Visualizee ↗: an AI rendering tool for architects and interior designers that grew after swapping a node editor for a chat box
- MRR
- $8,394 Verified 2026-10-21
- Total revenue
- $91,707
- Active subscriptions
- 294
Verified by TrustMRR
In one line: An AI rendering tool for architects and interior designers, with a watermarked free tier and monthly paid plans that remove the watermark and include a commercial license.
- Founder
- Piotr ObidowskiX ↗ · Solo
Timeline
- Revenue starts here, with very little in the first two years (self-reported)
- Revenue that month: $1,737 (TrustMRR)
- Revenue that month: $5,264 (TrustMRR)
- Revenue that month, the highest of the last 12 months: $9,929 (TrustMRR)
- Pauses all Meta ads after they brought in few payers (self-reported)
- Ships a new onboarding: signups who make a first render go from 55% to 76% (self-reported)
Compared with similar products
Sources: TrustMRR: Visualizee.ai; TrustMRR: Viztunr.com; X post by @pobidowski; X post by @pobidowski; X post by @pobidowski.
TrustMRR figures are monthly revenue as shown on 2026-10-23 (read-only sync from the payment provider).
4 moves Visualizee made
Dropped the node editor for a chat box.
The first version worked like ComfyUI and users found it too hard. He switched to a chat box with Vizzy, an assistant that writes prompts for users.
Moved from one-time payment to monthly plans.
Users only really started making images after this (self-reported).
Made SEO the main channel.
Asked which channel works best, he answered SEO.
Simplified sign-up.
One click to the first image, no password first, no paywall before the first image.
Sources: Starter Story interview 04:04; Starter Story interview 07:26; X post by @pobidowski; X post by @pobidowski.
How to copy it
Pick an industry and build a conversational image tool: upload a sketch or photo, the assistant asks a few questions and writes the prompt
Choose an industry whose people show images to clients, and build one page with Cursor or Claude Code: upload an image, the assistant asks a few questions, assembles a prompt and calls an image-editing API. No signup or payments yet. What the founder did: users upload a sketch or model and a chat assistant works out what they want so they don't need prompting skills (interview 01:33); his first node editor was too hard, and users just wanted to type a sentence and get an image (X post by @pobidowski).Expand ▾Collapse ▴
Choose an industry whose people show images to clients, and build one page with Cursor or Claude Code: upload an image, the assistant asks a few questions, assembles a prompt and calls an image-editing API. No signup or payments yet. What the founder did: users upload a sketch or model and a chat assistant works out what they want so they don't need prompting skills (interview 01:33 ↗); his first node editor was too hard, and users just wanted to type a sentence and get an image (X post by @pobidowski ↗).
Original page 1 ↗Original page 2 ↗Author post 3 ↗Original page 4 ↗
- A working page: upload an image, answer a few questions, get a result image.
- One of <your target users> makes a result from their own image and would show it to a client.
Build a one-page image tool for <your target users>: 1) upload a sketch or photo; 2) the assistant asks 2-3 questions for <your niche> (e.g. materials, time of day, style); 3) assemble the answers into a prompt and call an image-editing API; 4) show original and result side by side. Stack: Next.js + Supabase, secrets in environment variables, API parameters per official docs.
Used in step 1 to write this single-page tool; interview 09:09: he builds with Cursor and recently also Claude Code.
Used in step 1 as an alternative to Cursor for the same code; pick one.
Used in step 1 to build the web app; Next.js is in the founder's stack.
Used in step 1 to generate renders from the image and prompt; the founder's image service name is unclear, so this is our recommended substitute, not tested.
Measure signup to first image in PostHog and remove every obstacle before the first image
Track three PostHog events (signup, first upload, first result) as a funnel. Simplify signup with an optional password and show no payment prompts before the first image. What the founder did: one redesign made the first image one click, made passwords optional and removed the paywall before the first image, clearly lifting signup-to-first-image (X post by @pobidowski); in his view slow growth had really been the product being too hard to use (interview 07:53).Expand ▾Collapse ▴
Track three PostHog events (signup, first upload, first result) as a funnel. Simplify signup with an optional password and show no payment prompts before the first image. What the founder did: one redesign made the first image one click, made passwords optional and removed the paywall before the first image, clearly lifting signup-to-first-image (X post by @pobidowski ↗); in his view slow growth had really been the product being too hard to use (interview 07:53 ↗).
Author post 1 ↗Author post 2 ↗Original page 3 ↗Original page 4 ↗
- A three-step PostHog funnel with conversion recorded before and after the change.
- Signup-to-first-image conversion is higher after the change.
Below is my tool's PostHog funnel: signup, first upload, first result, plus every page and popup a new user passes before the first image. Please: 1) compute each step's conversion; 2) list every pre-first-image step that can be removed or made optional; 3) rank them by impact using only my data. Data: <paste>
Used in step 2 to measure the signup-to-first-render funnel; the founder monitors with PostHog.
Used in step 2 for one-click email login with an optional password; the founder uses Supabase for database, auth and storage.
Charge monthly with Stripe: a few free watermarked images, then tiers by image volume, and get the first payment
Create two or three monthly tiers in Stripe Checkout, split by images per month, at prices you set. Offer them when free credits run out or the user wants an unwatermarked download. What the founder did: one-time payments brought no recurring revenue, so he switched to monthly subscriptions, which worked, and argues for charging from day one (interview 07:26); today the free tier is watermarked and paid tiers scale by volume (visualizee.ai pricing).Expand ▾Collapse ▴
Create two or three monthly tiers in Stripe Checkout, split by images per month, at prices you set. Offer them when free credits run out or the user wants an unwatermarked download. What the founder did: one-time payments brought no recurring revenue, so he switched to monthly subscriptions, which worked, and argues for charging from day one (interview 07:26 ↗); today the free tier is watermarked and paid tiers scale by volume (visualizee.ai pricing ↗).
