Pick One Thesis or Kill Your Personal Brand

When a Forbes contributor, a CoinDesk editor, or a Blockworks producer is deciding who to quote on a story, they scan their source list for the person they file under a single topic. Not a "blockchain and AI and regulatory policy expert." One thing. One label. One go-to.

The founders who get quoted at 9pm on a deadline are the ones who own a specific corner so clearly that a journalist can retrieve the name in thirty seconds. The ones who never break through are usually trying to be authoritative on three or four things simultaneously, and in doing so, they're authoritative on none.

I've led founder profiling engagements for AI and Web3 companies, from placing GAIA's Matt Wright in Forbes on decentralized AI agents to positioning Fluence co-founder Tom Trowbridge as the DePIN voice for CoinDesk and Cointelegraph. The single thing that determines whether those placements compound or stall is whether the founder entered the engagement with one defensible thesis or a list of topics they'd like to "cover."

This post explains the three-step process I run at the start of every founder profiling engagement, and why the narrowing is the whole point.

Why three theses is actually zero theses

A founder shows up to our first call with a list that looks like this: DeFi liquidity design, AI agents in financial services, SEC regulatory posture on crypto securities. All legitimate. All genuinely interesting. All topics they could speak to with real authority.

The problem isn't knowledge. It's architecture.

A journalist covering a breaking story on on-chain liquidity fragmentation needs a source in under an hour. They don't search for "blockchain expert." They think of the two or three people they've quoted before on that specific topic and send a message. If you're the DeFi liquidity person AND the AI agents person AND the regulatory policy person, you're none of those people in their mental filing system.

I said this in a different context back in 2022 but it applies directly here:

your business needs to have a personality & your personality is built when you deliver your promises. Shilika Jain on X, 26 Oct 2022

A thesis is the personality of your public presence. When you dilute it across three topics, you're breaking the promise before anyone has the chance to hold you to it.

Step 1: The Topic Audit

The audit starts with a simple question: what do you actually write, post, and talk about publicly right now? Not what you want to be known for. What you're putting out.

Pull the last 60 days of content: X posts, LinkedIn articles, podcast appearances, conference talks, internal memos you've shared externally. Tag each piece with a topic. You're looking for clusters.

Most founders discover that they have two or three genuine clusters, and one or two things they posted once because the news cycle prompted it. The news-cycle posts don't count. What counts is what you return to without prompting.

Then ask yourself three questions:

1. What is broken or misunderstood in your corner of Web3 or AI right now, from your specific vantage point?
2. What will be true in 18 to 36 months that most people are not yet pricing in?
3. What have you personally learned from building that contradicts the conventional wisdom in your space?

You're looking for the intersection of what you actually know, what's wrong in the current conversation, and what you're positioned to be right about in the long run. That intersection is where theses live.

The audit will probably surface three to four candidate theses. That's fine. You don't pick all of them.

Step 2: The Competitive White Space Scan

Once you have candidates, you scan the coverage landscape. Who is currently owning each of those topics? How crowded is the source list?

For each candidate thesis, I look at six months of coverage in the relevant outlets. Who gets quoted on this topic in CoinDesk, The Block, Decrypt, Blockworks, and the relevant AI outlets? How many named sources are there? How diverse is the source pool?

A crowded topic isn't necessarily disqualifying, but it changes the angle you need. If there are already eight well-sourced DePIN experts getting quoted regularly, you can't just join the queue and call yourself a DePIN expert. You either need a sharper angle inside DePIN (compute markets, not storage, not connectivity) or you need a genuinely different topic.

A white space is a topic where there are fewer than three consistently quoted sources in tier-1 outlets, clear reader demand (journalists are covering it, just not with great sourcing), and a window of six to eighteen months before the field gets crowded.

When I worked with Tom Trowbridge at Fluence, the DePIN white space was still wide open in early 2024. The story angle that cracked it wasn't "here's what DePIN is" (educational content with no named source hook) but "revenue is the metric that separates real DePIN from marketing DePIN" (a concrete, arguable position). That distinction matters. A thesis is an argument, not a definition.

Step 3: The Five-Year Defensibility Test

A thesis that earns you coverage velocity over 24 months has to pass one filter: can you still be the credible expert on this in five years?

This filters out two categories of thesis that look attractive but collapse fast.

The hype-cycle thesis. "NFTs are the future of brand loyalty" was a perfectly reasonable thesis in late 2021. By mid-2023, it was embarrassing to have on your public record. If the thesis depends on a speculative market moment rather than a structural shift, it doesn't pass the test.

