Think of a plumbing and heating firm with a blog post called "Five benefits of a boiler service". It is fine. Grammatical, sensible. In substance it is the same post any competitor down the road could publish, changing nothing but the phone number.

That post was filed internally as thought leadership content. It was category content: the shared assumptions of an industry, restated politely.

The gap between those two things is what this article is about, and more usefully, how to test for it before you press publish.

The definition most articles give you, and why it fails

Search for a definition and you will get broad agreement. Thought leadership content contains original ideas. It offers a unique perspective. It builds credibility and shifts how people think. All true, all flattering, all completely useless at four o'clock on a Thursday when you are holding a draft and deciding whether it goes live.

None of those definitions give you grounds to say no. And a definition that cannot reject anything is not a definition, it is a mood.

Here is a working one. Thought leadership content is built on a claim the business has first-hand grounds to make, takes a position a reasonable person could argue with, and aims at a decision the reader is actually facing. Three conditions. Miss one and you have something else, which is fine, as long as you know what you have got.

Three things get confused with it regularly:

• Brand journalism: well-reported pieces about your sector, interviews, trend round-ups. Useful, often more readable than opinion, but the authority belongs to the people quoted, not to you.

• SEO explainer content: "what is X", "how does Y work". These exist to be found by people who do not know the answer. They should be clear and correct, not opinionated. A definition page arguing with itself is just confusing.

• Executive personal branding: the founder's LinkedIn posts. Different format, different rhythm, often genuinely good. But a 200-word post about a lesson learned is not a published position your sales team can send to a procurement committee.

Readers never see the label. Nobody has ever finished an article and thought "what strong thought leadership". They think: this person knew something I didn't. That is the whole measurement.

Four questions that separate real positions from filler

Run these on a draft. It takes about ten minutes and it is brutal the first few times.

1. The swap test

Could a direct competitor publish this piece unedited under their own logo? If yes, there is no leadership in it. None.

Most SME thought leadership fails this in the first paragraph, because it opens with the industry's shared assumption ("content is more important than ever for small businesses") rather than with something the author believes and others might not. The swap test is the fastest disqualifier we find, and it costs nothing to apply.

2. The disagreement test

Find one sentence in the draft that a knowledgeable peer could push back on. If every sentence would get a nod in a room full of experts, you have restated consensus, and consensus restated is the opposite of a point of view.

You do not need to be provocative. You need to be specific enough to be wrong.

3. The evidence test

Go claim by claim and ask what sits behind it. Your own client work. Your own aggregate data. A decision you got wrong and had to unpick. A regulation you had to implement before your competitors did. If the honest answer is "I read it somewhere", the claim either gets a named source or gets cut.

4. The decision test

Does the reader finish knowing what to do differently on Monday? Vaguely informed is not an outcome. "Audit your content"is not an instruction. "Open your three most recent blog posts and check whether any of them targets the same question"is.

What drains authority out of a draft

Flattened writing has tells, and once you can see them you cannot unsee them.

Hedge words everywhere: often, typically, many businesses, can help to. Sentences that all run to roughly the same length, so the whole piece hums at one pitch. Three-item lists that would apply equally to accountancy, dentistry and industrial coatings. A conclusion that restates the introduction in slightly different words, because there was nothing in the middle to conclude.

AI-assisted drafting makes this worse at scale, and we say that as people who use AI in our own workflow every day. The model is averaging the internet. The internet's average opinion on any topic is a careful, balanced, nobody-in-particular position, which is precisely the thing thought leadership cannot be. Most AI writing has a tell: it is competent, it is grammatical, and it sounds like it came from everywhere.

The fixes are editorial, not technical. Build a voice profile from the business's existing writing, so the draft starts from a real register rather than a default one. Keep a long banned-phrase list and actually enforce it. Vary sentence rhythm on purpose. Force a named thing into every abstract claim: not "modern security practices"but "MFA on the VPN and a 14-day patch window". That is roughly the editorial sequence we describe in how we research, brief and write every piece, and none of it is automatable in the way people hope.

Then the sceptical sub-editor pass. Read your own draft and write "how do you know?"after every assertion. Anything that cannot answer comes out. You will lose 200 words and gain the entire argument.

One honest caveat. If a business has barely published anything, there is no voice to learn from and no body of first-hand evidence to mine. Generating the first few pieces from the website alone will produce exactly the averaged mush described above. Those pieces have to come from a conversation: twenty minutes with the founder, a real client situation, an actual decision. Thin sites need interviews before they need volume.

You already have the raw material

Nobody needs to invent insight. Most businesses are sitting on more publishable evidence than they will use in two years.

Patterns across client work. The objection that comes up in every single sales call. The brief that went wrong and what you changed in your process afterwards. Back to the heating firm: the three failure patterns they see most often in ten-year-old combi boilers, with the model families involved, is a piece no competitor can copy, because the competitor has not seen those jobs.

