Services consulting
Content marketing built to be cited, not to fill a quota
Content marketing consulting means deciding which topics you should own, what depth each one needs, and what evidence a sceptical reader would require — then producing content that ranks and gets retrieved by AI answer engines.
What is topical authority and why does it matter?
Topical authority is the state of being a recognized, complete and credible source on a subject — not merely having one good page. It matters because search engines and AI retrieval systems both prefer sources that have demonstrated comprehensive coverage, and because a buyer who lands on any page of your site should find it worth continuing to read.
Building it is mostly a coverage problem. If you have three articles on a topic where a competitor has twenty that collectively answer every question a buyer asks, you are not competing on quality — you are simply not a candidate. This is why a gap analysis against the competitive set is the first step, and why content volume is sometimes genuinely the answer even though volume alone is not.
The distinction to hold onto: coverage is the goal, and volume is sometimes the only way to achieve it. Volume as an end in itself — a target of posts per month, justified by an article count — reliably produces thin pages that compete with each other and dilute the authority of the strong ones.
Do you need more content, or better content?
Usually better, and occasionally considerably fewer. The test is whether your existing pages are competing with each other for the same intent. Where they are, consolidating them into one authoritative page typically outperforms maintaining both, and frees the crawl budget the duplicates were consuming.
- Posts targeting the same query with near-identical intent — consolidate.
- Thin posts with no impressions after six months — consider folding into a stronger page or deleting outright.
- Posts that historically converted but have decayed — refresh rather than replace, preserving the URL and its accumulated signal.
- Posts on topics outside your actual expertise — these attract traffic you cannot convert and dilute topical focus.
What goes into a content brief?
A content brief exists to let a writer produce something that can rank without the writer having to be a subject-matter expert or an SEO. Vague briefs produce generic content. A useful brief specifies seven things.
- Target query and intent. The query, and what the searcher is actually trying to accomplish. Informational, commercial, transactional or navigational — this determines the format.
- The direct answer. The 1–2 sentence answer that must appear in the first paragraph. Writing it first forces clarity.
- Required subtopics. Every question a comprehensive answer must address, derived from what the top-ranking pages cover and what your own customers ask.
- Evidence required. Which statistics, studies, official documentation or first-party data the piece needs. This is what makes content citable and is the most commonly omitted section of a brief.
- Internal linking targets. Three to five specific URLs this page should link to and which should link to it, with anchor text suggested.
- AI extractability note. Confirmation that each H2 fully answers its own heading standalone.
- Entity and format requirements. Named entities to reference, and whether a table, list, FAQ or worked example suits the intent better than prose.
How do E-E-A-T and authorship signals work?
E-E-A-T — Experience, Expertise, Authoritativeness, Trust — is Google's conceptual framework for assessing content quality. It is not a direct ranking factor and there is no score you can optimize. What it describes in practice is whether a reader and a system can establish that the content was written by someone qualified to write it.
The practical signals are unglamorous: a named author with a real biography and credentials, consistent authorship across a topic cluster, primary-source citations, first-hand experience described specifically, and a visible connection between the author and the organization. Fabricating authority is both unethical and detectable, and it is the single most common failure in AI-generated content at scale.
Why does content decay and how do you fix it?
Content decays for identifiable reasons: the page mentions statistics or features that have changed, competitors have published something better, the query intent has shifted, or the page's supporting evidence has been superseded. The decay is usually gradual and goes unnoticed until the traffic step-change.
Refresh cycles should be scheduled rather than reactive. Quarterly review of your top pages by traffic and by citation is a reasonable cadence for anything commercially important; annually for the rest. The refresh is a genuine content update — new data, new screenshots, changed conclusions — not a date change in the footer, which Google can detect and which readers notice.
How do you write content that AI answers will cite?
AI retrieval lifts passages, so the writing unit is the section. Four practical changes to normal content production produce most of the effect.
- Lead with the answer. First one to two sentences of each section fully answer the heading. Do not build up to it.
- Make each section standalone. No "as mentioned above", no pronouns pointing at content further down the page.
- Anchor facts. Names, numbers, dates and named sources in the same sentence as the claim. "Cloudflare Turnstile is free" is citable; "it is pretty good value" is not.
- Phrase headings as questions. Match how a person would ask, because that is how the retrieval query gets formed.
These are not exotic techniques — they are what good reference writing has always done. The difference is that in 2026 they are also the conditions for being retrieved by a model. The full evidence is in the AI Overviews guide, and the process in AI search optimization.
Related services
Frequently asked questions
How much content should we publish per month?
That is a consequence of prioritisation, not a useful target in itself. We would rather you published one authoritative page that covers a topic completely than four that overlap. If your site genuinely lacks coverage on topics where you can win, then volume is the answer — but set it by gap analysis rather than by a monthly target agreed in January.
Is AI-generated content bad for SEO?
Not because a model wrote it, but because unreviewed generated content is typically generic, inaccurate and identical to everything else. Google's policy targets scaled content abuse regardless of authorship. AI is entirely reasonable for research, structure, outlines and first drafts, provided a qualified human verifies the facts and adds something only you could know — your data, your experience, your position.
How long should a blog post be?
Long enough to answer the question completely, and no longer. There is no target word count that ranks. A precise 900-word answer beats a padded 2,500-word article every time, because padding introduces unsupported claims and dilutes the passages most likely to be retrieved.
Do I need to update old content or delete it?
Update when the underlying information has changed and the page still has value — preserving the URL and its accumulated authority. Delete or consolidate when the page competes with a stronger page of your own, or targets a topic outside your expertise. Redirect rather than delete if it ever earned links.
Should I gate content behind a form?
Generally no. Gated content performs worse in search, is cited less by AI systems that cannot access it, and modern buyers are conditioned to expect an answer immediately. Gating works for genuinely proprietary research where the data itself is the product — annual benchmark reports, for instance. Ungated is the right default.
What is information gain and does it matter?
Information gain is content that adds something genuinely new: original data, a case study from your own work, a benchmark you ran, experience that cannot be synthesised. It is the most durable ranking and citation asset available, because it is the one thing a language model cannot generate — it has to come from somewhere. If you are not producing any, that is the highest-value gap in most content programs.
How do you know which topics to prioritize?
Gap analysis against the pages currently ranking for your commercially important queries, weighted by how close each gap is to being closable for a site of your authority. We look for topics where the ranking pages are weak, not where the competition is strongest — winning a topic you can plausibly own beats losing efforts on the biggest competitor.
Who should write the content — us, you or AI?
That is your decision and it depends on internal capability. What we insist on is that whoever writes it has access to subject-matter expertise from inside your business. Content written entirely from outside lacks the specific detail that makes it credible, and that detail is also what makes it citable.
Need content marketing for your business?
Get a free, no-obligation review of where you stand today on content marketing — and the three changes that would move the number that matters. No pitch deck.
Book a free 30-minute audit