I’m mostly using this as a blog about startups, product design, marketing, and investing — I’ll keep the emails to a minimum but may send from time to time.

Anyway, wanted to break the seal to share a quick announcement: Applications for Character Labs G6 are due Wednesday, April 8 (two days from now!)

Character Labs is our 5-week sprint program for technical founders. It’s the only startup accelerator that’s 100% focused on product-market fit. In Character Labs, founders work directly with our team, using our proven sprint methods, to accelerate toward product-market fit.

This will be our 7th group (we started with Group 0, naturally) and it just keeps getting better and better.

If you’re building something new and at all interested in working with us, visit character.vc/labs and read all the details there.

Any questions, please reach out!

Something I've been thinking about: how NARRATIVE & EVIDENCE explain almost everything in startup investing

🧠 Narrative is what everyone believes - the popular truths in markets, trends, technology, the future, etc - these beliefs travel via substack, X, linkedin, podcasts, and the coffee shops of SF...

VCs have a bad habit of pushing self-serving narratives (I try not to do this)

Founders have a bad habit of trying to fit the narrative (a couple months ago a lot of startups were suddenly "openclaw for _____")

Narratives can change quickly, or they can stick around a long time. e.g. we are 3-4 years into a period of very powerful AI narratives (which are mostly correct imho)

Meanwhile...

🔬 Evidence is what's really happening - at companies, on teams, in people's lives where they may or may not be paying attention to any of YOUR narratives (plus they have their own narratives)

Evidence does not travel quickly, or with ease - it lives in dashboards and spreadsheets and 1:1 conversations

Evidence is much more durable than narrative (think of the public markets, where the narrative has turned against legacy SaaS while the evidence is that those companies are still making a ton of money)

Narrative drives fundraising, until it doesn't

In time, evidence usually wins - but by then, it may be too late

The relationship between narrative and evidence is very important to understand if you are an investor (or trying to raise money from investors)...

🖖 There are three zones that matter

1. Strong Narrative + Strong Evidence = the definition of consensus (founders will raise easily, but beware taking on too much at too high a price; if you're an investor, you'll have to pay up for quality)

2. Strong Narrative + Weak Evidence = where hype lives (high status founders can raise easily if they say the right incantations; investors have to be very careful or they'll deploy their whole fund in 12 months)

3. Weak Narrative + Strong Evidence = opportunity zone for the patient and courageous (this can be very frustrating for founders, and I get it! otoh we have made some of our best investments here but it feels bad to be non-consensus)

(most of the time, Weak Narrative + Weak Evidence = the death zone. only the financially independent, delusional, or truly visionary/persistent survive here)

My advice to both investors and founders is simple: Know what zone you're in and what game you're playing

☝️ The most important thing: the only way to break the zones is through INSIGHT - the third dimension

If you have real insight (not blind faith) about something, you can build/invest in ways that step outside this framework and create breakthrough opportunity

You might even make it in zone 4

Just keep in mind, the more you leverage insight, the more you have to become a teacher... to customers, to investors, to everyone you work with...

But the reward (as with all teaching!) is the opportunity for impact that goes FAR beyond conventional thinking

• Talking out loud can be easier when you're not sure what you want to say

• You can do it while walking or otherwise away from the pressure and distraction of a "focused" work environment

• LLMs are really good at summarizing and organizing information

• By providing the LLM tons of context (in the form of your brain dump), it helps avoid slop (versus, say, "write me a blog post about X Y Z with these three main points")

My old favorite way to bridge the gap: Jot down little notes here and there in Google Keep on my phone. I still do this a ton (and it still works really well) but there's something freeing about the open-endedness of a long-running voice memo...

Have you tried Vibe Writing like this or with a similar workflow?

There are two kinds of founders: Product Founders and Marketing Founders.

Both perspectives are obviously critical, and the best founders embody both.

But you have to get the sequencing right. In general, it’s better to nail the product first, then market it.

Marketing Founders face a special challenge: They are susceptible to over-invest in marketing in early days, because they are good at it and it feels good.

When I talk to these founders, I offer them the TELEPORT TEST.

Here’s how it works:

Imagine you have a teleportation device. You can reach out into the world, grab an ideal customer, and teleport them right here into this room with you. Here’s the test: Will your product click with this customer? Can you confidently activate, monetize, and retain them?

Now, here’s the prescription: Until you can consistently pass the test, DON’T INVEST IN MARKETING. If you do, you’ll find yourself with the classic leaky bucket, adding more water (customers) only to find that the level (revenue or retention) never rises.

For a Marketing Founder, this is really hard advice to hear.

