Trends & new platforms

About 60 episodes credit timing as much as the idea. Four patterns recur. First, new capabilities and APIs: Notion’s API spawned a forms add-on and a Sheets export tool, ChatGPT’s launch and image upload triggered browser extensions and chart-analysis apps, and iOS’s AlarmKit let Jake Glacer’s alarm app work properly. Guests advise building ahead of the capability so you are ready on launch day. Second, incumbent shocks: Denis Yurchak launched a Skype alternative days after its closure was announced, and Thomas Sanlis launched when indie hackers were complaining about Product Hunt. Third, social trends: Cedric Roberge’s peptide app and the looks-maxing apps came from TikTok obsessions with no app yet. Fourth, growing platforms and markets (Airtable, AI app builders, Pinterest), where built-in distribution and 1% of a booming market are enough; Pat argues a growing market forgives mediocre execution. The counterweights: windows are short and copycats arrive fast; guests themselves call timing luck and hard to replicate; platform dependency can nearly kill a business (TweetHunter after the X takeover); Steph France prefers lasting human needs over hype; and Pat warns that a trend is only a vehicle — without marketing and a product that already solves a problem it has ‘no gas’. Revenue claims such as Base 44’s and Lovable’s ARR are second-hand.

Key points

• New APIs, model capabilities and OS features open short windows for missing features — 11 episodes. (Julien Nahum · Julien Nahum (Nov 2024) ▶ 01:39) (Leandro · Leandro (Jun 2025) ▶ 02:00) (Timo Koehler · Timo Koehler (Jul 2026) ▶ 03:45) +8 more
• Timing is a main difference-maker between projects that work and those that don’t — 9 episodes. (Domenico Gagliardi · Domenico Gagliardi (Oct 2025) ▶ 07:01) (Thomas Sanlis · Thomas Sanlis (Aug 2025) ▶ 05:54) (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 08:16) +6 more
• Build on growing platforms and markets: distribution is built in and 1% of a boom is enough — 7 episodes. (Andy Cloke · Andy Cloke (Aug 2025) ▶ 03:55) (Julien Nahum · Julien Nahum (Nov 2024) ▶ 02:26) (David · David (May 2026) ▶ 04:33) +4 more
• Turn social-media obsessions with no app yet into products — and be first — 6 episodes. (Cedric Roberge · Cedric Roberge (May 2026) ▶ 02:58) (unnamed guest · Nikita & Evgeniy (Sep 2025) ▶ 02:35) (Domenico Gagliardi · Domenico Gagliardi (Oct 2025) ▶ 06:50) +3 more
• Windows close fast; timing and platform bets carry luck and dependency risk — 6 episodes. (unnamed guest · Jake Glacer (Jul 2026) ▶ 17:19) (presenter · Jake Glacer (Jul 2026) ▶ 05:29) (Cedric Roberge · Cedric Roberge (May 2026) ▶ 06:58) +3 more
• Launch fast when an incumbent shuts down or its users revolt — 2 episodes. (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 08:23) (Thomas Sanlis · Thomas Sanlis (Aug 2025) ▶ 05:33)
• Host view: a growing market forgives mediocre execution — 2 episodes. (presenter · David (May 2026) ▶ 11:11)

Where guests agree

• Being early to a new capability, platform or trend materially helps a small product get noticed (Yasser Elsaid · Yasser Elsaid (Jul 2025) ▶ 09:12) (Saeed Ezzati · Saeed Ezzati (Aug 2024) ▶ 06:17) (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 04:48) +3 more

Where guests disagree

Should you chase trends? ⚠️

• Yes — trends and timing are the main difference-maker; ride what is hot on your channel (Domenico Gagliardi · Domenico Gagliardi (Oct 2025) ▶ 07:08) (unnamed guest · Nikita & Evgeniy (Sep 2025) ▶ 02:35) (presenter · Cedric Roberge (May 2026) ▶ 16:44) +1 more
• Be wary — prefer lasting human needs; a trend is only a vehicle without a product and marketing (Steph France · Steph France (Jun 2025) ▶ 08:35) (presenter · Nevo (Jul 2026) ▶ 13:14)

