Jason Zigelbaum saw a gap and filled it, funding the build with his savings and revenue from another app. It took two years to get traction, but then, he doubled down on a specific segment. Now, he's at $125k MRR — and he's solo.
Here's Jason on how he did it. 👇
Before building Zigpoll, I built a couple of SaaS products. Shopify absorbed one of them (an app called Metafields Manager), and I sold the other, so I needed a reliable income.
I grew up in e-commerce, mostly on the agency side, and for years, I watched brands pour money into ads and analytics to understand what was happening on their stores, then guess at the crucial part — why people did what they did. Analytics could tell them a cart was abandoned, but never why. This gap between measurable data and essential insights bothered me enough that I eventually built the tool I wished we'd had internally.
That tool is Zigpoll. It's a survey and customer feedback platform, built initially for e-commerce, but now also used by many SaaS teams. It offers post-purchase, exit-intent, and CRO surveys — questions that capture a customer at the exact moment they will answer honestly. One question on a thank-you page ("What almost stopped you from buying?") can teach more than a $15K CRO audit, and it does so for free, forever.
I run the company solo: no cofounder, no funding, no sales team. It took about two years to get traction, but it has doubled in revenue every year since. I started 2026 with about $1.03M ARR and am closing June around $125K MRR, roughly a $1.5M run rate. This represents about a 44% increase in the first half of the year. That's nearly half a million dollars of new annual revenue added in six months, achieved solo.
My current goal is $2M ARR, which means adding about $43K more per month.
Building Zigpoll cost far more time than money. I'm technical, so I built the first version of Zigpoll myself with a code editor, not a budget. Nights and weekends were the real currency. I've never taken outside money. I took no VC, no angel, and had no cofounder to split equity or decisions with. That was a deliberate choice as much as a circumstance. I'd spent enough time around other people's businesses to know I wanted one that was fully mine, one I could change on a Tuesday afternoon without asking permission.
Meanwhile, my other app supported me, requiring only a couple of hours per week. When you're a solo developer, your main cost is the income you're not earning while you build, and I funded that gap myself rather than diluting to cover it. Just a product I thought should exist, and enough runway in my own savings to find out if anyone agreed.
The most helpful factor wasn't a tool; it was the vantage point. My agency background meant I'd already watched dozens of e-commerce brands hit the exact same wall, so I wasn't guessing at the problem.
I built the tool, then used the tool to validate and iterate. My customers' answers, rather than my own opinion, informed almost every good roadmap decision I've made. That's a cheap way to build when you're one person, because you're not spending months guessing. You ship the small thing forty people asked for, they stick around, and it compounds.
In fact, the original idea did not focus on post-purchase surveys, but as demand emerged, I leaned into that use case. The same pattern repeated for exit-intent, conversion rate optimization, and order delivery surveys.
Here's the stack:
JavaScript from front to back
Express + Mongo on the backend
React on the frontend
As the scale increased, I leaned into using Redis as a caching layer, but the stack is very stable.
Being a solo developer, it's important to lean into reliable third-party tools whenever you can. These tools save a ton of time, and if you choose your spots well, they can save a ton of money too.

Zigpoll is a straightforward SaaS subscription. Customers pay monthly, and plans are tiered by the number of survey responses they collect. Because I run the entire operation solo, without a team, office, or sales floor, margins are what you'd expect from a software business with one employee. Real costs include infrastructure, business tools, and some paid acquisition. That's the quiet advantage of staying solo: the model expands faster than the cost base.
Pricing is a product surface, not a one-time launch decision. I redesigned mine several times, and each version taught me something a spreadsheet never would have.
But I'd say the biggest unlock wasn't a price increase; instead, I noticed which customers naturally expanded and then deliberately built for them. Find the segment that grows when you serve them well, and pour your energy there.
As far as chrun, it exists, as it does for anyone selling to small ecommerce brands, some of whom close or pause seasonally. I treat it as a number to keep boring rather than a story to obsess over. I keep it boring by using my own product to find out why people leave and fixing issues they report.
No single launch made the business. No viral post. No Product Hunt day. Zigpoll grew, and still grows, through a handful of channels quietly compounding. When I break down where new customers come from, no single channel tells the whole story.
The Shopify App Store is the biggest single source, accounting for around a third of new signups. Building Zigpoll as a Shopify app first proved to be my most important distribution decision. As a solo founder with no marketing budget, the App Store placed me directly in front of brands with the problem I solve, precisely when they sought a solution. I couldn't have bought that reach. I earned it by optimizing the listing, taking reviews seriously, and shortening the install-to-value path so people quickly understood its benefit.
Word of mouth accounts for the next quarter of signups, and it's my favorite because I can't fake it. Almost all of it traces back to agency operators. A freelancer installs Zigpoll on one client's store. It works, so they go to the next client's store, and the next. When I analyzed "How did you hear about us?" answers, the most common phrase was a version of "A freelancer I work with uses you on everything." So I stopped treating that as a happy accident and started building for it intentionally. The more I remove friction for those who install me everywhere, the harder that flywheel spins.
AI tools represent a channel that barely existed a couple of years ago. Roughly 14% of new signups now come through ChatGPT, Claude, and Gemini, as people ask an assistant what to use and get pointed to Zigpoll. That's my third-biggest channel, and I treat it as an SEO problem for a new kind of search. I ensure the content, docs, and positioning are clear and specific enough that a model recommending a survey tool easily explains what mine does and who it's for.
