SEO content audit
A verdict on every page — keep, merge, refresh, retire or differentiate
Years of publishing leave most established sites with an archive that competes with itself — four articles answering one question, top earners going stale, a long tail earning nothing. This is a fixed-fee content audit that gives every page a verdict, with the evidence beside the call: keep, merge, refresh, retire, or differentiate. Every change arrives as a brief, merges are specified so nothing valuable is lost, and the whole plan is sequenced for the team you actually have.
- A verdict on every URL — keep, merge, refresh, retire or differentiate
- The evidence beside each call; you review the verdicts before anything ships
- Merges specified section by section, so nothing valuable is lost
- A rollout sequenced and sized for the team you actually have
- Fixed fee, from £4,500, agreed before work starts
No commitment. I reply personally within two working days.
From archive to earning again, step by step
A clear path from "what's dragging the archive down" to "here's the rollout" — four deliverables your team can act on, then three optional extras if you'd rather not go it alone.
Step 1 — the diagnosis
First, what the archive is doing to itself
An archive that has stopped earning usually competes with itself. The diagnosis shows where, with the evidence to prove it:
- Every URL's evidence gathered: traffic, links, demand, decay
- Overlap mapped by meaning, so competing clusters surface
- The long tail earning almost nothing, quantified against the archive
See the real section from an audit
The archive is competing with itself
Eight years of publishing left 286 advice posts, and about 140 of them do nearly all the work. The biggest single problem is overlap: 61 posts compete with each other across 18 topics — three or four articles answering one question, which splits the ranking so none of them wins outright. A quarter of the archive earns less than one percent of its visits, and the top-earning regulatory guides state requirements that changed in January.
Step 2 — the verdicts
A verdict on every page, with the evidence
Keep, merge, refresh, retire — or differentiate, when a page's best move is saying something rivals can't.
- Keep, merge, refresh, retire or differentiate — every URL ruled on
- The evidence beside each call — no black-box scoring
- You review and veto before anything ships
See the real section from an audit
A verdict on every page — and what a merge actually decides
Every one of the 286 posts gets one of five verdicts, each with the evidence and the reasoning on its row. You review them before anything ships — you know things about these posts the data can’t see.
| Verdict | Posts | Share of visits | First move |
|---|---|---|---|
| Keep | 92 | 69% | Re-link the 12 strong guides no selling page points to |
| Refresh | 48 | 23% | The 9 guides quoting out-of-date regulations — facts first |
| Merge | 61 → 18 | 4.2% | Cluster 07 first — it proves the approach for the other 17 |
| Differentiate | 14 | 3% | The two fire-door flagships in cluster 12 |
| Retire | 71 | 0.8% | Wave 1: the 40 posts nothing arrives at from any channel |
And one verdict worked in full — a merge, the call most audits get wrong:
“Choosing a composite door” — 3 posts become 1
- The keeper
/advice/how-to-choose-composite-doors/— strongest rankings, most links from other sites, best structure. Stays at its address.- What the second post brings
/advice/composite-door-buying-guide/— its fitting-cost table (unique, and other sites link to it) and the security-ratings section carry over. Everything else repeats the keeper, so its address then forwards to it.- What the third post brings
/blog/2022/choosing-your-perfect-door/— the “common mistakes” FAQ carries over word for word: three of its questions show in Google’s “People also ask”, so the wording stays untouched. This address forwards too.- The merged draft
- The keeper’s structure plus the fitting costs, the security ratings and the mistakes FAQ — one introduction instead of three, updated for the 2026 fire-rating changes. About a day of your editor’s time to review.
- How I’ll know it worked
- The three posts earn 3,535 search visits a month between them. The merged post has to beat that within 8 weeks, tracked weekly — and if it hasn’t by week 6, the remaining merges pause while I find out why.
Step 3 — the hand-back
Every verdict comes back as a brief or a draft
Nothing lands as "update this". Every verdict is handed back as work your team can start the same day:
- Merges handed back as a drafted article and a review pack — never straight to live
- Refreshes and differentiations as briefs your editor can act on without a meeting
- Sequenced by value: quick wins live first, then the merges in payback order
See the real section from an audit
The work comes back as briefs and drafts, not a to-do list
Nothing lands as “update this”. A refresh comes back as a brief specific enough that your editor can start without a meeting; a merge comes back as a finished draft with everything you need to check the cut. Here’s a refresh brief, then what arrives with every merge.
