In 2018, Outreach & Salesloft raised hundreds of millions to own cold email. Lemlist launched with $1k on dynamic images. Now Heyreach (multi-sender) and Gojiberry (intent) run the same playbook. Categories are never saturated. Incumbents just get slow and leave wedges open.

i listed how 20 companies got their first foothold. 14 are at 10m+ in revenue, the fastest in 2 years gamma, recall[.]ai, riverside, heygen, instantly and a few others i wanted to know what the first wedge actually was, before any growth strategy i tagged each company by the type of wedge. a company can have two. 10 types came out of it, 28 tags in total:  1.⁠ ⁠new ai tech: 8  2.⁠ ⁠simplicity: 5  3.⁠ ⁠technical quality: 4  4.⁠ ⁠price: 2  5.⁠ ⁠specialization: 2  6.⁠ ⁠missing features: 2  7.⁠ ⁠targeting by buying signal: 2  8.⁠ ⁠targeting the person, not the company: 1  9.⁠ ⁠speed: 1 10.⁠ ⁠personalization: 1 the 3 that come back most:  1.⁠ ⁠new ai tech gamma. formatting slides took hours, so the product did it for you.  2.⁠ ⁠simplicity recall[.]ai. every software vendor was building its own meeting bot, a heavy job. they sold it as one api.  3.⁠ ⁠technical quality riverside. zoom and google meet compressed the stream, riverside recorded in high resolution. which of the 3 would you bet on if you were starting today?

for 15 years every product team got the same advice talk to your users before you build it made sense when building took months and burned real money now you can ship in a weekend i'm running a full discovery for my own project right now and i keep asking myself should i just build instead? where i've landed: interviews tell you what people say shipping tells you what people do when building gets cheap, building IS the discovery you still talk to users you just do it with something live in front of them

i read AI startup landing pages 10 times before i understand what they do and that's the paradox of this wave building software has never been easier being understood has never been harder when anyone can ship in a weekend, the companies that win are the ones people get in 5 seconds clarity is the new moat

everyone on here says the same thing: build software for unsexy markets. two problems with that advice. first, i rarely see concrete examples. which unsexy market, which company, what outcome? second, what i see on the ground in europe: companies in those markets are curious about ai. but curious is not ready. adopting ai-native tools means changing how people actually work. change management is brutal. most of these companies are waiting for one thing: concrete roi proof from big players before they move. so here's my honest question: are "unsexy" markets actually ready to buy ai-native tools, or are we building for buyers who don't exist yet? curious if anyone here sells into these markets. what's the real adoption story?

the best playbook for building a company used to be simple: raise millions, sell a 10 year vision, get everyone on board. investors, teams, customers. that playbook is dying. with AI moving this fast, who can honestly define a 10year vision? markets shift monthly code costs nothing barriers to entry are at historic lows what i'm seeing instead: small squads of very strong people launching companies on 2-3 year horizons. founders treating each company almost like a feature: bootstrap it, make it profitable, take dividends, reinvest in the next one. if the first product doesn't take, the same squad ships the next one fast. a one squad company. less glory, more cash. less narrative, more options. in a market nobody can predict, staying small, profitable and fast might be the least risky way to build.

200 guests into my podcast, one rule i've never broken: i never accept guests placed by agencies. it's never ended well. one time an agency pitched their VP product. 5 minutes into the intro call, i could tell she didn't want to be there. stressed, unprepared and clearly pushed into it. that episode would have been terrible. for her and for my listeners. here's what most companies get wrong about founder-led content: it's not about putting employees on podcasts for visibility. it's a culture of giving value before you ever expect to receive it. companies pushing their people onto podcasts just for exposure have it backwards. listeners feel it instantly. my guest criteria, after 200 episodes: 1. people who share value easily and say interesting things 2. people with real opinions, who dare to say them out loud 3. people who ask questions and aren't afraid to be wrong in public that's it. no PR filter. no agency pitch. the best thought leadership doesn't look like a strategy. it looks like generosity.

$40m ARR that's what some of the simplest software tools are making right now. linkedin post generators x scheduling tools nothing revolutionary meanwhile my feed is full of "software is dead". so who's right? my take: x is a bubble. a lot of smart people here are hypothesizing 2 years ahead of the market. their takes are useful for building your vision. but they're terrible for deciding what to build today. because users right now don't know every tool on the market. they still have painful problems. and simple products with great distribution still win. i don't know what software looks like in 2 years. maybe AI eats half of it. but if you wait for certainty, you never launch anything. talk to users. solve today's problems. distribute like crazy. that's the only playbook that works in every scenario.

Yesterday a founder told me his company was struggling. I told him to check 2 things. 1/ Does your product fall in one of these 3 buckets? • It makes your customer money Ex: lemlist. Outbound that shows clear revenue impact. That's why so many of these tools work. • It saves your customer money Ex: Ramp. Their literal promise is "save money". • It serves an obligation Ex: Pennylane riding the e-invoicing law. Deel on hiring abroad in compliance. If you're in none of them, you end up in the "nice to have" category. You might sell a bit, but product market fit will be brutal. Even more right now: AI is moving every market fast and nice to have products are the first to get cut. 2/ Is your market big enough to live on? You can start super narrow and vertical. But in B2B you need at least 10,000 businesses you can serve mid-term. No bucket, no market, no PMF.

One B2B interview can feed your content for 6 months. Most podcasters use it for 6 days. I've recorded 400+ podcast episodes (600k+ listens). One interview can become: •⁠ ⁠compilations with past episodes on the same theme •⁠ ⁠posts around each strong take from the guest •⁠ ⁠visuals on different angles posted on X & linkedin •⁠ ⁠a full post on one spicy opinion •⁠ ⁠a re-run 6 months later for everyone who missed it And if your main format is YouTube, it recycles into a podcast, a newsletter, a thread series. Create once. Distribute forever.

Installed Instinct last week. Didn't get it. Yesterday, I gave it a new shot and spent 10 hrs on it: voice notes, screenshots, Gmail, Notion. Everything on whatsapp. I realize for small tasks, I'd rather message it than open Claude. Maybe the interface is the breakthrough.