Content that needed more than a rewrite

A US-based digital agency with roughly 30 people brought Mediasphere in on a project basis to fix an outdoor gear client — around 800 products, tents, stoves, sleeping bags, packs. The agency ran the relationship; we worked under their name for 18 weeks. The end client never dealt with us and never knew we existed.

The website had 176 pages describing products and categories, written by different people over roughly six years. Nobody had gone back and looked at the whole thing.

• The pages were fighting each other. 28 buying guides covered only 11 real topics — 4 separate pages tried to answer “how to choose a sleeping bag.” Google could not tell which one mattered, so it split the credit and ranked none of them well. The same pattern held on categories: 52 pages covering 20 actual categories.
• The product pages described instead of answering. Materials, weights, dimensions — but not who it was for. Average time on a product page was 47 seconds, and 68% of users left without clicking anything.

Before writing anything, we read all 176 pages and compared them against what people were actually typing into Google — the questions, the phrasings, the intent behind each search.

Then we wrote the plan, not the pages. We listed what each type of page should say and in what order, decided which pages to merge or delete, and mapped which questions each product page had to answer before a buying decision was possible. The agency approved it without a single revision.

Weeks 1–2 · Briefing. Twelve hours across three sessions to examine the client's brand, their products, and establish the workflow. Then a deep dive into their website analytics and search data — delivered as weekly written updates rather than more virtual meetings.

Weeks 3–5 · The plan. Documented the site inventory, listed what each page type should say, and mapped what to merge and what to delete. The agency approved it without a single revision.

Weeks 6–16 · Writing. 96 pages fully rewritten and consolidated into 80 category pages and 32 guides. Out of that condensed schema we also produced 60 social posts cohesive across both channels. Cadence: roughly 12 pages a week at ~1.8 rounds of changes per piece.

Weeks 17–18 · Handover. Final pages, the plan document, the redirect map (old page → new page), and search recommendations. The agency delivered all of it to their client under their own name.

Measured nine months after the new pages went live, against the same nine months a year earlier.

• Monthly organic visitors from Google: +28%
• Pages ranking on Google's first page: 24 → 43
• Average time on product pages: +47%
• Users leaving without clicking: −9 percentage points

Traffic didn't rise because there were fewer pages — it rose because they stopped competing with each other. Rankings improved faster than visitors (79% vs 28%), which is expected: new rankings arrive at the bottom of page one where far fewer people click, and traffic follows a few months later.

Cost to the agency. Roughly three hours a week of their own team's time across eighteen weeks — including the twelve hours of briefing at the start — to run a project that rebuilt 128 pages.

Why this worked. Most content help is production: someone decides what the pages should say, briefs it, and a writer produces it. That works when the decisions are already made. Here they weren't. The pages were competing with each other and nobody had figured out which page should own which topic — so any brief would have repeated the original mistake. We started a level above the writing: first the plan (what each page type should say, what to merge, what to delete). Then the writing.

Mediasphere can partner with other agencies to support their existing complex ecosystems on a white-label basis.