Content Strategy
Content strategy decides which pages should exist, what each one owns, and how they connect. It is pillar and cluster mapping against real queries rather than a keyword list, and its main output is a decision about what not to publish.
Deciding which pages should exist
Most sites do not have a content problem in the sense of insufficient volume. They have an architecture problem: several pages competing for the same query, no page owning the commercially important one, and internal links placed by whoever happened to be editing.
A topic map fixes the architecture before anything is written. Each pillar owns a commercial topic and links down to the services it covers. Each cluster answers one specific question completely and links back up. If two pages could rank for the same query, one of them is consolidated.
The most valuable output is usually subtractive. A strategy that identifies eleven pages to consolidate and six to write is worth more than one that recommends forty new articles, and it is considerably cheaper to execute.
What a content strategy covers
Six workstreams. The audit usually changes what the plan looks like.
Query research
The questions buyers actually ask, separated from industry vocabulary nobody outside the sector uses, and weighted by commercial value rather than by volume.
Cannibalisation audit
Existing pages compared against each other to find where several compete for one query, which is the most common cause of a site that ranks worse than its content deserves.
Pillar and cluster map
A structure where each pillar owns a commercial topic and each cluster owns one question, with the target query and intent recorded per page.
Internal linking specification
Which page links to which, and why, written down per page so the structure survives whoever edits it next.
Editorial calendar
Owner, reviewer and review date per piece, because unowned content decays and undated content cannot be judged.
Measurement plan
Success defined as qualified enquiries against named query sets, with a dated baseline, rather than as total sessions.
How a strategy engagement runs
Inventory
Every existing page catalogued with its current query, position and commercial relevance, so decisions are made against reality.
Audit competition
Internal cannibalisation identified and each conflict resolved by a decision: consolidate, differentiate or retire.
Map topics
Pillars and clusters defined with one target query and one stated intent per page, and gaps identified against the question inventory.
Specify links
Internal linking written page by page, with the reason for each link recorded so later editors do not undo it.
Plan and baseline
Editorial calendar with owners and review dates, and a dated performance baseline against the named query sets.
Strategy next to the work it directs
How content strategy relates to the disciplines that execute it.
| Dimension | Content strategy | Content engineering | SEO |
|---|---|---|---|
| Unit of work | The topic architecture | The template and markup | The page |
| Question answered | Which pages should exist? | Can the structure be produced at scale? | Should this page rank? |
| Main output | A topic map | Patterns, schema, QA gates | Position and enquiries |
| Time to effect | Immediate as a decision | 6 to 12 weeks | 2 to 6 months |
| Fails when | Volume is mistaken for coverage | Structure depends on discipline | The site cannot be crawled |
| Runs | Before writing | Alongside writing | Continuously |
What a strategy cannot do
A topic map does not write the content, and it does not supply the expertise the content needs. Where nobody in the organisation can answer a cluster question from experience, the page will either not get written or will restate what competitors already say. The remedy is consultant time, not another brief.
It also cannot create demand. Where the map reveals that the questions buyers ask have very low volume, the honest conclusion is that search is not the channel, and the budget belongs somewhere that creates awareness instead. We would rather report that in week six than build a calendar around it.
- The most valuable output is usually what not to publish.
- Cannibalisation is the most common reason good content underperforms.
- One target query and one stated intent per page, written down.
- Measure qualified enquiries against named query sets, not total sessions.
Questions buyers ask about this
How many cluster pages does a pillar need?
Enough to cover the questions buyers actually ask, and no more. Six clusters that each answer something completely are worth more than twenty that restate the pillar in different words, and they are cheaper to keep current.
What do you do about pages that already compete?
The audit names them explicitly. Usually the answer is consolidation into the stronger page with a redirect from the weaker, rather than publishing a third page on the same subject. Occasionally the right answer is genuine differentiation, and the map records which was chosen and why.
Do you write the content as well?
We can, and the strategy is designed so that a writer can work from it. The strategy engagement itself delivers the architecture, the specification and the calendar. Where the content requires expertise your team holds, the more efficient arrangement is usually consultant interviews written up rather than external authorship from scratch.
How is this different from keyword research?
Keyword research produces a list. A strategy produces decisions: which pages exist, what each owns, what gets consolidated, how they link, and how success is judged. The list is an input to that, and on its own it tends to generate volume without structure.
How do you measure whether the strategy worked?
Against a dated baseline, by non-branded organic performance on the named query sets and by qualified enquiries from the pages in the map. Total traffic is excluded deliberately, because it rises comfortably on queries that will never buy anything.
Marked up as FAQPage structured data, matching the visible text exactly.
Still not sure this is the right service?
Answer four questions and we will tell you which one fits — or that none of them do.
Tell us what you are trying to fix.
A short conversation about the objective, the constraints and the timing. If we are not the right fit, we will say so.