“How much does a composite door cost?”
- Diagnosis
- Ranks 11.2 for “composite door prices” — page two, on 8,400 impressions a month for 31 clicks. The demand is there; the position isn’t. “Composite door cost fitted” gets 0 clicks off 1,140 impressions, because the page never answers it directly.
- Why it’s worth doing
- One position off page one on the highest-intent buying query the range should own. This is a refresh, not a rewrite — the page is strong, it’s just buried and out of date.
- What’s winning now
- The top ten are 2026-dated price guides that lead with fitted costs and a price-by-size table in the first screen. This page opens with a definition.
- What to change
- Reword the title to lead with “Composite door prices in 2026”; add a fitted-prices section (the door and the installation, costed separately); surface a “prices by size and glazing” table; and link to the composite range and the how-to-choose keeper. No new research — the numbers are already on the page.
And every merge arrives as a pack, so you can check the cut before anything goes live:
- The merged draft
- A web preview to read and a document to comment on — one introduction, no repeated sections, every link inside the article kept live.
- What was cut
- Every passage that didn’t make the final draft, in full, with the reason and what replaced it. In cluster 07: 11 repeated sections dropped, all three customer reviews kept word for word.
- Where every sentence came from
- The draft colour-coded by source — this one is 57% the keeper, 38% the two folding-in posts, 5% joining edits. Read it to check the working, or skip it if you trust the cut list.
- Images and buttons
- All kept, nothing dropped on your behalf — and tagged: 9 of the 41 arrived with the folding-in posts and want your eye, the rest already live on the keeper.
- Review notes
- A short list splitting the clear fixes I can make on your word (a couple of stale product names carried over from the live page) from the judgment calls that stay yours (does one long section earn its place near the top?).
The standing rule: no address changes hands until the link check is done. Every link another site points at a merging or retiring post is found first and given a destination on purpose — handled carelessly it’s lost, handled properly it flows to the post that stays.
Step 4 — the wrap-up review
A then-and-now read, once the work is done
When the rollout's complete (whenever that lands), I take you through what shipped and how the archive responded, measured against where it started:
- A recap of what shipped — merges, refreshes, retires, quick wins
- The success measures quantified: then vs now, against the baseline
- A straight read on what moved, and anything that still needs a look
See the real section from an audit
What shipped, and how the archive responded
Once the rollout’s complete (whenever that lands), I take you through where it landed. It’s a clear before-and-after on the measures the audit set as its markers of success, not an ongoing dashboard.
What shipped:
| The work | Done |
|---|---|
| Accuracy fixes & quick wins | 9 regulatory guides corrected, 40 dead posts retired, 23 internal links repointed |
| Merges | 18 clusters folded into their keepers, old URLs redirected |
| Refreshes & differentiations | 48 refresh briefs actioned, 14 competing pairs split |
And how performance moved — the success measures, then vs now:
| Success measure | At the audit | Now | Call |
|---|---|---|---|
| Overlapping topics | 18 topics, 61 posts splitting the rankings | 18 topics, one owner each | Landed |
| Regulatory accuracy | 9 top guides quoting out-of-date rules | All verified current | Landed |
| Archive size | 286 posts | 172 — a 40% cut | Landed |
| Search visits | 41k/mo baseline | 44k/mo — up 7% on a smaller archive | On track |
Optional extras
Optional extra
Track the clean-up back to the business
The clean-up ships in waves — retire, merge, refresh — and each wave is meant to move a number. A private dashboard shows whether each one did. From £2,500, then £250 a month if you keep it on.
- Every rollout wave annotated on the chart the week it ships
- Keep, merge and refresh cohorts each tracked against their own benchmark
- Carried through to conversions and revenue, not just sessions
Optional extra
Guardrails, so the overlap doesn't grow back
The clean-up changes the pages, but not the publishing that built the overlap, and twice-cleaned archives grow it back. Guardrails are the standing checks that catch the drift early. Scoped to your site — what needs watching sets the price.
- A publish gate: new titles checked against the topic map before drafting
- Decay alerts when a kept or refreshed page starts losing ground
- A standing link check so broken journeys never quietly build back up
Optional extra
Bring me in to oversee the rollout
The audit hands your team everything to run it themselves. When you'd rather it was steered, I stay on to do the steering. On hand throughout, scoped per project:
- The rollout managed and kept on track, wave by wave
- Dependencies and priorities sequenced as the work ships
- Each change reviewed and signed off before it goes live — and me on hand throughout
Every audit is a fixed fee, from £4,500, agreed before work starts — it covers most archives up to ~1,000 pages of content; larger archives are scoped individually.