The borrowed thesis. Some founders want to own a topic because a more famous founder already owns it and the topic is getting coverage. If your thesis is a variation of what a better-known expert is already saying, you're not a source, you're an echo. Journalists don't quote echoes.

A defensible thesis has to be connected to something you are genuinely building or have genuinely done. For the GAIA team, the thesis was decentralized AI agents with node count as proof. Not AI agents broadly. Not decentralization broadly. The specific intersection, backed by verifiable data that compounded every time node count grew. That's why the Forbes placement in December 2024 wasn't a one-off. The Decrypt piece in January 2025, the Binance Square piece in April 2025, the multiple podcast appearances, all of them returned to the same thesis because the thesis kept generating new proof.

If your thesis can survive being wrong about the specific hype cycle and can be validated by the work you're actually doing for the next several years, it passes.

What the selection actually looks like

To make this concrete: a founder comes to me running a cross-chain interoperability protocol. They want to talk about DeFi, AI agents, and regulatory clarity. Here's how I'd walk through it.

DeFi is crowded. Unless they have a genuinely specific angle inside DeFi (not "interoperability enables better DeFi" but something structural about where liquidity flows between chains that other people aren't saying), I'd shelve it. AI agents is adjacent to their actual work but not central, and the space is filling up fast. Regulatory clarity is evergreen but brutally competitive for sourcing because every lawyer and every exchange executive has an opinion.

Cross-chain interoperability security, specifically the failure modes of bridge architecture at scale, is a white space. It's connected to what they're building. It can survive a bear market because security failures happen in every cycle. And there are fewer than four consistently quoted sources on it in tier-1 outlets.

That's the thesis: the person who explains why bridges keep failing and what the architecture needs to look like instead. Not because it's the most exciting topic on the list. Because it's the one they can actually own, defend over five years, and generate compounding coverage from.

The coverage velocity argument for narrowing

Every founder who has pushed back on this process has the same concern: "If I narrow to one thing, I'll miss opportunities when other topics are in the news."

This is backwards. The narrower your public thesis, the faster you accumulate source credibility in that specific topic, and the more journalists trust you to give them usable quotes. A journalist who has quoted you twice on bridge security is far more likely to call you when the next bridge exploit happens than to call someone who sometimes talks about bridges among nine other things.

The narrowing accelerates, it doesn't restrict. Once you're the recognized expert on one topic, getting quoted on adjacent topics becomes easier because the journalist already trusts your analytical rigor.

I've seen this play out across engagements. A steady cadence of founder appearances around one narrative compounds into tier-1 pickup. The founders who try to cast wide end up in mid-tier outlets doing generalist interviews. The founders who commit to one defensible corner end up being called by Forbes.

In web3 all that matters is online reputation! Shilika Jain on X, 2 Nov 2022

And online reputation in PR terms means journalists can name-search you and find a coherent, consistent record on one topic. Not a scatter plot of opinions across a dozen subjects.

The two founder profiles where this applies differently

Pre-seed or seed stage: Your thesis is more about establishing the intellectual framework for the problem you're solving than about your personal background. At this stage, you're building the concept. The thesis should map directly to the category insight that made you start the company. A founder building a DePIN compute network should own the insight about why centralized cloud providers fail for AI workloads, not a vague "decentralization is the future" take.

Series A and beyond: At this stage, you have real data. Node count. Transaction volume. Customer case studies. Revenue. Your thesis should be verifiable by the work, not just arguable from first principles. The strongest founder profiles at this stage are the ones where the journalist can ask "can you back that up" and the answer is a number, a customer, or a deployment.

The thesis selection process is the same in both cases. What changes is the evidence available to defend it.

The practical output of a profiling engagement

When I run the audit-scan-defensibility process with a founder, the output is a single thesis statement of one to two sentences, a set of three to five sub-arguments that support it, a list of the outlet-topic combinations where it's white space, and a content calendar that starts generating proof points immediately.

That's the input to every pitch, every op-ed brief, every podcast appearance brief, and every conference speaker application. Everything is downstream of the one sentence.

If you can't write your thesis in one sentence, you haven't picked one yet.

---

If you're building in Web3 or AI and you're not sure which thesis to own, or you've been posting for months without landing tier-1 coverage, that's usually where I find the problem. Book a 30-min teardown with Shilika and we'll run the audit together.