Marketing Founders are good at marketing, and marketing gets attention, and attention creates dopamine. It’s a powerful cycle. (I feel it every time I make a post here!)

BUT but but! I can see that your hackles are rising.

🙋‍♂️ “But you also tell me to test with customers. How can I find customers to test with if I don’t talk publicly about my company?”

Alright, you got me. You’ll need to use outreach (a form of marketing) to recruit customers for testing. But these efforts should be small and focused, and you should stop after you filled your next panel.

🙋‍♂️ “Distribution is the key risk for my business. I know I can build the product, but reaching customers is the hard part.”

Cool. Run some experiments until you’re confident. But remember, marketing doesn’t solve customer problems — product does.

🙋‍♂️ “What about branding?”

In rare cases where you can’t launch for a year+, then it MIGHT be sensible to invest in building your brand so there’s familiarity, trust, and credibility when you do launch. But for most startups, branding is a later thing.

🙋‍♂️ “What about press?”

Good point. If there’s an external news event (like a fundraise) you can use to your advantage, go for it. But the key words are “to your advantage” — try to delay until you can use the attention for acquisition (after you pass the Teleport Test), hiring, or fundraising.

🙋‍♂️ “But marketing is free JZ, I just post on social. It’s organic!”

No, it’s not. Every calorie spent on marketing takes energy away from getting the product right.

Ultimately, building new products is all about solving important problems for people you care about. Until you’re confident you can do it consistently, I think you owe it to yourself (and your customers) to stay focused.

Use the Teleport Test. Build first, market later.

Not an a/b test, but completely different approaches to positioning, differentiation, product architecture, workflow, etc.

This works REALLY WELL because it snaps people’s brains into “Shopping Mode.” I don’t know why, but Shopping Mode is powerful—we all “shop” all the time, and we’re much better at comparing options than we are at reacting yes/no to single option.

A few teams in Character Labs are doing this right now. It’s shockingly easy to do, especially if you are building with AI.

Here’s how it works, high level:

1. Use sketching or other solo work to generate different concepts. DON’T do this as a group discussion, or your multiple concepts will converge to one blah compromise.

2. Pick out the ones you want to test by identifying the specific hypothesis for why you think it’ll work. E.g. “if we make the homepage super simple and get people into using the product ASAP, they will be more likely to understand it, like it, and convert”.

3. Use whatever prototyping tools you like best to create these as separate apps. Here’s a trick: If you are fluent in multiple tools, use a different tool for each concept—this will keep them from looking like flavors of the same thing.

4. THIS IS KEY: Give each approach a different name and visual style. In other to activate Shopping Mode for your customers, they have to believe they are looking at multiple different products. If you show them versions A, B, and C of the same product, they will slip into Feedback Mode and try to give you advice. This is 90% less useful than their raw reactions as they shop.

5. Interview your customers, and during the interview, ask them to look at a couple of sites/apps/products/whatever with you. Don’t tell them these are different flavors of the same thing. If you did it right, they won’t notice—they’ll think they are just “shopping” for a solution.

(You can’t do this as an a/b test, because: a. it’s too slow to fix all the bugs and ship production-ready code, and b. you want to be there when your customer is shopping)

Congratulations, you just learned more in 5 days than most founders do in 5 months :)

The team at Inductive Bio are incredible at this, which is especially important in the biotech/pharma world, where trust is hard to build and lead times are long.

Josh dropped a ton of gems during the call yesterday, but one phrase stood out:

“We just want to solve this for you.”

This was Josh and Ben’s mantra in the early days. They weren’t precious about product. They weren’t afraid of consulting. They did lots of things that didn’t scale.

It’s so hard for product builders to adopt this obsession with solving customers’ problems.

Usually they got into building products because they... love building products. And to be clear—long-term, you need to be great at building products if you want to succeed.

But in the short-term, it can be a handicap. If you are always thinking “how do I build the product to solve this customer’s problem?” you’ll sometimes miss out on opportunities to just... solve the customer’s problem.

So, here’s my advice (with thanks to Josh!)...

If you’re pre PMF, adopt this as your mantra: “We just want to solve this for you.”

Become obsessed with solving problems for your customers, and I promise that good things will happen.

“Rational” is an invention — it’s a mode we can switch into, but it’s not where we spend most of our time.

By framing “irrational” as a negative, people think of it as an exception; a problem to be fixed.

In reality, it’s normal. We should be clear about it, and design our systems and our products to acknowledge that.

Like Dan Ariely so famously said: We are predictably irrational.

A cool 1500 people signed up.

We didn't talk about how to vibe design or vibe code (there are already lots of how-tos available for that).

We didn't talk about the most hyped tools you have to be using right now (we're not cool enough to be up-to-the-minute).

