11 Thought Leadership Examples That Actually Worked

The short answer

The thought leadership examples that hold up share one trait: each told its audience something true that the audience could not get elsewhere, and kept doing it for years. Eleven worth studying, grouped by what they teach:

• Original research: DORA, Stack Overflow Developer Survey
• Product data: Okta Businesses at Work, Gong Labs
• A framework and a cadence: Stratechery, Paul Graham's essays, Amazon's shareholder letters
• Open playbooks: Netflix culture memo, Basecamp's Shape Up, the GitLab Handbook
• Expert-led editorial: First Round Review

What makes a piece of thought leadership actually work?

Thought leadership works when it changes how a specific audience understands a problem, and when that audience can trace the change back to a source it trusts. Reach and volume do not decide it. A report nobody cites, or a newsletter that restates the news, has not led anyone's thinking.

Four traits separate the examples below from the much larger pile of content that calls itself thought leadership:

1. Evidence nobody else has. Proprietary data, a survey run with rigor, or first-hand experience written in detail.
2. A claim stated plainly. The piece says what the author believes, not that "the issue is complex."
3. A rhythm kept for years. The same report, letter or newsletter returns on schedule, so each piece builds on the last.
4. A format the audience can reuse. Numbers to cite, a framework to apply, a playbook to copy.

How the list was built: every example has a public primary page (the report, the newsletter or the handbook itself), and no outcome is claimed beyond what the publisher states. Where a source does not publish results, the entry says what was published and why the design works, not how much revenue it produced.

11 thought leadership examples worth studying

1. DORA: research that became a shared vocabulary

DORA is a research program run by Google Cloud that studies how software teams work and how that work connects to delivery performance and organizational outcomes. It publishes an annual DORA Report and describes itself as "the longest running academically rigorous research investigation of its kind."

Why it works: DORA did not stop at findings. It turned them into a small set of named software delivery metrics (on its current guide: change lead time, deployment frequency, failed deployment recovery time, change fail rate and deployment rework rate) that teams can measure on their own systems. It also publishes a Core model that, in its own words, "deliberately trails the research," so practitioners get only the most established findings.

The result is visible in other companies' products: GitLab's documentation, for example, includes a page on tracking DORA metrics.

What to borrow: give your findings names and a way to self-assess. DORA's site offers a Quick Check that lets a team compare itself with the research. A finding readers can apply to their own numbers travels further than a finding they can only quote.

2. Stack Overflow Developer Survey: community data, published openly

The Stack Overflow Developer Survey asks developers, every year, "about the tools they use, how they learn, and how they work," and Stack Overflow presents the results as "the largest survey of people who code." The results page keeps every past edition, going back more than a decade.

Why it works: the survey owns a recurring question that its audience cares about, and it answers it at a scale no single team could reach. The decisive detail is openness. The raw data is downloadable and licensed under the Open Database License, with past years kept in a public GitHub repository, so analysts, journalists and researchers can build on it.

What to borrow: if you run a survey, publish the data, not only the charts. Every third-party analysis of your dataset carries your name into places your marketing never reaches.

3. Okta Businesses at Work: what only your product can see

Okta's Businesses at Work report looks at how organizations use applications and services, based on what Okta describes as "anonymized data from our network of thousands of customer organizations (both big and small), applications, custom integrations, and millions of daily authentications." Okta has published it every year for more than a decade.

Why it works: nobody else can publish this data. An identity provider sees which apps companies actually connect, so its rankings of the most popular and fastest-growing apps are evidence, not opinion. Okta also turns sections of the report into standalone pages that readers can link directly.

Okta notes that the findings reflect its own customer base rather than the whole market. That caveat strengthens the report: readers know exactly what the numbers describe.

What to borrow: ask what your product sees that no survey can. Then say clearly whose behavior the data represents.

4. Gong Labs: one question, one data-backed answer

Gong Labs is where Gong, a revenue intelligence platform, publishes research from its data team. Gong describes it as "our original research engine," built on "billions of sales conversations, deals, and workflows."

Why it works: the format is small and frequent. Instead of one large annual report, Gong Labs publishes short pieces that each answer a question salespeople already ask. Recent titles include "Do execs really reply to cold email? Here's what the data says" and "When (and how) to multi-thread when selling to executives."

