Thought leadership is writing or speaking that offers an original, well-supported point of view on a problem your readers care about, published under a real person's or organization's name. LinkedIn's guide to thought leadership for marketers describes a thought leader as someone who “has formed a clear, unique and defensible point of view,” and defensible is the useful word: the view has to be one a reasonable reader could dispute. It is a reputation, earned one piece at a time, not a content format. The examples that work share three things: a claim someone could disagree with, evidence the author has and others don't, and a reader who can use it.
Below are eight published pieces, each with the claim it makes and what to borrow, then eight samples in the formats most people actually write. The samples' people, companies and data are invented, so their numbers illustrate the form, not real findings.
Thought leadership formats
Most thought leadership takes one of a handful of forms. The right one depends less on the audience than on the evidence you have.
| Format | Best for | Example on this page |
|---|---|---|
| FormatOp-ed or guest essay | Best forOne argument, tied to the news, in a publication others trust | Example on this pageMarc Andreessen in The Wall Street Journal |
| FormatEssay | Best forAn idea that needs room, published on your own site | Example on this pagePaul Graham |
| FormatAnnual letter | Best forA leader's account of the year and the reasoning behind decisions | Example on this pageWarren Buffett's shareholder letters |
| FormatNewsletter | Best forA recurring point of view on one beat | Example on this pageStratechery |
| FormatTalk or keynote | Best forOne idea, simple enough to repeat after one hearing | Example on this pageSimon Sinek at TEDxPuget Sound |
| FormatResearch report | Best forOriginal data that others will cite | Example on this pageThe Edelman Trust Barometer |
| FormatPrinciples or culture document | Best forHow an organization actually decides | Example on this pageThe Netflix culture memo |
| FormatBook | Best forA body of thinking in one place | Example on this pageStripe Press |
| FormatLinkedIn posts and articles | Best forFrequent, short points of view under a person's name | Example on this pageSamples 1 and 7 below |
Several formats have their own page on this site: the structure of an op-ed, annotated white paper examples, and how to write a newsletter. This page is about what makes any of them leadership rather than content.
Eight real thought leadership examples
These are well known for a reason. Each is described from its own published source, with at most a few words quoted.
An op-ed: “Why Software Is Eating the World”
Marc Andreessen, co-founder of the venture capital firm Andreessen Horowitz, published this piece in The Wall Street Journal on August 20, 2011. Against talk of a technology bubble, he argued that software companies were taking over large parts of the economy and that their valuations reflected real businesses. The claim readers could dispute was that this wasn't a bubble. The evidence was a run of industries where a software company had already won, Amazon over Borders and Netflix over Blockbuster among them, and he disclosed his own firm's investments in the sector.
What to borrow: a thesis short enough to repeat, argued against the conventional wisdom of the moment, by someone with the standing to make it.
An annual letter: Berkshire Hathaway's shareholder letters
Warren Buffett's letters to Berkshire Hathaway shareholders are posted on the company's website going back to 1977. They are written for shareholders, in plain language, and they explain not only the year's results but the thinking behind the company's decisions. An annual letter is thought leadership with a deadline built in.
What to borrow: write for an intelligent reader who doesn't work in your field, and explain decisions, including the ones that went badly. A yearly letter from a founder or a managing partner can do the same job at any size.
An essay: “Do Things that Don't Scale”
Paul Graham, a co-founder of the startup accelerator Y Combinator, published this essay on his website in July 2013. The claim a reader could dispute is that a startup's early growth should come from work that can't scale, such as recruiting users one at a time. The evidence is what he watched at Y Combinator: the Airbnb founders going door to door in New York, and the Stripe founders' habit, which he calls the “Collison installation,” of setting a new user up on the spot. The title is an instruction, and the essay is addressed to one kind of reader.
What to borrow: a title that tells the reader what to do, advice that contradicts the default, and a reader defined tightly enough that it can be specific.
A newsletter: Stratechery
Ben Thompson started Stratechery in 2013 and has written it full-time since 2014. It covers the strategy and business side of technology and media. Weekly articles are free, and subscribers get three Daily Updates a week, so the free pieces show the thinking and the paid ones serve the regular readers.
What to borrow: one beat, a fixed schedule and a recognizable way of analyzing things, kept up for years. A newsletter earns its standing over many issues, which is also why a slipping schedule hurts it.
