Content creation
Every brief starts with the search, not with a word count. We map what people are really asking around a topic, structure the page to answer it in the order they ask, and write it so it reads like a person wrote it — because that is what keeps ranking.
published pages
needs refreshing
writes a word
earning today
Five kinds of page, five different jobs.
A buying guide and a comparison page are not the same thing and should never be written the same way. Switch type to see what we make and what each one returns.
- Smart heating, explained1,840 words what is smart heating 4,180 1,840 6 1
- What a full install actually costs1,620 words smart heating cost 2,960 1,620 41 2
- Grants you can still claim in 20262,140 words heating grants uk 3,410 2,140 28 3
- Heat pump vs combi, honestly1,980 words heat pump or combi boiler 1,870 1,980 33 2
- Why your radiator is cold at the bottom1,120 words radiator cold at bottom 5,240 1,120 9 1
Last year's pages are still working.
This is the argument for content, and it's the one that takes patience to believe. Each band is a quarter's worth of publishing. The bands don't disappear — the pages written eighteen months ago are still bringing people in this month, for nothing.
That's the opposite of paid, where traffic stops the day the budget does. It's also why the first three months of content feel like nothing is happening, and why the businesses that quit at month four never see the part that makes it worthwhile.
- You own every page. There's no rent, and no algorithm to be evicted by.
- The cost per visit falls every month a piece stays live.
Your best page is dying quietly.
Nobody warns you about this one. A page peaks somewhere around month eight, and then starts losing perhaps 3% of its traffic a month — slowly enough that nobody notices, because everyone is looking at the total rather than at individual pages.
It isn't that the writing got worse. Competitors published something newer, the facts aged, and the search itself shifted. Left alone for two years a strong page quietly halves.
A refresh — updated figures, the questions people now ask, a restructure to match what currently ranks — costs a fraction of a new piece and usually recovers more traffic than a new piece would have won. Which is why we budget for maintaining the library, not just for adding to it.
Most content fails at the brief.
By the time a draft is bad it's too late — the decisions that made it bad were taken before anyone opened a document.
What a brief contains
Every one of these is decided from what already ranks, not from an opinion about what would be nice to write.
-
Query
"how much does smart heating cost"
Informational, comparing before buying -
Length
1,600–1,900 words
Median of the current top five, not a target we invented - Answer A price range in the first 80 words. Bury it and they go back to the results.
- Cover Install costRunning cost GrantsPaybackWhich system
- Links Three internal, to the pages that sell
Right topic, wrong intent
Both of these rank. Only one is written for somebody who is about to spend money.
Always something moving
A month's work sits at four different stages at once, so publishing never depends on one person having a good week.
The order we do it in.
Publishing more is the easiest thing to sell and the least likely to work. Everything here happens before volume.
Most sites we take on are already sitting on pages that could rank and don't. We inventory everything, measure what each piece actually earns, and separate the ones worth refreshing from the ones worth merging or deleting. Thin pages competing with each other for the same query are a common and completely fixable problem — and fixing it costs nothing to write.
A page can rank first for a term with four thousand searches a month and produce almost nothing, while a page ranking third for six hundred fills your diary. We prioritise by what the searcher is trying to do, and we'll show you the working — including when the honest answer is that a topic isn't worth writing about.
The query, the intent behind it, the questions that must be answered and in what order, a length taken from what currently ranks rather than invented, and the internal links it has to carry. Nearly every piece of content that fails, fails here — by the time a draft is wrong, the decisions that made it wrong were taken days earlier.
Answer in the first eighty words. Structure so it can be skimmed and still make sense. One clear action. No throat-clearing introduction about how important the topic is. Search engines have spent years getting better at spotting content written for search engines, and the defence against that has always been the same: write for the person.
Published is not finished. Pages decay from about month eight, and a refresh usually recovers more traffic than a new piece would have won for less than the cost. Part of every month goes on updating what's already live — which is the part almost nobody budgets for and the reason most content libraries quietly shrink.
Content feeds everything else.
Almost every other channel is downstream of having something worth reading. This is where that gets made.
Search engine optimisation
Decides which pages are worth writing. Content is how the SEO roadmap actually gets built.
Pay-per-click
Tells us which messages convert before we commit a quarter to ranking for them.
Social media
One long piece cuts into a month of posts, so the same research pays twice.
Email marketing
The thing worth subscribing for, and the material every newsletter is cut from.
Content creation
Briefed from what already ranks, written for the person, and maintained after publishing.
You're hereWeb development
A slow page loses the reader before the first paragraph, however good the writing is.
Find out what's worth writing.
Send us the domain and we'll come back with which of your existing pages are decaying, which are competing with each other, and the five briefs we'd write first — before you've committed to anything.