Instead, we opened the book on our “missing manual” for building products that cut through the noise and actually solve real problems for customers.

We pulled up an empty Miro board and started building out a framework of context, big ideas, and advice.

We shared our unique approach to product design, based on lessons from working hand-in-hand with 300+ teams.

I was gonna write up the big lessons we shared at the event, but Mark Hubbard did it for me. Pasting those below with a few edits and much gratitude to Mark!

BIG IDEAS

1. Differentiation is Crucial: Successful products solve problems in a radically different way from existing solutions.

2. AI Output is Generic by Nature: AI-generated outputs tend to be generic because they are based on common data. Differentiation requires human insight and creativity.

3. AI Stresses Everyone: The rapid pace of AI development can be overwhelming, but it's important to focus on quality and differentiation.

4. Quality Matters: Obsessing over quality and details is important in an era when more and more teams are putting out slapdash work.

5. Context is Key: If you want to produce differentiated output with AI, you need to provide differentiated context (from customer data, personal insight, etc). In the words of Tyler Cohen, "Context is that which is scarce."

6. AI Can Reproduce Design, But Not the Process of Design, with credit to M.G. Siegler

ADVICE

1. Don't Outsource Your Thinking: Use AI for grunt work but maintain your unique insights and thinking.

2. Start on Paper: Begin with sketches to capture your intuition before using AI tools. If it's slow and hard, that means it's working.

3. Ask Questions Before Choosing Tools: What needs to be true? > What experiment do I need to run? > What prototype do I need? > What's the best tool(s) for ME (not someone else) to use to build that prototype?

4. Slow Down to Save Time: Taking time to think critically can lead to higher quality outcomes. Thinking is slow, but it's how you tap into unique insights and differentiation output.

5. Use AI as a Collaborator: AI should be used to speed up processes and explore ideas, not to replace human creativity.

1️⃣ Wealth Manager

I love personal finance and investing. I'm fascinated by behavioral science. I think I'm pretty good at synthesizing information and creating tools to help people act in their own best interest. Downside? Probably too much optimization; not enough new and different.

2️⃣ Music Producer

I love music, and I used to write/perform/record/produce music in my teens and 20s. It turns out what I do at Character Capital is a lot like production... find great people, back them, support/enable/empower them, but ultimately get out of the way and recognize that they are the geniuses and stars, not me. Downside? Probably even harder to find success than startups.

3️⃣ Yacht Captain

I love sailing and boating, and unlike most people, I actually enjoy managing, maintaining, and (occasionally) repairing boats. I enjoy creating repeatable systems and designing great experiences. I think I'd make a darn good yacht captain. Downside? Being away from home for months at a time.

4️⃣ Graphic/Marketing/Brand Designer

This one is not SO different than what I do in this universe. But the context is different: Clients instead of portfolio companies. Campaigns instead of products. Offline (mostly) instead of online. Fun fact: When I graduated from college, this was the career I pursued. I applied to a bunch of jobs at marketing/design/interactive agencies and got exactly zero of them. On a whim, I applied for a job at a startup. Fortunately, I caught Matt Shobe's attention and... the rest is history.

Some context: I’ve helped 300+ teams make and test prototypes to test business hypotheses...

90% of these follow a similar pattern — a realistic-looking simulation of a frontend that teams use to decide whether it’s worth building a technical or business backend (using “backend” here to mean software, logistics, data, or even a business process that’s not trivial to stand up).

The team at Hatch was in a different situation.

For their new AI-powered donor scoring, they already knew — through previous prototype testing and customer interviews — that customers REALLY wanted comprehensive and reliable scoring of prospective donors.

In other words, they had answered the question “Do our customers want AI donor scoring?”

Now they were prototyping the backend — experimenting with models, prompts, pipelines, and evals.

It was time to answer the question, “Can we reliably produce AI donor scores that are accurate and useful?”

But they needed a quick way to display and interact with those scores. And here’s where that clever prototype came in…

Instead of building a functional frontend or relying on rudimentary plain text views, Ben created a dashboard in Coda — it could pull real scores from the prototype backend and provide visualizations, filters, and sorts with virtually zero design or dev effort.

In the near future, you might write code for the test frontend instead.

Either way, here’s the pattern:

1. Identify the tough question or hypothesis, and focus your efforts on getting that hard part right.

2. Isolate it from the other aspects of the problem space.

3. For the rest (the easy parts), leverage AI or other tools to keep these lower-risk areas low cost.

I’m curious… what are the most clever prototypes you’ve seen recently? How do you think vibe coding and AI tools will change this?