A talk: “How great leaders inspire action”
Simon Sinek gave this talk at TEDxPuget Sound in September 2009. Its idea is simple: leaders who inspire people start by explaining why they do what they do, not what they do. One idea, a structure the audience can hold in their heads, and examples that make the abstraction concrete.
What to borrow: a single framework people can repeat after one hearing. For building a talk around one idea, see how to give a TED talk.
A research report: the Edelman Trust Barometer
Edelman, the communications firm, says it has studied trust for more than 25 years, and it publishes a new Trust Barometer every year. A communications firm, in other words, publishes the yearly data on the very thing it advises clients about.
What to borrow: pick one question that matters to your clients and ask it the same way every year. From the second year on, readers can see what changed, and that comparison is evidence no one else has.
A culture document: the Netflix culture memo
Netflix publishes its culture memo on its careers site, built around four principles, and dates its first culture deck to 2009. The memo says plainly that the culture, and the document, will change over time. It is thought leadership about management, written for candidates and employees, and it doubles as a recruiting tool.
What to borrow: write down how your organization actually makes decisions, specifically enough that someone could disagree, and publish it with a promise to revise it.
Books: Stripe Press
Stripe, the payments company, runs a publishing imprint, Stripe Press, with the tagline “Ideas for progress.” It publishes works about technological, economic and scientific advancement, among them Stewart Brand's Maintenance: Of Everything, Part One and Nick Sleep's Nomad Letters. Neither is about payments.
What to borrow: back ideas bigger than your product, in work good enough to stand without the company's name on it.
Eight original samples to borrow
Most thought leadership is smaller than a TED talk: a post, a newsletter, an op-ed in a trade journal. These samples show the shape of each format at that scale. The people, companies, surveys and results are invented.
We stopped paying sign-on bonuses to caregivers in 2023. Retention went up.
The bonus was $1,000, paid after 90 days. It brought people in, and a lot of them left on day 91. When we asked caregivers who stayed what mattered, almost none of them mentioned money up front. They mentioned knowing their hours.
So we moved the bonus budget into a guarantee: every caregiver who wants 30 hours a week gets paid for 30 hours, even when a patient cancels. It costs us more in a slow month and less in every other way.
Two years later, one-year retention is 71%, up from 52%, and missed visits are down by more than half.
Home care runs on reliability, and so do the people who provide it. I'd like to hear from other operators who have tried something similar.
Why it works. The first line is a claim with a result, and the evidence is the company's own. The post explains the mechanism, admits the cost, and ends by inviting peers in, which turns the post into a question for the industry. More examples of the form are in LinkedIn post examples.
Last August, on the hottest evening of the year, the state's grid came within 900 megawatts of rolling blackouts. That same week, enough battery storage to cover the gap sat finished and unconnected in three counties, waiting for permission to plug in.
The wait was not about safety or cost. It was about information. Developers like us have no public map of where the grid has room for new storage, so we apply blind, the utility studies each application one at a time, and most of them fail.
The Public Utility Commission can fix that this fall by requiring utilities to publish that map. It would cost them little and save every summer after this one.
Why it works. The news peg and the scene come first, the thesis lands in the third paragraph with a named decision-maker and a date, and the author's stake is visible rather than hidden. The second paragraph does the job of a nut graf: it explains the problem in terms any reader can follow.
Write the memo before the meeting
Every engineering organization I have run has had the same problem by its second year: more meetings, worse decisions. The fix that worked every time was not fewer meetings. It was a rule that any meeting called to make a decision starts with a written memo, sent the day before, of no more than two pages.
The memo does three things a slide deck can't. It forces the author to finish the thought, because a sentence has to have a subject and a verb and a bullet doesn't. It lets the quietest person in the room arrive with an opinion. And it leaves a record, so that a year later, when someone asks why we chose the expensive database, the answer is a link and not an argument.
Here is how we write them, and the three ways they go wrong.
Why it works. The title is an instruction, the first paragraph names the problem and the rule, and the second argues for the rule with three concrete reasons. The last line promises the practical part, which keeps a skeptical reader going.
The State of Warehouse Work 2026: key findings
- Predictable schedules now outrank pay as the main reason warehouse workers stay, cited by 46% of those planning to stay a year or more.