AgencyAnalytics
B2B SaaS
Argos
Retail
Barclaycard
Financial Services
Bravissimo
Fashion
CMO Group
Ecommerce
Dunelm
Home Retail
Hindawi
Publishing
IEDigital
Fintech
Jobber
B2B SaaS
MOO
Ecommerce
Technogym
Fitness & Sport
The Range
Retail
Virgin Experience Days
Leisure & Gifting
Sitting on an archive that stopped earning?
Tell me how big it is and what it's supposed to do for the business. I'll tell you whether a content audit is the right next step — and if it isn't, what is.
Before you enquire
Common content audit questions
What does a content audit cost?
A content audit is a fixed fee, from £4,500, agreed before work starts — and the fee includes the verdict review and my support while the rollout ships, not just the document. The exact price depends on the size of the archive: the fee covers most archives up to around 1,000 pages of content, and larger archives are scoped individually. If an audit isn't the right next step, I'll say so before you spend anything.
What does an SEO content audit include?
An SEO content audit is delivered as a written document: an executive summary with the traffic baseline, a verdict on every URL with the reasoning, the merge and refresh briefs specified section by section, a worked merge draft with its hand-back pack, a sequenced rollout and a then-and-now wrap-up review once the work lands — plus a data sheet with all the per-URL evidence. Ongoing monitoring and regression guardrails are an optional add-on.
Will we lose traffic by removing content?
You won't lose traffic if the removals happen in the right order, which is the point of the audit. Nothing retires until its value has somewhere to go: useful sections move to the surviving page, links get redirected, and every wave is measured against the baseline before the next one ships. The pages being removed typically carry almost none of the traffic — in the sample, a quarter of the archive earns 0.8% of it.
How is this different from a content quality audit?
This SEO content audit is an inventory decision: it works through everything you have published and rules on each piece — keep, merge, refresh, retire or differentiate — based on how it performs and what it duplicates. The content quality audit answers a different question: why pages fail the quality bar, and which templates keep producing pages that do. A big archive that has stopped earning usually needs the content audit; a template-built site usually needs the quality audit. If it's unclear which fits, say so in the enquiry and I'll point you to the right one.
What about the content I'm missing — the gaps my competitors cover?
Finding the gaps your competitors cover is gap analysis — a separate piece of work I can add. I'd usually run it after this audit, not alongside it. When content is underperforming, the fastest gains are almost always in fixing what's already there — merging what competes, refreshing what's gone stale, retiring what drags the rest down. Fill the gaps with fresh content before that, and you build on shaky ground: the new pages inherit the same problems, and a moving baseline hides what's working. So the order I'd recommend is clean the archive, prove it has recovered, then find the gaps worth filling. Say so in the enquiry and I'll scope both, in that sequence.
Do you write the merged and refreshed articles?
No — writing the content isn't a service I offer. The two kinds of change come back differently: a merge arrives as a draft to review and publish (several posts consolidated into one), and a refresh or new section arrives as a brief that says exactly what to rewrite or add, for your team to write. If you'd rather the rollout was steered, I can oversee it — managing the sequence and signing each change off before it's live.
How much of our team's time does the audit take?
Very little during the audit: a short scope call, a few minutes setting up access, then the verdict review — the one meeting that matters, where we walk the calls together before anything ships. The rollout afterwards is sized to your team from the start: in the sample, the whole programme is cut for one content editor, with an owner and a week window on every row.
How do you decide what gets retired?
Evidence, then your veto. A page reaches the retire list when it has no traffic, no links, no demand for its topic and nothing worth salvaging — and every verdict is reviewed with your team before anything ships, because a page can matter for reasons the data can't see: a print campaign, a warranty document, a range that's coming back.
How long does a content audit take?
A content audit usually completes in two to three weeks, depending on the size of the archive and how quickly access is set up. The verdict review is booked as soon as the document is ready.
Three patterns hold the archive down. Clusters split their own rankings; the top earners have gone stale; and a long tail earns almost nothing while it drags down how the good posts are judged. Fix those three and the rest is housekeeping.