Zigpoll: Solo at $125,000/Month — the Day It Realized It Had Misread Who Its Real Customer Was

Dollar amounts in this article include a rough conversion at ¥150 to $1.

No co-founder, no funding, no sales team. Jason Zigelbaum runs ‘Zigpoll,’ a survey SaaS for e-commerce sites, entirely on his own, and reached $125,000 MRR (roughly ¥18.75M/month) in June 2026. Annualized, that’s around $1.5M (roughly ¥230M).

More interesting than the growth trajectory itself is what drove it: not a new feature, not a new channel, but the discovery that he had misread who his own customer actually was. And that misread came straight out of textbook SaaS pricing design.

The numbers, in full

What it sells

Zigpoll is a tool for embedding post-purchase, exit-intent, and CRO (conversion rate optimization) surveys into e-commerce and SaaS sites. Billing is subscription-based, tiered by the volume of responses collected.

The product philosophy is captured in one of his own lines: ‘Analytics tells you a cart was abandoned, but it never tells you why.’ It’s positioned as a tool for capturing, in the moment, the reasons that live outside the behavioral log.

The founder came up through the agency side of the e-commerce industry. He’d previously built two SaaS products — one (Metafields Manager) was acquired by Shopify, and the other was also sold. He used that income to fund Zigpoll’s development. In the early days, it was a nights-and-weekends project alongside his day job, at ‘a few hours a week.’

Where the turning point was

Zigpoll’s turning point wasn’t a feature launch or a marketing campaign. It was the moment reading onboarding data revealed that the assumed customer and the actual growing customer were different people.

He’d originally assumed his user was ‘an in-house team at an e-commerce brand.’ In reality, the fastest-growing segment turned out to be agencies and freelancers managing multiple client stores. One person would install it on one store, then install it again on the next project, a spreading pattern.

The problem was in the pricing design. He had locked integrations and AI features behind the top-tier plan, a standard SaaS packaging move. But that design directly penalized the agency segment that used integrations heavily across multiple client environments. His own words nail it: ‘I was restricting the exact segment I wanted most. I didn’t realize it until I actually sat down with the onboarding data.’

The move he made was simple. He moved integrations into the standard plan and deliberately started building for agency operators. The result: a 24% increase in revenue per account with no price increase. The number on the price sheet didn’t move, yet the amount extracted from a single account went up: growth that came from unblocking a bottleneck, not from adding anything new.

Why the agency segment mattered so much

This deserves a structural look rather than a surface reading.

It multiplies, to start. An in-house team is one company, one account, period. An agency or freelancer installs it once per client they manage. For the same acquisition cost of landing one customer, the number of installs that follow is an order of magnitude different.

Referrals happen automatically, too. He cites a real example: ‘a freelancer we work with uses your tool on every single project.’ It expands without being asked, and it gets referred without being asked. In fact, roughly 25% of sign-ups come through word of mouth, the second-largest channel overall. That’s a zero-cost sales channel operating on its own.

Above all, this functions as the best possible signal of product-market fit. As he puts it: ‘Look at who’s referring you and expanding without being prompted, and build relentlessly for that segment.’ It’s a way of measuring success not by what you sold, but by what’s growing on its own, unpushed.

The channel mix backs this structure up. The largest is the Shopify App Store, at about 33% of all sign-ups, a place where brands already grappling with the problem are lined up directly in front of it. Next is word of mouth at about 25%. And roughly 14% now comes through generative AI. ChatGPT, Claude, and Gemini. The remainder is split among Google search, YouTube, LinkedIn, podcasts, and paid ads. He frames the AI-sourced traffic as ‘SEO for a new kind of search’: a product whose use case can be explained in a single clean sentence gets recommended, while one with an ambiguous positioning gets skipped.

The cost of the misread, and what’s still weak

The most expensive mistake he cites is exactly this, misreading his real customer for a long stretch. The cause was putting a single-in-house-team user image in place first, then applying standard SaaS packaging logic on top of it without validating it against the actual expansion pattern.

In other words, the growth here wasn’t produced by starting something new so much as by releasing growth that was already happening, but that his own design had been blocking. The upside was inside, not outside.

He’s also matter-of-fact about churn. Small e-commerce businesses are seasonal, and a certain number of customers close their stores or pause. He treats this ‘not as a story worth chasing, but as a number that should stay boring’, asking about the reason for cancellation in his own product and fixing whatever issues come up.

Another regret he mentions is about publishing openly. He wishes he’d ‘gone public from the start,’ and says, ‘I only seriously started writing openly, including numbers, this year. It’s compounded faster than almost anything else.’

What’s replicable, what isn’t

What’s easy to bring home is the method of validation: read the onboarding data, find who’s expanding and referring unprompted, and remove the design obstacle that’s locking that segment out. This procedure itself can be executed regardless of scale. His point on the timing of listening to customers is specific too: ‘Asking “what almost stopped you from buying” on the thank-you page tells you more than an email three days later.’ And a threshold for decision-making: ‘If 40 of your first 50 responses say the same thing, you don’t need the 800th. What you need is the courage to go fix it.’

That said, the non-transferable conditions are just as clear. He’d already built two SaaS products, with one acquired by Shopify, and that exit income funded Zigpoll’s development period. Having that separate income stream is exactly what let him endure roughly two years at just a few hours a week before traction hit. And the Shopify App Store, which supplies a third of sign-ups, isn’t a channel that works the same way for everyone, being placed at the point of decision in a market where problem-aware businesses are already gathered mattered enormously. $125,000/month, solo, can’t really be explained without that placement.

He’s stated he plans to keep going solo and bootstrapped. His reason for choosing a structure with ‘no co-founder, no funding, no sales team’ is that he wanted a business he could change on a Tuesday afternoon without asking anyone’s permission. The agility to restructure pricing plans the day after spotting a misread is itself a byproduct of this structure.

Related reading

• How solo-run Carrd reached $2M ARR
• Plausible: $1M ARR with a 4-person team and zero ad spend