Ten structures. Your material inside them.
Templates don't write thought leadership — they stop you staring at a blank composer while the good story from Tuesday fades. Each one below tells you when to use it, gives you a skeleton you can paste straight into any model alongside your notes, and shows a worked example so you can see what 'specific' actually looks like.
What is the best template for an AI-written thought leadership post?
The objection teardown: quote the pushback in the buyer's words, name the concern underneath it, give your answer, and end on one claim a reader can disagree with. It works because the raw material already exists in your week, so the model has something specific to shape instead of inventing an opinion.
- A template fixes structure; you supply the specifics.
- Use two or three regularly, not all ten.
- Paste the skeleton plus your notes into the model.
- Rewrite the opening and closing line yourself.
- 120–250 words is the working range.
- Do post templates make you sound like everyone else?
- Only if you leave the template doing the thinking. A template fixes the structure — where the tension goes, where the evidence goes, where the claim lands. The specifics you drop into it are what make the post yours, and specifics are exactly what a model cannot invent for you.
- How many templates should I actually use?
- Two or three. Rotating ten structures every week makes your feed feel restless and gives you nothing to be known for. Pick the two that match the material you naturally collect — usually the objection teardown and the lost-deal lesson for sellers — and use the rest occasionally.
- Should I paste the template into ChatGPT or write it myself?
- Paste the template plus your raw notes and let the model produce the first draft, then rewrite the opening and the closing line yourself. Those two lines carry the voice and the judgement; the middle is structure, which is the part a model handles well.
- How long should a LinkedIn thought leadership post be?
- Between 120 and 250 words for most of these templates. Long enough to carry one specific story and one claim, short enough that a buyer reads it standing in a queue. If a draft runs past 300 words, it usually contains two posts.
How to use these
ach template is a container, not a script. Pick the one that matches the material you actually have, paste the skeleton into a model together with your raw notes, and ask it to fill the structure using only what you gave it. Then rewrite the first line and the last line by hand — those two carry the voice and the judgement, and they are the two a model consistently softens.
If you have not set up the capture habit that feeds these, start with the full AI thought leadership workflow. Templates without source material just produce faster wallpaper.
The ten templates
Ordered roughly by how often a working seller can use them. The first three will cover most of your weeks.
The objection teardown
Use it when: A buyer pushed back this week and your answer was better than usual. The most reliable template a seller owns, because objections repeat.
"[The objection, in the buyer's exact words]" I heard that again [this week / three times this month]. Here's what's underneath it: [the real concern behind the stated objection]. [What you said, or what you'd say now.] [The one line a reader can quote back: the claim.]
"We already have someone doing this internally." Heard it twice this month. What they usually mean is: I don't want to explain a new line item to my CFO. So I stopped arguing about capability and asked what the internal person had to stop doing to take it on. Both times, the answer was the thing their boss actually cared about. The competitor isn't the other vendor. It's the opportunity cost nobody has priced.
The lost-deal lesson
Use it when: You lost something and the real reason is not the reason in the CRM. Highest-trust template there is, because almost nobody publishes losses.
We lost a deal [timeframe] because of [the real reason]. Not [the reason it looked like from outside]. [What actually happened, two or three sentences, concrete.] What I'd do differently: [the specific change]. [The claim other people in your position should take from it.]
We lost a deal last quarter because our pricing page did the talking before we did. Not on price. On a forwarded link. The champion sent it to two people we never met. They read "starting at", assumed double, and killed it before we got a second call. What I'd do differently: ask early who else will see this, and give them something written for that person. Every asset a champion can forward is part of your pitch, whether you're in the room or not.
The question asked twice
Use it when: Two different buyers asked the same thing in the same week. That is a market signal, and answering it publicly saves you the third call.
Two [buyer role]s asked me the same question this week: "[The question]" Short answer: [the direct answer in one sentence]. Longer answer: [two or three sentences with the nuance and the trade-off]. If you're asking it too, [the thing to check first].
Two RevOps leads asked me the same question this week: "How do we know if the data is bad or the process is bad?" Short answer: change one and see if the number moves. Longer answer: bad data shows up as inconsistency between sources; bad process shows up as consistency you don't like. If every report agrees and the number is still wrong, it isn't a data problem. If you're asking it too, start by counting how many places the same field gets typed by a human.
The contrarian take
Use it when: You genuinely disagree with something your industry repeats. Only use it when you would defend the position on a call — a borrowed contrarian take collapses in the comments.
[The accepted wisdom] is mostly wrong. Here's what people miss: [the reason it became accepted]. In practice, [what actually happens, with a specific instance]. The version that works: [your alternative]. I'll take the disagreement in the comments.