Frameworks

• Platform ideation framework (six steps) — Andy’s method for finding micro-SaaS ideas on growing platforms. (Andy Cloke · Andy Cloke (Aug 2025) ▶ 04:57)
  • Find a growing platform (e.g. via Exploding Topics)
  • Find a pain point (forums, Reddit, Twitter)
  • Borrow a proven add-on/pattern from a more established platform and make it feel native
  • Check you can integrate (public API, marketplace, extension SDK)
  • Napkin maths: platform users × problem frequency × willingness to pay (use the established platform’s price points)
  • Work out whether the platform will crush you (roadmap, support forums, how native the feature feels)
• Trend-to-app process (five steps) — Cedric’s process for finding and launching a trend-based app if starting over. (Cedric Roberge · Cedric Roberge (May 2026) ▶ 08:38)
  • Live on TikTok/Instagram and find what people are obsessed with
  • Check whether apps already exist (a few is a good sign)
  • Mine every competitor review for likes, dislikes and wishes; feed them to Claude or ChatGPT for your app idea
  • Get niche influencers posting before launch
  • Have a clear marketing/distribution plan before building
• Why Yadaphone worked — Three components he credits. (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 08:16)
  • Validated market (tweet, then Reddit sales)
  • Timing around a big news event (Skype closure)
  • Enough accumulated skill to build something good
• Generative-AI substitution idea framework — Find tasks that people currently pay to do manually and that existing or imminent generative AI could significantly enhance or replace; anticipate where AI is heading. (Ramsri Goutham · Ramsri Goutham (Jun 2025) ▶ 06:32)
  • Identify manual tasks people pay for
  • Check whether current or imminent generative AI can enhance or substitute them
  • Track generative AI progress to spot newly possible, cheaper or simpler use cases
• Criteria for an affiliate niche — What Alex looks for in an uncrowded niche to target. (Alex Goldberg · Alex Goldberg (Mar 2024) ▶ 06:11)
  • New (little content exists)
  • High margin in the product category
  • Lots of activity or momentum
  • Staying power (tried and true)

Tactics

• Build ahead of model capability — Start on something current models can’t quite do and bet the next model will; he had Chatbase ready when the ChatGPT API launched and tweeted a ‘48 hours’ build that went viral. (Yasser Elsaid · Yasser Elsaid (Jul 2025) ▶ 08:56)
• Watch platform API changes for app ideas (host reflection) — New OS APIs (AlarmKit, widgets, Screen Time) trigger a wave of apps; build in the short window after the change. (unnamed guest · Jake Glacer (Jul 2026) ▶ 16:50)
• Launch into a competitor’s community backlash — Launch an alternative when users publicly complain about the incumbent (indie hackers unhappy with Product Hunt featuring only big products). (Thomas Sanlis · Thomas Sanlis (Aug 2025) ▶ 05:33)
• Get into articles about a departing competitor — Search keywords like ‘Skype alternative’, find the ranking blog posts, contact the authors via X, LinkedIn or email, and ask them to mention you or replace the old brand with yours. (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 10:32)
• Build on top of a platform with a big existing community — When Notion released its API, he built an integration for its fans, cannibalising part of the community and solving a sub-issue. (Julien Nahum · Julien Nahum (Nov 2024) ▶ 02:26)
• Find ideas from trends and sentiment on X — All his projects started from trends he tracked on X. (Domenico Gagliardi · Domenico Gagliardi (Oct 2025) ▶ 06:50)
• Ride a platform news moment with an explainer thread — When X open-sourced its algorithm, he read the GitHub code and wrote a thread on how it works and what makes posts go viral. (Alex Finn · Alex Finn (Jun 2025) ▶ 02:31)
• Write long-form X articles on a trending topic — With 200 followers he wrote an X article on using OpenClaw with Postiz and got half a million views; X was pushing articles because people read them longer than posts. (Nevo · Nevo (Jul 2026) ▶ 05:33)
• Ride an adjacent wave (MCPs) — Adding a searchable MCP section when MCPs arrived drove a second wave of adoption. (Pontus Abrahamsson · Pontus Abrahamsson (Nov 2025) ▶ 11:05)
• Be first to cover new AI releases — Covering a new model early lets small channels rank. (Jacky Chou · Jacky Chou (Mar 2026) ▶ 08:01)