Each headline promises an answer, and the answer comes from data the reader cannot collect alone. For B2B thought leadership aimed at practitioners, that combination is hard to beat: the content is useful the same day, and it reminds the reader where the data came from.

What to borrow: list the ten questions your buyers ask most often, then check which ones your own data can answer.

5. Stratechery: a framework built one article at a time

Stratechery is Ben Thompson's newsletter on the strategy and business side of technology and media. Thompson has written it for more than a decade. Weekly Articles are free, and three Daily Updates a week are for paying subscribers; Stratechery says it has subscribers in more than 85 countries.

Why it works: Thompson built a recognizable framework, Aggregation Theory, and applied it to company after company. Stratechery has a dedicated page that collects the articles behind the theory, and states that the free Weekly Articles "are the foundation of the subscription-only Daily Update, consulting, and speaking engagements."

The second reason is independence. Stratechery's ethics statement says Thompson does not hold individual stocks in companies he writes about and does not take speaking engagements with companies he covers.

What to borrow: name the lens you use and apply it in public, repeatedly. A framework readers can apply without you is what turns a writer into a reference.

6. Paul Graham's essays: strong claims in plain language

Paul Graham, a cofounder of Y Combinator, publishes essays on startups, work and writing as plain web pages on paulgraham.com, with a short list of suggested starting points at the top of the index.

Why it works: the essays follow the standard Graham describes in his essay on how to write usefully: useful writing "tells people something true and important that they didn't already know, and tells them as unequivocally as possible." Each essay commits to a claim. There are no images, no gated downloads, no calls to action.

The investment per piece is high. In another essay, Graham writes that he usually spends weeks on one.

What to borrow: one claim per piece, argued as strongly as the evidence allows and no further. Fewer pieces with real conviction beat a steady stream of hedged ones.

7. Amazon's shareholder letters: a position repeated every year

Amazon publishes a letter to shareholders every year, and one detail explains much of its authority: each new letter ends with a copy of the first one. Andy Jassy's letter, for example, closes with a postscript saying that, as Amazon has always done, the original shareholder letter follows, and then reprints the company's first letter in full.

Why it works: the first letter took a clear position, under the heading "It's All About the Long Term," and argued that Amazon would be judged by the value it created over the long term rather than by short-term results. Reprinting it every year turns one document into a standing commitment that each new letter is measured against.

What to borrow: write down the few beliefs that will not change, and show readers, year after year, how your actions still match them. Consistency over time is a form of evidence.

8. The Netflix culture memo: a playbook anyone can read

Netflix publishes its culture memo on its jobs site as a long web page, describing a culture "based on four core principles": the Dream Team, People Over Process, Uncomfortably Exciting, and Great and Always Better. The page itself refers back to the company's first culture deck, the slide presentation the memo grew out of.

Why it works: the memo is specific enough to be uncomfortable. It describes the "keeper test," in which managers ask "if X wanted to leave, would I fight to keep them?" and it explains the reasoning instead of listing values. Readers can disagree with it, which is exactly why they discuss it.

What to borrow: publish the operating principle, including the parts that will put some people off. A culture or method document that could belong to any company teaches nothing.

9. Basecamp's Shape Up: an open-sourced method

Shape Up is a book by Ryan Singer that describes how Basecamp plans and builds product work. It is free to read online, with a print edition for sale.

Why it works: Shape Up gives teams a complete, named method rather than general advice. Work happens in six-week cycles, which the book explains as "long enough to finish something meaningful and short enough to feel the deadline from the beginning." Projects get an "appetite," the time the team wants to spend, instead of an estimate: fixed time, variable scope. A "betting table" decides which pitches get the next cycle.

Every team that adopts the vocabulary spreads it, and every mention points back to Basecamp.

What to borrow: if your company works in an unusual way that works, write it down in enough detail that another team could run it. The free version is the marketing.

10. The GitLab Handbook: radical transparency as content

The GitLab Handbook is, in GitLab's words, "the central repository for how we run the company." Printed, it runs to more than 2,000 pages, and as part of GitLab's value of transparency it is "open to the world," with an invitation to suggest changes through a merge request.

