What did your last opinion piece cost the reader to believe? That question sounds odd until you notice that belief always has a price. Executives spend real money learning this the slow way: a ghostwriting retainer, a year of weekly posts, a content agency producing competent paragraphs, and at the end of it, no inbound, no speaking invitations, no journalist calling for a quote, because nothing published was an opinion anyone needed the author specifically to hold. When a reader accepts your claim, they spend something: attention, the risk of repeating it to their boss, the chance of being wrong in public. Most executive content fails because it asks readers to pay that price for opinions that cost the author nothing to hold. “Culture matters.” “AI will change everything.” “Customer obsession wins.” Free for you to say, worthless for anyone to believe.
The fix is a filter we call the Earned Opinion Test, and it has three parts. An opinion is earned when you paid for it with experience (you ran the experiment, took the loss, made the call), when it is falsifiable enough that someone could disagree without being an idiot, and when it is specific enough that your competitor could not publish the same paragraph by swapping the logo. Run every draft against those three parts before you write thought leadership content under your own name, because readers run the same test unconsciously, and so do the AI engines now deciding which voices to cite.
Start from decisions, not topics
Topic-first writing produces the content graveyard: pick “the future of logistics,” summarize what everyone knows, add a closing line about embracing change. Decision-first writing starts from a moment you chose under uncertainty. Why you killed a product line that was still profitable. Why you pay above market in one role and refuse to in another. Why you turned down the bigger client. Decisions force everything the Earned Opinion Test demands: stakes, a position someone could oppose, and details only you possess. Keep a running list of the ten hardest calls you made each quarter; that list is the editorial calendar nobody can copy. Decision-first writing also fixes the consistency problem that kills most executive publishing programs by month four, the sense of having nothing left to say, because a business that makes decisions keeps refilling the list whether or not you feel inspired.
The objection arrives immediately: my hardest decisions are confidential. Some are, and the workable answer is abstraction one level up, writing the principle the decision taught rather than the deal that taught it. You cannot name the client you fired, but you can write “the three contract clauses I now refuse after a seven-figure engagement went sideways” without identifying anyone. The numbers can be ranged, the names removed, and the lesson keeps its teeth as long as the cost stays in the story. What kills the piece is removing the cost along with the names, which is how decision stories decay back into the platitudes the whole method exists to escape. If the sanitized version reads like something a consultant who never lived it could write, it failed the test; put it back in the drawer and pick a decision you can afford to tell.
Argue against a reasonable person
Weak thought leadership pretends opposing views do not exist. Strong pieces name the smartest version of the other side and engage it, because readers trust writers who demonstrate they understand the disagreement. The structure is old and works: here is the conventional position, here is why reasonable people hold it, here is the specific evidence that moved me off it, here is where my view would break. That last element, stating what would change your mind, is rare enough in executive writing that it functions as a credibility signal all by itself.
A practical drill for finding the reasonable opponent: write the strongest one-paragraph case against your position before drafting the piece, as if a smart competitor were writing it. If you cannot produce that paragraph, you do not understand the topic well enough to publish on it yet, and the research that fixes the gap will improve the piece more than any editing pass. If you can produce it and it defeats your draft, you just saved yourself from publishing something a commenter would have dismantled in public. The pieces that survive this drill come out with a quality readers can feel but rarely name: the sense that the author has already been everywhere the reader’s skepticism might go, and came back with answers.
Write the version only you could write
Before publishing, run the swap test. Replace your name and company with a competitor’s. If the piece still works, it was not thought leadership, it was category content wearing your byline. The unswappable ingredients are always the same three: your numbers (real figures from your operation, even ranged for confidentiality), your scars (the failure that taught the lesson, with the cost stated), and your position in the system (what your seat lets you see that analysts and journalists cannot). When you write thought leadership content around those three ingredients, imitation stops being a risk because imitation requires living your last decade.
The swap test also explains why most ghostwritten executive content fails even when the writing is competent. A ghostwriter without access to the numbers, the scars, and the seat produces swappable prose by definition, because the unswappable material lives in the executive’s head and calendar. The arrangement that works keeps the thinking with the principal and delegates only the polishing: a monthly hour of recorded interrogation, the writer drafting from the transcript, the executive cutting anything they would not say across a dinner table. What never works is handing a writer the topic list and the brand guidelines and expecting earned opinions to emerge, since opinions cannot be earned on someone else’s behalf.
Pick formats that survive being quoted
Earned opinions travel as fragments: a pulled quote, a screenshot, a citation in an AI answer. Build pieces around one quotable claim stated in one or two sentences, supported beneath. The formats that hold up are the argued essay (one claim, evidence, counterargument, implication), the decision memoir (the call, the context, the result, the principle), and the data take (one number from your operation and what it overturns). Formats that collapse under quotation, the listicle of platitudes, the trend roundup, the vendor pitch in a trench coat, also fail the engines, which extract and attribute claims rather than reprinting essays whole. A claim that survives out of context is a claim that compounds, and that is the practical reason to write thought leadership content as arguments rather than vibes.
Practice the compression deliberately: before publishing anything, write the one-sentence version you would want quoted, and put that sentence in the piece verbatim, early. If you cannot produce the sentence, the piece has no claim yet, only a topic. If the sentence embarrasses you with its boldness, you are probably close to something worth publishing, because the discomfort usually marks the spot where you stopped hedging. Executives consistently underestimate how much hedging costs: a claim wrapped in “in many cases, depending on context, it may be worth considering” cannot be quoted, cannot be disagreed with, and therefore cannot build a reputation, since reputations are built out of the disagreements a writer survives.
Publish where standing accrues
The same essay earns differently depending on where it lives. Your own site builds the canonical archive engines associate with your name. A trade publication byline adds editorial validation. LinkedIn adds reach and the discourse layer. The common mistake is treating the three as interchangeable and posting everything everywhere simultaneously, which wastes the trade outlet’s exclusivity expectations and buries the canonical version under its own syndication. The sequence that works: canonical version on your site, adapted version pitched to one trade outlet per quarter, fragments to social with links back. Over a year that produces a citation trail, your name plus your claims, distributed across domains the engines treat as independent confirmation, which is how a person becomes the answer to “who is worth reading on this.”
Consistency of the byline matters as much as placement. The same name, the same title framing, the same one-line bio across every venue lets both editors and engines resolve the scattered pieces into one entity, while a byline that shifts between roles and descriptions fragments the trail you spent a year building. Pair the publishing sequence with a personal page that lists every placement, because that page becomes the canonical record an AI engine retrieves when someone asks who you are, and an executive with twelve argued essays across four respected venues gets described very differently by that machine than one with a hundred unsourced social posts. The engines, it turns out, run their own version of the Earned Opinion Test: they cite the voices whose claims are specific, attributed, and independently confirmed, which means writing to pass the human test and writing to pass the machine test became the same job somewhere around 2025.
A cadence note, because the question always follows: how often can one executive produce work that passes the test? For most operators the honest answer is monthly, sometimes less, and the instinct to fill the gaps with lighter content is the instinct to resist. Audiences do not penalize silence; they penalize dilution, because every platitude published under your name retrains readers on what the byline is worth. The portfolio that builds a reputation over two years is roughly twenty pieces that all passed, not a hundred where the good ones drown.
One discipline matters more than any tactic here: publish only when you have an earned opinion, even if that means once a month instead of twice a week. The executives who become reference points are not the loudest. They are the ones whose every piece passed the test, until the byline itself became the signal.