---
title: Stuck with no traction? Remove the friction before the first success
slug: removing-product-friction-pivot
summary: Visualizee sat at ~$100–150/month for two years until a node-graph UI became a chat with a prompt assistant (and one-time payments became subscriptions) — ~$10K/month; Cal AI won by turning calorie "homework" into one photo; activation friction defines what a category can do.
domain: mobile-apps
tags: [product-strategy, friction, activation, pivot, ai-apps, simplicity]
related: [app-mvp-single-problem-scope, differentiating-in-saturated-app-categories, app-founder-case-studies-common-patterns, product-experiment-process-okrs-and-activation, adopting-subscriptions-and-who-should-subscribe]
sources: [yt:TUN8Y8__QGU, yt:OxE79rpGioI, yt:OV09yPLf8BI, yt:s8Rped5LuWw]
updated: 2026-10-24
---

# Remove the friction before the first success

## TL;DR
- **Before blaming marketing, check whether the product is too hard to use.** Piotr blamed slow growth on marketing for two years; the real problem was usability.[^1]
- Visualizee v1 was a **ComfyUI-style node system** (upload → connect nodes → write → generate). Users said it was too hard; v2 is a **ChatGPT-like chat** with an assistant ("Vizi") that asks follow-ups and writes the prompt.[^2][^3]
- Result (self-reported): from ~$150/month to ~$8.2K MRR, ~$10K/month with one-time payments; 300+/day landing visits, 5–10 trials/day, ~25% trial-to-paid.[^4][^5]
- Same pivot: **one-time payments → monthly subscriptions**, "from day one".[^6][^1]
- Cal AI's whole edge was friction removal: typing meals turns tracking into "homework"; a photo replaced it and users switched trackers.[^7][^8]
- Advice: "Don't build the most powerful product you can. Build the simplest product your customer needs."[^9]

## Details
**How Piotr saw it ([[people/piotr-visualizee]]).** After the change, signups started rendering — the data showed the fix.[^6] He stuck with the idea because competitors with similar products clearly had customers, so the problem had to be his execution.[^10] He held on too long to the belief that the original design was great.[^11] Stack: Cursor and Claude Code, fal.ai GPUs, Next.js, Supabase, Vercel, Stripe, PostHog, Sentry, MailerLite.[^12]

**Activation friction is category-structural.** Symmetry's founder: in a gym tracker, activation means getting someone to the gym and logging a workout — "crazy friction" — while in a calorie app activation is just snapping a photo.[^13] Gleam's founders concluded retention is largely structural to category and hard to game, so they stopped obsessing over it and pushed on what worked (marketing, monetization).[^14]

**Novel-but-heavy flows backfire.** Gleam tried voice onboarding; it was "completely catastrophic".[^15]

## Disagreements / caveats
- Visualizee is a **web SaaS for architects**, not a mobile app; the lesson (friction to first output) transfers, the pricing ($35/$80 per month) doesn't.[^4]
- Starter Story frames the result as "a ChatGPT wrapper won"; the simpler UI and the subscription switch shipped together, so their effects can't be separated.[^6]

## See also
- [[app-mvp-single-problem-scope]]
- [[product-experiment-process-okrs-and-activation]]

[^1]: yt:TUN8Y8__QGU @ 08:03 — Piotr (Visualizee)
[^2]: yt:TUN8Y8__QGU @ 04:03 — Piotr (Visualizee)
[^3]: yt:TUN8Y8__QGU @ 05:38 — Piotr (Visualizee)
[^4]: yt:TUN8Y8__QGU @ 02:26 — Piotr (Visualizee)
[^5]: yt:TUN8Y8__QGU @ 03:15 — Piotr (Visualizee)
[^6]: yt:TUN8Y8__QGU @ 07:12 — Piotr (Visualizee)
[^7]: yt:OxE79rpGioI @ 05:27 — Jake Castillo
[^8]: yt:OxE79rpGioI @ 07:55 — Jake Castillo
[^9]: yt:TUN8Y8__QGU @ 09:36 — Piotr (Visualizee)
[^10]: yt:TUN8Y8__QGU @ 08:49 — Piotr (Visualizee)
[^11]: yt:TUN8Y8__QGU @ 07:12 — Piotr (Visualizee)
[^12]: yt:TUN8Y8__QGU @ 08:49 — Piotr (Visualizee)
[^13]: yt:OV09yPLf8BI @ 08:15 — Mauro (Symmetry)
[^14]: yt:s8Rped5LuWw @ 31:04 — Kelvin Zhang
[^15]: yt:s8Rped5LuWw @ 08:32 — Kelvin Zhang