- Workers whose shifts are posted at least two weeks ahead are half as likely to quit in their first six months.
- Only 18% of managers said schedules were a retention problem at their site. Among their workers, 61% said they were.
How we did it. We surveyed 2,412 hourly warehouse workers and 388 managers in 14 states in August 2026, and matched survey answers to six months of anonymized job records from 61 sites we staff.
What it means. The cheapest retention tool most warehouses have is a scheduling policy. Page 9 shows what the sites with the lowest turnover do differently.
Why it works. Three findings, each a number a reader could quote, including one that contradicts managers' beliefs. The method sits on the first page, not in an appendix, because careful journalists and skeptical readers look for it before they cite anything.
The Dock Door, issue 112: the whiteboard by the time clock
I visited three warehouses this month, and the best-run of them had the cheapest piece of technology: a whiteboard next to the time clock, with yesterday's numbers and one handwritten line from the shift lead.
This week's argument is that operations information belongs where people already stand still. Screens on the floor get ignored because nobody stops in front of them. The line at the time clock happens twice a shift, every shift.
Three things to try before you buy another dashboard:
- Move the one number your shift cares about to the time clock, the break room or the dock door.
- Let the shift lead write one line a day, by hand.
- Ask in a month whether anyone read it. Then ask what they would change.
Reply and tell me where people stand still in your building.
Why it works. Each issue makes one argument from something the writer saw that week, and gives readers three things to try. Asking for replies can turn readers into sources for future issues, one way a newsletter on a narrow beat keeps finding material.
Five years ago our maternity unit closed. It was losing money, we couldn't staff it, and the nearest other unit was 40 minutes away. It was a reasonable decision. It was also wrong.
In the year after it closed, the number of women in our county who gave birth without a single prenatal visit doubled. Some of them delivered in our emergency department, which is not where anyone should have a baby.
We reopened the unit in 2023. It still loses money, though less than it did. This morning I want to talk about why we decided that some services are worth running at a loss, how we chose which ones, and what it did to the rest of the hospital.
Why it works. The speaker opens by admitting a mistake, then gives the evidence that changed her mind: the doubling of births with no prenatal visit, and babies delivered in the emergency department. The admission carries weight because the evidence behind it is specific. The last paragraph tells the audience exactly what the next 20 minutes will cover.
How we decide at Gridline
- The person closest to the problem decides. Everyone else advises.
- Decisions that can be undone get made in a day. Decisions that can't get a written memo and a week.
- We write down why, not just what, so the next person can disagree with the reasoning instead of guessing at it.
- Disagree before the decision, commit after it, and reopen it only with new information.
- When a subcontractor tells us the software is wrong, we assume they are right until we have watched them use it.
We will revise this list every January. Last year's version, and what we changed, is linked below.
Why it works. Each principle is specific enough to break, which is what makes it a principle and not a slogan. The fifth is about the company's customers, so the memo tells prospective clients something too, and publishing the old version shows the principles are used.
Subject: Pitch: Schedules, not pay, keep warehouse workers (new survey data)
Dear Mr. Patel,
Our survey of 2,412 hourly warehouse workers, released next Tuesday, found that predictable schedules now outrank pay as the main reason they stay. I'd like to write a 900-word piece for your readers on what that means for managers, with three scheduling changes the lowest-turnover sites in our data have made.
I run research at a staffing company that places 9,000 warehouse workers a year. The full report and method are attached.
Thank you for considering it,
Ines Duarte
Why it works. The pitch leads with the finding, says what the reader of this magazine gets, and gives the author's standing in one line, and it offers the method up front, because a trade editor will ask for it anyway.
Finding your point of view
The bar is high. In LinkedIn and Edelman's 2021 B2B Thought Leadership Impact Study, as reported on LinkedIn's marketing blog, 71% of decision-makers said less than half of the thought leadership they consume gives them valuable insights. The steps below aim at the part that does.
- Start from an argument you have had
Think of a position you hold that a reasonable colleague disputes. If nobody would disagree with it, it isn't a point of view yet; it is a value statement.
- Find the evidence only you have
Your data, your cases, your customers' questions, your mistakes. This is your standing, the ethos that makes readers listen, and it is the one thing a competitor can't copy.