Earlier this year, at one of our Foundation Sprint Workshops, CTGT founder Cyril Gorlla joined us to share his journey — from Character Labs to YC, to a $7M seed round and working with some of the largest companies in the world.

A few things that stood out:

1️⃣ Get Started, Not Perfect

Cyril and Trevor Tuttle were still grad students when they joined us in Character Labs. If they had waited until everything was perfect to make the perfect plan before starting their company at the perfect moment, they might still be waiting.

2️⃣ Take Action to Generate Information

Cyril and Trevor had been focused on research and technical development. They were confident in their technical approach, but they needed to generate information from the market. They did this in Labs by testing prototypes with real customers every week.

3️⃣ Story Matters

CTGT spent most of Character Labs working on their sales deck. Seriously!! Isn't that wild?

Actually, it's not so wild. They had a differentiated technical approach based on unique insights and experiences, but they didn't quite know (yet) how to talk to customers about it.

4️⃣ Tap Into Your Conviction

The team came into the Labs with incredible passion for their solution. It was one of the things that stood out from their application and our interview with them.

And as Cyril shared last week, it's essential to tap into your conviction - it's the heart of your differentiation and the motivation that'll help you keep going when things get tough.

5️⃣ But Make Sure You Test It

Ultimately, conviction doesn't matter if your approach doesn't click with customers. The world's best founders balance deep conviction and enthusiasm for their approach with humble and honest validation that comes from repeatedly and pointedly testing with customers.

These lessons from CTGT’s journey aren’t just inspiring, they’re instructive. They remind us that momentum matters more than perfection, and that great startups are built through continuous learning, not one big bet.

That's what Character Labs is all about.

1. It’s just a hypothesis until you test it

I've worked with ~300 software teams and, a long time ago, I accepted that none of us really know how people will react to our products.

Be clear about your hypothesis and find a way to test it ASAP.

2. Everything is downstream of product-market fit

Hiring, fundraising, press, and technology are all important, but ultimately don't matter if you can't find product-market fit.

(My definition is Rob Kaminski's: “PMF is when you can repeatably find, sell, serve, and retain customers.”)

Optimize everything for accelerating toward PMF. Including hiring — only hire figure-outers, not doers!

3. Markets are made up of people

When you start, focus on making something that clicks with 1 person, then 5, then 50, then 500...

Before you can find product-market fit, you have you find "product-person fit."

4. If you can’t find 5 customers to test with (or 50), you might be in the wrong business

Founders often ask, "How can I find customers to test with?" Or, "Can you introduce me to potential customers?"

Finding customers is the whole point! Customer intros are like training wheels; the goal is not to get better training wheels, but to get the wheels off ASAP so you can ride on your own.

Like Fan (J) Li, I strongly believe "you can network yourself to 4-6 of any type of user.”

And if you can't, it might be a bad market, a bad business, a poorly defined target, or another red flag.

5. Focus on one risk at a time

Building a startup is like opening the invitation box at the beginning of Glass Onion — you have to solve one puzzle before you can move on to the next.

Constantly ask yourself: "What's the next most important thing I need to figure out?"

If you try to do it all at once, you'll probably fail (and drive yourself crazy doing it!)

6. Maintain your freedom of movement

The best founders take action (talk to customers, test prototypes, launch stuff) to generate information about what's working or not working, and then use that information to inform decisions and change course if necessary.

But when you make a big decision that limits your freedom of movement — like a corporate partnership, an oversized bet, etc — that breaks this loop, and you stop learning.

7. Intuition is your most important muscle — strengthen it always!

You'll never have perfect information, so every decision depends on your intuition. THIS IS OKAY.

In fact, as a founder, your ability to make high-context, intuition-rich decisions is one of your advantages over incumbents.

Try to spend 80%+ of your time on activities that build your intuition — each one will make you better, stronger, and more likely to succeed.

p.s. If you’re a founder thinking through product-market fit, customer testing, or go-to-market risks, we work on these questions every day at Character Labs. Check out character.vc/labs to learn more or stay in the loop for the next program.

• We created sprint methods (including the Design Sprint and Foundation Sprint) that are used by the world's most successful startups

• We make time to personally run these sprints with you, to help you reach product-market fit faster

• We will treat you with character (pun intended!), whether we invest or not

If you're in the early days of building something new, we'd sincerely love to talk. Here's a bit more about what we do and how we do it...

With Fund 2, we remain focused on world-class technical founders using AI to solve previously unsolvable problems with software.

Our first investment in Fund 1 was Phaidra, whose AI control systems improve reliability and efficiency for datacenters and other industrial facilities. Phaidra has raised more than $60 million from Index, Collab Fund, and others, and is now working with NVIDIA, AWS, and other datacenter operators.