Stories & examples

• He used Skype to call offices abroad; when Microsoft announced its closure, Pieter Levels tweeted encouraging people to build an alternative. He built a prototype over a weekend, launched on Reddit and X and got sales within minutes. (Denis Yurchak · Denis Yurchak (Oct 2025) ▶ 04:48)
• Timing the ChatGPT API launch: Chatbase already worked before the ChatGPT chat API came out; when it launched, he tweeted that he had built a new product in 48 hours and it went viral amid the API buzz. Lesson: Be ready before the capability arrives. (Yasser Elsaid · Yasser Elsaid (Jul 2025) ▶ 09:12)
• Fintech comparison pivots to telemedicine: The site began comparing fintech companies; seeing $100 million Series A rounds going to telemedicine with no review content, they pivoted to health and wellness and were among the first to review platforms that became household names. Lesson: Go where money is flowing but content doesn’t yet exist. (Alex Goldberg · Alex Goldberg (Mar 2024) ▶ 02:36)
• A roommate suggested taking peptides, then a peptide TikTok appeared on his For You page; peptides were huge on social media with no successful app, so he studied every existing app’s reviews and built the best version. (Cedric Roberge · Cedric Roberge (May 2026) ▶ 05:57)
• Telegram bot to mobile app: When ChatGPT added image upload, the brothers built a Telegram bot to analyse stock charts, then turned it into the Chart Detector AI app. Lesson: A new capability in a major AI API can open a wave of app ideas. (Timo Koehler · Timo Koehler (Jul 2026) ▶ 03:45)
• He already knew the Sheets API from client work when Notion released its official API; searching the Notion subreddit for sheets/Excel/CSV keywords showed many people trying to export Notion data to Sheets, so he built an MVP in two weeks. (Leandro · Leandro (Jun 2025) ▶ 01:58)
• Cracking an AI detector: After a TikTok video about an AI detector claiming 99% accuracy, he reverse-engineered it until every response came up human, then built StealthGPT. Lesson: Spotting an emerging push-back to a trend can reveal a product. (Jozef Gherman · Jozef Gherman (Jul 2024) ▶ 06:23)
• COVID timing: When COVID hit, tech companies redirected in-person conference budgets into content while still funded, pouring money into developer-space marketing just as Draft.dev started. Lesson: Market timing can accelerate a niche service. (Karl Hughes · Karl Hughes (Oct 2023) ▶ 10:25)
• TweetHunter nearly killed by the X takeover: When Elon Musk took over X, TweetHunter was almost killed while making 200K per month. Lesson: Platform or AI shifts can kill a product overnight; a portfolio makes the business resilient. (Tibo · Tibo (Dec 2025) ▶ 16:32)
• Athletes Thread and the NIL ruling: Having worked in licensed merchandise for years, they saw an opportunity when the name-image-likeness ruling let student athletes monetise, and launched merchandise co-branded with colleges and athletes. Lesson: Existing licences and industry knowledge let them move quickly when a regulatory change opened a market. (Karthik Shanadi · Karthik Shanadi (Apr 2023) ▶ 01:56)
• Before iOS 26’s AlarmKit API, third-party alarm apps were gimmicky and unreliable because they relied on app sounds and workarounds. (Jake Glacer · Jake Glacer (Jul 2026) ▶ 05:14)

Numbers

• Duolingo monthly revenue and his 1%-of-market estimate: $33 million/month; $4 million/year at 1% USD (self-reported, approx.) (Blake Anderson · Blake Anderson (Dec 2024) ▶ 07:17)

Caveats & warnings

• Platform-API app opportunities have only a short window (host reflection). (unnamed guest · Jake Glacer (Jul 2026) ▶ 17:19)
• Pat notes luck and timing played a part: the AlarmKit change coincided with Jake having already shipped. (presenter · Jake Glacer (Jul 2026) ▶ 05:29)
• He says the idea blew up almost overnight, so it’s hard to give a general method. (Cedric Roberge · Cedric Roberge (May 2026) ▶ 06:58)
• Being early helped but isn’t always replicable advice. (Dawson Botsford · Dawson Botsford (Apr 2024) ▶ 09:29)
• Not every simple project will work; timing was crucial and he built very early. (Saeed Ezzati · Saeed Ezzati (Aug 2024) ▶ 06:23)
• Building agentic features alone is ‘boring’ and won’t drive revenue without marketing and a product that already solves a problem; beware trend-hopping. (presenter · Nevo (Jul 2026) ▶ 13:14)
• Avoid markets in decline, such as tools for newspaper owners. (Diego Roshardt · Diego Roshardt (Jul 2025) ▶ 11:12)
• A project that fails in January may work in July — don’t quit only because of timing. (Thomas Sanlis · Thomas Sanlis (Aug 2025) ▶ 04:12)

How much to trust this

• Evidence: 59 episodes, 59 independent speakers (repeat guests counted once; Pat and Gus counted as one voice).