a genuinely useful page in its own right rather than a signpost. A thin index page helps nobody and ranks for nothing.
The practical version. Write it, then plan around it.
Producing the anchor usually reveals which supporting pages are actually needed, which is more reliable than deciding in advance.
What this looks like on our own site. This page.
You arrived from a hub covering the subject broadly. This page handles one part of it in depth. That is the arrangement described.
Finish One Subject Before Starting Another
A half built cluster underperforms in a way that tempts businesses to abandon the whole approach. This is the commonest reason content strategies get judged failures. It is a sequencing problem rather than a strategy problem.
Why half built is worse than small. No coverage.
Three of eight planned pages does not deliver a third of the benefit. The subject is not covered, the internal structure is incomplete and nothing has been demonstrated.
What businesses do instead. Spread out.
Start a subject, get impatient, start another. After a year there are four subjects each a third covered. None of them performs.
Why the temptation is strong. Novelty.
A new subject feels like progress. Finishing the fourth supporting page of an existing one feels like admin. It is the thing that produces the result.
What finishing looks like. The questions answered.
Every reasonable question about that subject has a home somewhere on the site. That is a checkable condition rather than a feeling.
The conclusion drawn wrongly. Content does not work.
We have inherited sites with a dozen abandoned half subjects, where the owner concluded the whole approach was ineffective. The approach was never tested.
The rule worth adopting. One at a time.
Nothing new begins until the current subject is complete. Restrictive, though it is the single change that most improves outcomes.
A Realistic Volume
A small business that completes a handful of properly chosen pages covering one subject has a content strategy. A business that commits to publishing something every week has a promise. It is the promise that gets abandoned by about March.
Why the weekly commitment fails. It has no end.
Open ended obligations lose to whatever is urgent. A finite plan with a visible finish survives a busy quarter because the end is in sight.
What a finite plan looks like. Countable.
A named set of pages covering one subject, with each one assigned. Somebody can point at it and say how much is left, which is what keeps it moving.
What we would ask a client to commit to. Completion.
Not a frequency. Finish this subject, then decide whether to start another. That conversation happens once rather than every month.
Why fewer pages can outperform more. Depth.
Eight substantial pages covering a subject properly beat thirty thin ones spread across six. The second arrangement also creates competition between your own pages.
What to do if capacity is genuinely small. Narrow the subject.
Cover something smaller completely rather than something large partially. A narrower subject fully answered is a real asset.
Where the craft sits. Our blogging material.
How to write the pages, plus how to keep them worth reading, belongs there rather than here.
When To Revise It
A keyword map should be revisited when something changes rather than when a date arrives. Three things genuinely warrant it. A quarterly review in the diary is not one of them.
The business changes. The clearest trigger.
A new service, a new area, a different type of customer or something dropped. The map describes what the business sells, so it moves when that does.
A subject is finished. The planned trigger.
Completion is the natural moment to decide what comes next. It is the only revision most businesses need in a year.
Real data has accumulated. The valuable one.
Once pages have been live for a few months, your own search reporting can correct assumptions the original research got wrong. Block seven covers that.
What is not a trigger. The calendar.
Redoing research because it is March produces churn. Nothing about search demand respects a quarterly cycle. A plan revised constantly is never executed.
What a revision usually involves. Less than expected.
The subjects rarely move much. The phrasing, the emphasis and the ordering do, which is an amendment rather than starting again.
When to genuinely start over. A different business.
A substantial change in what is sold or to whom. That warrants fresh research rather than editing the existing map.
Let Your Own Data Correct It
After a few months of pages being live, your own search reporting becomes better evidence than the original research was. It reports what actually happened on your site rather than estimating what might happen on any site.
What it reveals first. Unplanned queries.
Pages appear for searches nobody targeted. That is information about what the page is genuinely good at. It frequently suggests a better page than the one that was written.
The second finding. Weak appearances.
A search where a page appears consistently but low is demonstrated demand you can already partly serve. That is a far better target than a phrase from an estimate.
What it corrects. The assumptions.
Which subjects mattered, how people phrase things and which pages carry the weight. All of it was guesswork until the data arrived.
Why it beats redoing the research. Specific to you.
No third party knows what your site appears for. Our guide to using search reporting for research covers reading it.
What to do with the findings. Amend the map.
Add the subjects the data revealed, reorder what remains and remove anything the evidence says nobody wants. That is a short exercise rather than a new project.
The discipline. Wait for enough.
A fortnight of data corrects nothing. Give pages long enough to accumulate a picture before acting on what they show.
Know When It Is Done
A subject can be finished. Once every reasonable question a customer would ask has a home on the site, the work changes from production to maintenance. That is a legitimate state and hardly anybody in this industry will tell a business it has been reached.
Why nobody says it. Ongoing work.
A supplier paid monthly to produce content has little incentive to announce that a subject is covered. That is worth naming plainly.
How you know. Diminishing questions.
When the remaining ideas are things no customer has ever asked, you have reached the edge of the subject. Continuing produces pages for nobody.
What maintenance looks like. Smaller and cheaper.
Keeping prices current, updating what has changed, improving pages against what now outranks them. Real work, considerably less of it.
What comes after. A decision, not a default.
Start another subject. Or stop and let the finished one work. Both are valid and the second is rarely offered as an option.
The warning. Do not manufacture subjects.
Inventing topics to sustain a publishing schedule produces the thin content that causes problems elsewhere. Our blogging material covers updating rather than adding.
Where the rest sits. The hub.
All twenty-two guides are on the keyword research guide.
Four subjects,
each a third
covered, none
of them working.
A half built cluster does not deliver a fraction of the benefit. It delivers none, because the subject is not covered and nothing has been demonstrated. We have inherited sites with a dozen abandoned half subjects where the owner concluded content does not work. It had never been tested.
How we plan a build:
A supplier paid monthly to produce content has little reason to tell you a subject is covered. We will.
Every guide.
One subject.
What the words mean, what the numbers in a keyword tool actually are, how to find the searches worth targeting, how to turn a list into a publishing plan and what to do when a page you relied on slips.