Aggregate internal data, handled responsibly. "Across the 40 client sites we crawled last quarter, the most common gap was X"is publishable if the sample size is stated plainly, the data is anonymised, and no number is rounded up to sound better. Forty is forty. Do not write "dozens"because it sounds bigger, and do not write "nearly 50".

Your own operating decisions. Why you priced a service the way you did. Why you turned work down. Which tool you dropped and what it cost you to migrate. A SaaS founder who walked away from a six-figure enterprise deal because the integration would have wrecked the roadmap, and who then writes up the trade-off honestly, has produced something no agency could ghostwrite from desk research.

Build the capture habit: keep a running note of things you said out loud to a client this month that visibly surprised them. That note is your editorial calendar.

Fitting it alongside everything else you publish

A site cannot be all opinion. Position pieces need somewhere to point, which means definition pages and practical guides doing the unglamorous work of answering questions directly. The mix matters more than the ratio, but a site of pure argument with nothing underneath it reads as bluster.

Cannibalisation is the quiet killer of thought leadership programmes. Commission five pieces circling the same theme and they split their own signal, compete with each other, and leave you wondering why none of them ranks. Checking existing pages and previously published articles before commissioning is part of the editorial job, not housekeeping. A proper content gap analysis against competitors and your own archive catches this before a word is written.

Be realistic about search volume. Many thought leadership topics sit at very low monthly UK search volumes, and some have no volume at all. Judging them on rank alone will make you kill your best work. Success more often looks like a prospect quoting the article back at you in a sales call, or a procurement contact citing it as the reason you made the shortlist.

Cadence is itself a signal. A steady rhythm reads as a point of view developing over time. Twenty pieces in a fortnight reads as a campaign, and Google's spam policies specifically name scaled content abuse, meaning large volumes of pages produced mainly to manipulate rankings rather than help people. A ramp is safer and, frankly, more believable.

The ten-minute review before anything goes live

1. Swap test. Could a competitor run this unedited? Yes means reject, not revise.

2. Disagreement test. Name the sentence someone could argue with. If you cannot find one, there is no position.

3. Evidence test. Every claim gets a first-hand source or a named external one, with a date. Unsourced numbers come out.

4. Decision test. Write down, in one line, what the reader does differently after reading it.

5. Sub-editor pass. "How do you know?"after every assertion. Cut what cannot answer.

Write the rejection rule down before you start, because you will not want to apply it to a piece you have already paid for. Ours is simple: fail the swap test, the piece does not publish.

Grade on two axes, not one. A search score tells you whether a piece can be found. An integrity score tells you whether it deserves to be read. Those two disagree more often than people expect, and when they do, integrity should win. In practice that means killing a well-optimised article that says nothing new, which is an uncomfortable thing to do and the main reason most programmes drift towards filler. It is also why our content marketing platform grades every piece twice rather than handing over a draft and hoping.

Last thing: whoever's name goes on it has to read it properly. Not a summary, not the headings. Google's guidance on creating helpful, people-first content leans heavily on demonstrated experience and on who is behind the content, so a byline attributed to an expert who never read the piece is both an E-E-A-T risk and, more immediately, a reputational one when a client asks them about it.

And when a draft fails? Narrow the scope. Do not add words. A 900-word piece about one failure pattern you have personally fixed will outperform 2,000 words about the industry in general, every time.

Frequently Asked Questions

What counts as genuine thought leadership content?

Content built on a claim your business has first-hand grounds to make, taking a position a reasonable expert could disagree with, aimed at a decision the reader is actually facing. The quickest check is the swap test: if a competitor could publish it unedited under their own logo, it is category content rather than leadership.

How is thought leadership content different from a normal blog post?

A normal blog post answers a question the reader typed into Google, and it should do that clearly and without editorialising. Thought leadership argues something. It draws on your client work, your data or a decision you made, and it accepts that some readers will disagree. Both belong on a healthy site, and they do different jobs.

Can AI-assisted drafting produce genuine thought leadership content?

Only with real editorial input. Left alone, a model produces the averaged consensus of everything it has read, which is the opposite of a position. Feed it first-hand evidence, a voice profile built from your existing writing, and a proper human review pass, and it can do the structural work well. Skip those and you get something competent that sounds like nobody.

How often should a small business publish thought leadership pieces?

Steadily rather than in bursts. One or two solid position pieces a month, sitting alongside practical guides and definition pages, reads as a developing point of view. A sudden spike of twenty pieces reads as a campaign, and large volumes produced mainly for rankings fall under Google's scaled content abuse policy, current at the time of writing.

How do you measure whether thought leadership content is working?

Rarely by rank alone, because many of these topics have very low UK search volume. Look at assisted conversions, time on page from returning visitors, and the softer signals: prospects quoting the piece back to you, inbound enquiries that reference it, a shorter sales conversation because the objection was handled before the call. Track those alongside search data and you can see where your existing content is actually earning its place.