What changed
A new August 2026 Indie Hackers interview provides detailed operating evidence from Zipchat cofounder Ruslan Leteyski about rebuilding after Checkout X, a Shopify checkout product he says reached $8M ARR before platform changes made the product nonviable. Zipchat, founded in 2024, now sells an AI sales agent with plans based on AI replies and founder-reported revenue of about $167,000 per month, roughly 700,000 conversations per month and approximately $2M ARR.
Why it matters
The case connects platform dependency, pricing design, burn, financing and distribution in one operating history. For indie SaaS builders, the useful lesson is not the revenue milestone but how a prior single-platform failure changed the team's approach to monetization, channel diversification and dependency risk.
Checkout X exposed existential platform risk
Leteyski says Checkout X, bootstrapped from 2017, reached $8M ARR before Shopify changed its checkout platform and the product had to shut down. The historical revenue figure is founder-reported, but the broader lesson is concrete: a business built on a privileged platform surface can lose product viability when the platform owner changes that surface.
Zipchat adopted usage-linked revenue
Zipchat says its plans are priced by AI replies, with additional usage fees when customers exceed plan limits. The commercial logic is land-and-expand: merchants start with one surface such as website chat and add more channels as usage grows. Current Zipchat materials publicly show plans such as $49, $129, $249 and $499 tiers and usage-related overages.
The rebuild was not immediately profitable
Leteyski reports that Zipchat ran at roughly a $20,000 monthly loss for almost a year due to salaries, model costs and advertising. He says the company later reached break-even and used revenue-based financing tied to MRR rather than relying only on equity financing.
Hiring stayed constrained until revenue supported it
The founder says the company stayed extremely lean until about $50,000 MRR, with three founders covering product, sales/support and advertising. The team avoided hiring until break-even and affordability improved. This is self-reported operating history, but it is specific enough to examine as a bootstrapping decision.
Distribution is deliberately diversified
Zipchat says it tests YouTube, founder-led social media, SEO/GEO, app partnerships, Shopify App Store optimization, outbound email/calls, conferences and paid acquisition. Leteyski reports sending more than one million outbound emails per month and describes channel development as a multi-year process rather than a single growth hack.
Model and channel dependency are both being spread
The founder says Zipchat moved from OpenAI to Anthropic and is now working with multiple major model providers, while the product itself expands beyond a single website-chat surface. That does not eliminate dependency risk, but it suggests a deliberate attempt to avoid repeating the single-platform exposure that hurt Checkout X.