"Always be adding value" is mostly wrong. It became accepted because it's the polite opposite of pestering people. In practice it produces a lot of forwarded articles nobody asked for. I've had buyers tell me the "value-add" emails were the reason they stopped replying. The version that works: only reach out when you know something they don't about their situation. If you don't, don't. I'll take the disagreement in the comments.
The surprising customer quote
Use it when: A customer described the problem or your product better than your own marketing does. Their language outperforms yours because it is the language buyers search with.
A customer said this to me [timeframe]: "[The quote, verbatim, permission-safe or anonymised]" I've been describing it as [your worse version] for [time period]. [Why their framing is better — one or two sentences.] [What you changed because of it.]
A customer said this to me on Tuesday: "I don't need more content. I need to stop sounding like a brochure." I'd been describing the problem as "content velocity" for about a year. Velocity is a metric. Sounding like a brochure is a feeling, and it's the one that actually stops people posting. We changed how we open every demo because of one sentence.
The mini teardown
Use it when: You can show a small, concrete before-and-after — a message, a page, a call opening. Show the work; don't just assert the improvement.
Before: "[the weak version]" After: "[the stronger version]" What changed: [the single principle, not five]. Why it matters: [the effect on the reader or buyer].
Before: "Just checking in to see if you had any thoughts on my last email." After: "You mentioned the board meeting was the 14th. Did the budget line survive it?" What changed: the follow-up now references something only I would know about them. A check-in asks the buyer to do the remembering. A specific question does the remembering for them.
The framework post
Use it when: You have a repeatable way of doing something. Frameworks get saved and re-shared, which is the closest thing to compounding reach on LinkedIn.
Every time [situation], I run the same [number] checks: 1. [Check] — [why] 2. [Check] — [why] 3. [Check] — [why] Most people skip [the one that matters most], which is why [the common failure].
Every time a deal goes quiet, I run the same three checks: 1. Did the champion change jobs — half of "went dark" is just a leaver. 2. Did we ever meet the person who signs — if not, we were never in a deal. 3. Did the problem get cheaper to ignore — budgets don't vanish, priorities do. Most people skip the second one, which is why forecast slippage always looks like a surprise.
The myth-bust
Use it when: A widely repeated claim in your field is measurably false or badly misused. Requires evidence you can defend, not just an opinion.
Myth: [the claim as people state it]. Where it came from: [the origin or the grain of truth]. What's actually true: [the correction, with your evidence]. What to do instead: [the practical replacement].
Myth: posting more often always grows your reach. Where it came from: it's true at the very start, when almost nobody has seen you. What's actually true: past roughly two substantive posts a week, the extra posts mostly compete with each other for the same audience. My own worst-performing week was the one with five posts. What to do instead: keep the second post and spend the third post's time replying to comments on the first two.
The number you can defend
Use it when: You have a real figure from your own pipeline, product or process. Never use this template with a statistic you found in a blog post you did not read.
[The number], from [where it comes from — your own data]. Context: [what it's measured across, so it's honest]. What it tells me: [the interpretation]. What it doesn't tell me: [the honest limitation].
Eleven of our last forty inbound conversations started with someone saying they'd seen a post. Context: that's our own pipeline over one quarter, and it's self-reported on the first call. What it tells me: publishing is doing more of our top-of-funnel work than our outbound sequences are. What it doesn't tell me: whether those eleven would have found us anyway. Attribution on this is soft and I'm not going to pretend otherwise.
The prediction with a stake in it
Use it when: You believe something about where your market is going and you are willing to be publicly wrong. Vague predictions are worthless; put a date and a condition on it.
By [specific timeframe], [the specific prediction]. Why: [the mechanism driving it, not just a trend]. What would prove me wrong: [the falsifier]. Saving this so you can hold me to it.
By this time next year, "we use AI to write it" will be a neutral fact in B2B, the way "we use a CRM" is now. Why: the tooling gap closes faster than the taste gap. Once everyone drafts with a model, the differentiator moves back to whether you have anything to say. What would prove me wrong: buyers starting to treat AI disclosure as a trust signal in itself, rather than ignoring it. Saving this so you can hold me to it.
Choosing your two
Being known for something requires repetition, and repetition requires a structure you can fill without thinking. Pick the two that match the material your job produces naturally: sellers usually land on the objection teardown and the lost-deal lesson; founders on the contrarian take and the number they can defend; marketers on the mini teardown and the framework.
Run those two for a month before adding a third. If a template keeps producing drafts you do not want to publish, it is the wrong template for the material you have — not a discipline failure.
Templates cannot rescue a thin note.
Frequently asked questions
Do post templates make you sound like everyone else?
How many templates should I actually use?
Should I paste the template into ChatGPT or write it myself?
How long should a LinkedIn thought leadership post be?
Can I reuse a template with the same story later?
Skip the blank composer.
SelloWise holds your voice, ICP and objection library as persistent memory — so every draft starts from your material, in a structure that fits it.