I help technical founders make complex products understandable, trusted, and sold. 25+ years across enterprise tech, B2B2C startups and global GTM + writing
Europe 🏖️ office 💻
Joined January 2022
- Tweets24.8K
- Following600
- Followers1.2K
- Likes67.3K
Pinned Tweet
Funny that AI checks the exact three things I've been beating founders about for a decade now ... with mixed success 😉
But if anyone asks AI (any AI) about your company it will first check 3 things before it says anything nice or vague about you;
1. A clear, specific description of what you actually do. Vague positioning like “innovative solutions for modern businesses” gives an AI model (or your friends or customers) nothing concrete to repeat.
2. Specific claims about who you serve and what problem you solve get cited; generic ones get skipped.
Consistency across every source that mentions you. Your website, LinkedIn, review sites, and directory listings all feed the same model.
When they contradict each other on basics like team size or founding year, it's a problem and the AI engine hedges instead of committing to an answer.
3. Third-party proof, not just your own claims.
Case studies, press mentions, and reviews hosted outside your own domain give an AI engine something more trustworthy to cite than your homepage copy alone.
Just to be clear, this worked before and works now as well, so you better get to it 🙏 please 🙏
The problem is that almost everyone today is building apps. Founders and vibecoders hope they'll find product-market fit after they launch.
- Great UI.
- Clean onboarding.
- Beautiful branding.
But:
- Zero demand.
- Zero distribution.
- Zero validation.
Successful founders always study first:
- What's the problem.
- What's already working.
- What's the way distribution works.
All that before building. Thinking about distro is still underrated.
Fun?
Scary?
Exciting? … which vibe do you feel?
Met some interesting characters outside @Figure_robot HQ yesterday morning…
They didn't have much to say. Luckily, @adcock_brett did!
New Moonshots episode coming soon 🤖👀
33
Split the work by JBTD (Jobs To Be Done) and you’ll save and learn more …
longer article about it in comments 👇
everyone's posting fable 5.2 videos.
nobody's explaining how they made them.
i made this vox-style documentary on the $6.3B binance hack. fully in chat, no after effects
my biggest learnings:
1. make your ai study your references one large frame every quarter second, not skim them. this was a huge unlock.
2. use a project manager chat. one chat turns your thoughts into a brief, then hands the whole job to a fresh fable chat. my random thoughts stopped turning into rebuilds and it saved me so much usage
3. turn your notes into a numbered fix list. it changes only what's on the list and leaves everything else alone
want the prompt i used? comment "heist" (it's free)
1
31
Motivation for European founders — ClickHouse just bought Xata! Romanian founders, Postgres branching that AI agents can spin up and throw away in seconds.
Cool tech that now sits in front of 5,000+ ClickHouse customers overnight. That's distribution Xata would need years to build alone.
Not saying every startup should aim for an exit.
Not saying every startup should aim at becoming someone’s feature.
I guess the question is simpler - who already has your buyers, and why aren't you talking to them yet?
Partner, integrate, pilot together … or wait until they build it themselves.
Who's your "ClickHouse"?
AI selling to SMBs and Enterprise will get way more crowded soon — well, if rumors that Meta is building teams to enter that field are true … 🤔
When do you think?
1
34
When you start your workshop about your creator business with showcase how you burned out on long days of calls in your corporate job or previous business …
… I think you lose half of the audience below 30 because they’ve never experienced anything like that and they possibly never even heard about it. 🤷
1
2
49
Pete Bright (aka BFG) | Lab2094 retweeted
SEO is dead.
Cold outbound is dead.
Organic reach is dead.
Gated PDFs are dead.
Content is dead.
Reddit ads are dead.
What's not dead?
For only $99 you can buy my course to find out:
Pete Bright (aka BFG) | Lab2094 retweeted
Quest update: they’re all printed now, and coming with me to Oslo.
Job’s not done.
@KingOfQueenz @JColeNC
A month ago @JColeNC dropped “The Fall-Off”
On release day I instantly made an artwork about it, felt like something I needed to do.
Little did I know that picture would turn into a quest to make an inspired one for every track.
24/24✅
👇🏻
ChatGPT will soon show you how often it mentions your business … good or bad?
OpenAI is testing a dashboard that shows how often ChatGPT mentions your business and how many clicks it sends. Not live yet.
I think it’s pretty obvious preparation for hard monetization - the attention extraction (2005 Facebook style) time is coming 🤓 but it's here's the thing for technical founders: you'll see the score, not the prompts.
So you'll find out you're invisible, but not why.
Why is usually boring. I talk about it a lot — your site never says clearly who it's for and what it replaces, and a model can't repeat what you never said.
So before buying another AI visibility tool, maybe ask 10 buyers how they'd describe you. If they can't, ChatGPT can't either 😉
1
41
Blatant reminder for all chronically online — no AI prompt or even video call can ever match
the value of in-person deep-dive or intellectual debate with an off-topic chat over coffee.
Such are the facts!
Schedule one ☕ for the upcoming week!
That’s how it’s done people ✅
You repeat your vision,
you repeat your vision,
and you repeat your vision,
because there’s always someone who haven’t heard it or forgot and may like it 💪
Selling is not about pushing your thing, talking about features, nor talking about your successes and victories.
Selling at its best is about listening and offering a solution to your prospect’s problem.
Best sales, BD, GTM people are problem solvers, not sleak talkers. They listen more than they talk.
Sure, you also need to talk about features, successes and yourself — but only within the story of how these solve the problem and when they build trust. Not more, not less.
I regularly write about my experiences with finding the right paths to sell and the right people to sell to -> join BuildBetter newsletter and find something for your business 👇
1
35
BuildBetter is the newsletter for the first time technical founders that know distribution matters as much as a product;
buildbetterhq.substack.com/
6
Who else feels that messegas like:
- “usage limit reached"
- “model unavailable”
are coming faster and more often?
Don’t have sharp stats on it. It’s just a feeling that I’m seeing daily demonstration of compute shortage.
32
You definitely need to spend something — whether it’s money, effort, time ir all of them
Sales don’t just happen
Trust doesn’t just happen
Distribution doesn’t just happen
Brand recognition doesn’t just happen
Marketing lesson you must understand: You need to spend money to make money.
This does not only apply for VC backed companies with huge budgets.
So often founders tell me "We're not spending money cause we have no users."
Brother, the reason you have no users is because you spend no money! You just wasted 2 years building a product that no one knows of.
You need to spend to make.
1
34
What does it mean to be a great product team?
What does it mean to be great at building things in the AI age?
Maybe it means being revenue-driven team rather than product-driven team 🤔
Agree or disagree?
22
Not that I’m solely into cybersecurity but it’s a hot topic these days and also one of a few areas where funding flows continuously even in Europe.
If you’re building in cybersecurity then I believe you need to step aside from virtual GTM motions and actually show people that you’re human that can be trusted — that will earn you a chance to show your product too.
Trust is built much faster f2f - so I’d suggest you turn events like it-sa, Black Hat Europe, etc. into pipeline. Go offlline while everyone is polishing online sequences!