- Name the reader and the decision
Who will do something differently after reading it, and what will they do? “Warehouse managers will post shifts two weeks ahead” is a reader and a decision. “Leaders in our space” is neither.
- Let the evidence choose the format
Data becomes a report or a pitch. A story becomes a post or a keynote opening. An argument tied to the news becomes an op-ed. A method becomes an essay or a principles memo.
- Publish under a name, then keep going
One piece is an opinion. A reputation comes from a series on the same beat, each piece with its own evidence. Pick a pace you can keep, and keep it.
Thought leadership vs content marketing
The two overlap, and a company can do both. The difference is what the reader gets out of the piece if they never buy anything.
Thought leadership
- Argues a position someone could dispute
- Uses evidence the author owns
- Useful even to readers who never buy
- Signed by a person, or by a firm that stands behind it
- Measured in reputation: invitations, citations, replies
Content marketing
- Answers questions buyers search for
- Often summarizes what is already known
- Leads toward a product or a sign-up
- Often unsigned or signed by a brand
- Measured in traffic and conversions
A how-to guide can be both, if it argues for a method the author has tested and others haven't. A white paper is thought leadership when it makes a case the reader could reject, and a brochure when it doesn't.
Writing thought leadership for others
Much executive thought leadership is drafted by ghostwriters and communications teams, and that is fine on one condition: the ideas, the evidence and the final approval belong to the person whose name is on it. The writer's job is to find the argument in an interview, structure it and make it sound like the author on a good day.
- Interview for positions, not topics. “What do you believe about hiring that your peers don't?” gets an op-ed. “What should we write about?” gets a listicle.
- Keep a positions file. Five or six claims the author holds, each with its evidence and the stories that go with it. Every piece draws on the file, which keeps the body of work consistent. Before each new piece, write one line for each test: the claim someone could dispute, the evidence the author owns, the reader who can act. If a line stays empty, the piece isn't ready.
- Never invent the evidence. No made-up anecdotes, numbers or quotes. If a piece needs a detail, ask for it.
- Get sign-off on every word. The author answers for the piece in public, so they approve it in private first.
The business side of this work, from rates to contracts, is in how to become a ghostwriter.
Thought leadership mistakes
- Predictions with nothing at stake. “The future of work is hybrid” costs the author nothing if it is wrong. A claim worth publishing is one that could turn out to be wrong.
- The trends listicle. Five trends everyone has already read about, in an order nobody will remember.
- A sales pitch in disguise. If every argument ends at your product, readers tend to notice, and the paragraphs before it lose their credibility.
- Borrowed opinions. Restating what the leading voices in your field already said makes you an audience, not a leader.
- Generalities from the top. A senior title doesn't make a broad point specific. If you are the executive, go deep on the thing you know best.
- Stopping after three pieces. A reputation is built over many pieces, so a burst followed by six quiet months leaves readers with little to remember.
If you can't find the objection yourself, ask the Strokes Assistant for the strongest one to your thesis, then answer that objection in the piece instead of an easier one.
Frequently asked questions
Is thought leadership the same as content marketing?
No, though they overlap. Content marketing answers the questions buyers search for and leads toward a product. Thought leadership argues a position, uses evidence the author owns, and is useful to readers who never buy anything. A single piece can be both when it argues for a tested method and also happens to explain what the company does.
How long should a thought leadership article be?
As long as the argument needs, within the limits of where it runs. Newspapers and trade journals set their own op-ed lengths, so check the outlet's guidelines before you pitch. For articles published on LinkedIn, LinkedIn's guidance for company Pages suggests planning for 500 to 1,000 words. Essays and reports can run longer if every section adds evidence.
Can thought leadership be ghostwritten?
Yes, and much executive thought leadership is drafted by ghostwriters or communications teams. It stays honest when the ideas, the evidence and the final approval belong to the named author. The writer interviews the author, finds the argument and drafts it in their voice, and the author approves every word and answers for it in public.
Who should write thought leadership?
Anyone with standing on a specific problem: a specialist with data, a manager who tried something new, a founder who changed a practice. A senior title is no substitute for evidence, and a specialist with original data often has more to say than an executive with general views. Start from the problem you know best and the evidence only you have.