The remaining channels collectively add up: Google search, some YouTube, and paid social. I write on LinkedIn nearly every day, appear on podcasts, and co-market with my integration partners. None of those are huge acquisition channels individually, but they accomplish something the App Store cannot. They build trust and familiarity, so when someone encounters my listing or asks an AI about surveys, the name isn't unfamiliar. I share the tactics and real numbers behind the business, and the product sells itself as a byproduct of its usefulness.
If I could start over, I'd start building in public from day one. I only got serious about writing openly and sharing the real numbers this year, and it has compounded faster than almost anything else I've done for distribution. It builds trust before someone reaches the product, and it forces me to think clearly about the business because I explain it out loud. If you're building something now, don't wait until you have a milestone worth announcing. The audience compounds like revenue does, and both reward the years you can't get back if you start late.
But here's my main growth advice: Find the channel where the platform handles distribution for you and commit to being the best option there. For me, that was the Shopify App Store. Then, observe which customers refer you without prompting, and build for that segment relentlessly, because word of mouth is the only channel that becomes cheaper as you grow instead of more expensive.
My most expensive mistake was misunderstanding the customer for too long. For a long time, I pictured my user as a single in-house ecommerce team, and I priced and built accordingly. At one point, I even gated integrations and AI features to the higher plans, a standard SaaS packaging approach on paper.
In practice, it quietly punished my best-growing segment — agency operators running Zigpoll across a dozen client stores — by charging them extra just to connect tools every client already used. I throttled exactly the people I most wanted, and I didn't see it until I read the onboarding data closely.
Fixing it meant rebuilding the pricing so integrations live on the standard plan, and then building for that segment on purpose. That correction largely explains why my revenue per account climbed 24% this year without a single price increase.
I wish I had committed to the agency operator segment far earlier. My data showed the signal for months before I acted on it, and every good roadmap decision I've made since came from listening to how those operators use the product differently than I imagined. The lesson generalizes: Watch who refers you and expands without being asked, and relentlessly build for them sooner than feels justified.
My advice: Listen to your customers.
The first mistake is thinking you listen when you only hear the loudest few. Founders who tweet ask Twitter, founders in a Slack group ask the group, and everyone mistakes the vocal minority for the market. Real listening is quieter and more structured than that. It means asking people who are using or leaving, right at the moment when they can tell you something true, then reading the answers closely enough to act. Most of my best decisions this year came from data sitting in front of me for months.
The most valuable feedback is almost never a feature request. It's the reason someone didn't buy, or the place they quietly wasted ten minutes and never mentioned it. Ask a customer, "How did we do?" and you get a rating you can't use. Ask "What almost stopped you from buying?" and you get a list of every objection your product page didn't close, written by the people who bought anyway. Timing matters as much as the question. A survey on the thank-you page right after checkout will out-teach any email you send three days later, because you caught them while the reason was still fresh.
And you need far less of this than you think. People tell themselves they'll act once the data is statistically significant, so the survey sits and the decision waits. Qualitative feedback doesn't work like a conversion test. You're looking for patterns in why people do things, and patterns show up fast. When forty of your first fifty responses say the same thing, you don't need response number eight hundred. You need the nerve to go fix it. Collecting is never the hard part. Acting on what you hear, especially when it contradicts the thing you were excited to build, is the whole game.
Here's what surprised me most. If you listen well enough, your customers won't just tell you what to build. They'll tell you who you are for. I spent a long time picturing my user as one kind of person, but my own onboarding answers revealed that real growth was coming from a completely different segment I hadn't been building for. The customer knew before I did.
That's what I'd tattoo on a new founder if I could: You are not the smartest person in your own business. Your customers are, collectively. The sooner you build the habit of asking them and believing them, the faster everything else gets easier. It's the reason I built a feedback company in the first place, and it's still the only strategy I fully trust.
My near-term goal is a number I think about most mornings: $2M ARR. I like having a number that's close enough to feel real but far enough that I can't coast to it. The number itself was never the prize, though. I am chasing the proof that a single person, listening carefully and shipping consistently, can build something that compounds without a team, a board, or a funding round pushing it.
My bigger goal is to stop being just an e-commerce tool. Zigpoll started on Shopify because that's the world I came from, but I built a feedback engine, and every business needs to understand its customers, not only those selling physical products. SaaS and app companies face the exact same problem in a different costume. They guess why people don't convert, why they churn, and what to build next. That's the same gap I set out to close for e-commerce brands.
So, I want to widen the aperture: making Zigpoll the obvious way for any business to ask its customers the right question at the right moment and turn the answers into decisions. The more value I can add across the whole ecosystem — through integrations, through AI tools easily recommending it, through serving the agency operators who run it across dozens of stores — the more useful the product becomes to people I wasn't even thinking about when I started.
And I want to do all of it the same way I've done it so far: solo, bootstrapped, changeable on an afternoon, built on what customers tell me rather than what I assume. The freedom to do that is the entire point, and hitting $2M without giving any of it up would mean more to me than a much bigger number achieved someone else's way.
You can follow along on LinkedIn and X. And check out Zigpoll!
Leave a Comment
The segment discovery is the most transferable lesson here. I'm in early customer discovery for a coding tool, and the same thing happened I kept pitching developers with sophisticated workflows and getting polite nods. When I started talking to people who didn't have their own workaround yet a tax lawyer doing no-code projects, a school admin they described the pain in more detail and with more urgency than anyone I'd deliberately targeted. The people who haven't built their own fix are the ones who'll actually pay.
Also, the GEO stat caught my eye 14% of signups from ChatGPT/Claude/Gemini recommendations. Building in the AI space myself, the products that get surfaced by LLMs are the ones whose use case fits in one sentence. "Post-purchase surveys for Shopify stores" is exactly that. Same forcing function as positioning for humans, just a faster feedback loop.