We have repeatedly seen the incredible potential of applying new AI capabilities to unsolved problems. Our portfolio includes Fathom, Inductive Bio, Extend, Reclaim, Orbital Materials, Weave, and other companies who have grown into significant businesses (and raised significant follow-on financing) since our initial investments.

But capital is only half our mission. New technologies don’t matter unless they work for customers. And that’s where the other part of our mission comes in...

Over the past 20+ years, we have helped 300+ teams design new products and bring those to market.

We do this with our sprint methods — the Design Sprint, the Brand Sprint and Name Sprint, and most recently, the Foundation Sprint, with more on the way! — working collaboratively with founders to rapidly prototype and test products, simplify complex decisions, and validate hypotheses.

These methods have become exceptionally well-known among founders and tech operators through our books Sprint and Click. We continue sharing our methods through case studies, resources, templates, and tools, and via presentations at amazing companies like OpenAI, Anthropic, DeepMind, Google, Notion, Stripe, and more.

From day one of Character, we designed our firm so the three of us — the founders and general partners — can personally spend time leading sprints with our portfolio companies.

Our support goes way beyond office hours, board meetings, intros, and advice. We actually clear our calendars and work hand-in-hand with founders, leading sprints, sketching, writing, prototyping, and applying personal insights and lessons to their toughest problems.

We plan on doing this a very long time. Armed with a new $52 million fund, we're excited to continue our work with a few key improvements:

1. Character Labs is now our primary starting point for new investments. We run the program twice a year and accept up to 15 companies per group.

2. Our maximum check size is now $2 million, allowing us to lead and co-lead a select number of seed rounds.

3. As announced last year, Jake Knapp is now a general partner. Jake helped Eli and I start Character back in 2021, and his contributions are immeasurable.

4. Kristen Brillantes is now working with us as an advisor and part-time head of operations. Her capabilities, taste, and community have become integral to Character. We originally worked with Kristen 10 years ago at GV, and we're thrilled to be collaborating once again.

We love working with founders, and we're grateful to continue with even more resources! Thanks all!!

(You can read more of our story in Alex Konrad’s recent profile of Character.)

Here’s the idea…

When starting a company, the best founders engage in a tight loop of:

1. Taking some action in the market (talk to customers, test prototypes, launch stuff), to...

2. Generate information about what's working or not working, and then...

3. Use that information to inform decisions and change course if necessary. Then back to step 1 👆

But you can easily and unintentionally break this loop.

All too often, I see founders make a big decision (a corporate partnership, a round of financing, a fixed roadmap, an oversized bet, etc) that limits their freedom of movement and breaks the loop between steps 2 and 3.

In other words, you get new information, but you can't do anything about it, because you're locked into the big decision.

Before you know it, 3 months have passed... 6 months... and you're no longer building a startup, you're executing a plan. And the plan might be (probably is?) wrong.

These freedom-of-movement-limiting decisions can be tough to avoid because they often look like exciting opportunities.

My suggestion? Look at every big decision through the "freedom of movement" lens. Ask yourself, will this increase or decrease my freedom movement?

If it's limiting, proceed with caution.

When founders are building new products, the most important steps are:

• To clearly define the opinionated hypothesis behind the product

• To test the components of that hypothesis, AQAP

Some founders do this intuitively.

In fact, I'd argue that most successful founders have developed this intuition to a great extent.

A few succeed despite not doing steps 1 and 2.

But most founders don't clearly identify and test their hypotheses, and as a result, they fail quickly or slowly.

I've personally seen this play out from working with 300+ teams over the past 20 years. In 90% of cases, looking back, it's clear whether they did steps 1 and 2 — and you can see how that affected their odds.

At Character Capital, we look for founders who already operate in this way. We call it "hypothesis-driven," and it's one of the criteria we score when evaluating potential investments.

But like a coach working with elite players, we also want to help the best in the world get better.

And if you're not working this way now, we've created a few methods you can use to start.

The Foundation Sprint and Design Sprint are our methods for helping founders build products that click with customers.

And we run these sprints almost every week with portfolio companies.

Want to see how we do it? Next month we're hosting a 2-day workshop in San Francisco where you can deep dive on the methods and run a full Foundation Sprint with our team.

We are super excited to share it with you all, and this episode of Jake & JZ is a deep dive into the process of writing the book and how it fits into our investing strategy.

We also talked about:

• How to get early access to Click

• The history of the Foundation Sprint

• Books that inspired us: Psychology of Money, Power of Habit, Building a Storybrand, The Heath Brothers’ books

• Why we had to share the Foundation Sprint