Why it works: the handbook is not written as marketing. It is the actual operating manual, from engineering processes to sales methods, which is why it is credible. GitLab also publishes its handbook-first approach to remote work, where decisions are documented before they are announced.

What to borrow: some of your most authoritative content may already exist inside the company as process documentation. Publishing even one well-maintained section shows how you think more convincingly than an essay about how you think.

11. First Round Review: experts speaking directly

First Round Review is the editorial publication of venture firm First Round Capital, launched more than a decade ago. Its about page states that the knowledge that can transform how people build technology is "trapped in other people's heads," and promises to "let experts speak directly" and to give tactics readers can use the same day.

Why it works: the Review puts practitioners at the center. It chooses not to use bylines, so the expert's ideas, not the writer, carry each article, and the pieces go deep on how a specific operator solved a specific problem.

The firm benefits indirectly: founders who learn from the Review know where it came from.

What to borrow: if your own team lacks the expertise, interview the people who have it, and edit their knowledge into something a reader can act on.

What do these thought leadership examples have in common?

The examples share a structure more than a style. Read side by side, five patterns repeat:

• The evidence is exclusive. Product data (Okta, Gong), a survey at scale (DORA, Stack Overflow), or an internal method nobody else has documented (Basecamp, GitLab, Netflix).
• The claim is explicit. Graham's essays, Amazon's long-term argument and Netflix's keeper test all take a position a reader could argue with.
• The rhythm is fixed. Annual reports, weekly articles, yearly letters. Authority accumulates because each piece arrives when readers expect it.
• The output is reusable. Metrics to adopt, datasets to download, methods to copy, frameworks to apply. Readers do the distribution.
• Promotion is absent from the piece itself. None of these examples sells inside the content. The company benefits because the reader remembers who taught them.

The last point matters most for B2B teams. Buying cycles are long, and the pieces that work are the ones a prospect saves, forwards or cites months before a sales conversation.

What makes AI-generated content worth publishing?

AI-generated content is worth publishing when it meets the same bar as the examples above: exclusive evidence, a plain claim and something reusable. The model can help with the third; it cannot supply the first two.

None of the eleven examples could have come from a prompt. DORA's findings come from research, Okta's from customer data, Shape Up from years of practice. A language model asked to "write a thought leadership post about sales" can only reproduce the consensus it was trained on.

An AI-assisted piece earns its place when:

• The evidence came first. The draft starts from sources, data or notes the author chose, not from a topic.
• The claim is the author's. A person decides what the piece argues; the model helps say it clearly.
• Every claim is traceable. Each number and assertion can be followed back to its source in seconds.
• The author would sign it without edits to the argument. If the reasoning would change under questioning, it is not ready.

Some tools are built around that order. Kindal, for example, reads the sources a user chooses, writes a brief when something changes, and drafts posts written from your briefs, with each draft showing the brief lines it came from. Whatever the tool, the test is the same one Graham sets for essays: does the reader learn something true and important they did not already know?

How can a small team apply these patterns?

A small team can apply these patterns by picking one evidence source, one format and one rhythm, then holding all three for a year. Large research budgets explain the scale of DORA or Okta, not their logic.

A practical starting sequence:

1. Inventory your exclusive evidence. List what your product, customers, support queue or sales calls reveal that outsiders cannot see. Anonymized patterns from a few hundred customers can be enough for a credible finding.
2. Pick the format that fits the evidence. Data suits short Gong-style posts that each answer one question. Experience suits a playbook like Shape Up. Interpretation suits a newsletter or essay series.
3. Name your recurring lens. One idea you apply again and again, the way Stratechery applies Aggregation Theory, gives readers a reason to come back.
4. Commit to a rhythm you can keep. A monthly data post or a quarterly report sustained for a year beats a weekly cadence abandoned after six weeks.
5. Make every piece reusable. Add a metric, a checklist, a template or the dataset itself.
6. Watch your market for fresh evidence. Competitor moves, filings and specialist sources supply angles between your own data releases; a short list of tools that monitor your market can cover this without an analyst.

Start smaller than feels ambitious. The examples above became reference points because they were still publishing years later, not because their first edition was large.