- Every tier pays in Stripe test mode; the plan picker appears when free credits run out; Stripe shows the first real payment.
- Stripe shows the first subscription not paid by you.
My image tool gives <n> free watermarked images; subscriptions are tiered by monthly images: <tier names, prices, counts>. Write: 1) English popup copy for when credits run out or an unwatermarked download is wanted; 2) an English pricing comparison table; 3) backend pseudocode that removes watermarks and grants credits per tier on Stripe webhook events (event names per official docs). Use only my numbers.
Used in step 3 to collect monthly subscriptions; the founder takes payments with Stripe, and TrustMRR's data comes from his Stripe.
Find image-intent search terms in your niche with Ahrefs, build a landing page with an upload box for each, and submit to Search Console
Enter a few seed terms in Ahrefs Keywords Explorer, keep terms whose intent is "I want an image made", and build one landing page per term with the step 1 upload box above the fold and before/after examples below. Connect Google Search Console and submit the sitemap. What the founder did: asked for his best channel he answered SEO (X post by @pobidowski), uses Ahrefs for it (interview 09:09), and stopped the Meta ads he tried after they didn't pay back (X post by @pobidowski).Expand ▾Collapse ▴
Enter a few seed terms in Ahrefs Keywords Explorer, keep terms whose intent is "I want an image made", and build one landing page per term with the step 1 upload box above the fold and before/after examples below. Connect Google Search Console and submit the sitemap. What the founder did: asked for his best channel he answered SEO (X post by @pobidowski ↗), uses Ahrefs for it (interview 09:09 ↗), and stopped the Meta ads he tried after they didn't pay back (X post by @pobidowski ↗).
Author post 1 ↗Original page 2 ↗Author post 3 ↗Author post 4 ↗Author post 5 ↗
- A batch of landing pages is live, each with an upload box above the fold; the sitemap is submitted in Search Console.
- These pages get impressions and clicks in Search Console and bring signups.
Below is my Ahrefs keyword export: term, volume, difficulty. Please: 1) keep only terms with "I want an image made" intent and drop "what is" terms; 2) pick the batch most worth doing first by volume and difficulty; 3) write an English title, H1, one hero line and 3 FAQs per term; 4) use only my data. Export: <paste>
Used in step 4 to find the industry's render search terms; interview 09:09: he uses Ahrefs for SEO.
Used in step 4 to submit the sitemap and see which pages get impressions; our recommendation, not mentioned by the founder.
Email people who cancel with a single question and decide what to change from the reasons
When a subscription is canceled in Stripe, add the email to a MailerLite group and send one automatic email asking only why they canceled (a few options plus other), then group the replies by reason. What the founder did: he tallied cancellation reasons and found users only need it when they have a project, which clashes with monthly billing (X post by @pobidowski, X post by @pobidowski), using MailerLite for email automation (interview 09:09).Expand ▾Collapse ▴
When a subscription is canceled in Stripe, add the email to a MailerLite group and send one automatic email asking only why they canceled (a few options plus other), then group the replies by reason. What the founder did: he tallied cancellation reasons and found users only need it when they have a project, which clashes with monthly billing (X post by @pobidowski ↗, X post by @pobidowski ↗), using MailerLite for email automation (interview 09:09 ↗).
- A cancellation triggers the email automatically; a table of reasons and counts.
- After a change aimed at the top reason, cancellations for that reason drop.
My image tool will email people who cancel with a single question. Please: 1) write the English email: one line of thanks, one question "why did you cancel", 5 options plus "other"; 2) group and count the replies below; 3) for the top reason, give one concrete product or pricing change direction, based only on the replies. Replies: <paste>
Used in step 5 to send a survey email automatically after a cancellation; the founder runs email automation in MailerLite.
Used in step 5 to trigger the email on cancellation and to build the new pricing option.
Sources
- TrustMRR
- TrustMRR public Markdown page (sync time, monthly revenue, subscriptions)
- Starter Story video interview (2026-09-02): 01:33 revenue dashboard, pricing and subscriber count
- Starter Story video interview: 04:04 the first version was a node editor that users found too hard
- Starter Story video interview: 07:26 switching one-time payments to monthly subscriptions
- Starter Story video interview: 09:09 tech stack and tools
- X: "Visualizee lifetime: about $88K since August 2023" (2026-10-21, self-reported)
- X: the founder answers that his best acquisition channel is "SEO" (2026-10-21)
- X: about 466 visitors and 79 signups a day; 55% of new users made a first render (2026-10-18)
- X: three changes that took signup-to-first-render from 55% to 76% (2026-10-17)
- X: 87% of paying users' renders are edits of their own image (2026-10-19)
- X: $821 on Meta ads brought in about $45 (2026-10-16)
- X: cancellation reasons breakdown (2026-10-20)
- LinkedIn: "For two years, my startup made about $150 a month" (2026-05-29, founder)
- Pricing page visualizee.ai/pricing
- Visualizee website
- trustmrr.com
Next case
One-click AI YouTube documentary generator, founded July 2026
Business model: Three subscription tiers; prices in the comparison table.
Data source: TrustMRR, checked 2026-10-26
