Loading page…
An AI marketing stack for independent restaurants that gives away the website, online ordering and loyalty for free, charges for the AI layer on top, and sells the whole thing exclusively door-to-door — $3.5M ARR growing about 400% year over year with zero dollars spent on marketing
about 400% year-over-year growth, described as accelerating · roughly $6K average ACV per location · 300–400 demos a month · 25+ full-time staff, three engineers plus the founder · around EBITDA break-even (some months slightly up, some slightly down) · $450K of non-dilutive financing from Founderpath on a 16-month payback, about $35K a month fixed
"We are a solar company, a pest control company, a roofing company" — Carl Turner sells restaurant software the way roofs and solar panels get sold, on the doorstep, because "every AI cold caller or human cold caller will hit a wall of us having a beer."
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Carl Turner's pitch is deliberately unglamorous: "You can think about us as a super boring marketing agency, just that there is no one actually in the agency and it's AI." Swipe.by sells to SMB brick-and-mortar owners, and inside that it goes narrower still — independent restaurants. His read on the customer is the whole thesis. A restaurant owner "works in the business, not on the business": they make the food, they call Spectrum when the Wi-Fi is out, they call the oven guy when the oven breaks, they serve the food when two staff members did not show up. Marketing is the first thing that falls off that list, which is why the product runs on autopilot rather than handing the owner another dashboard with a magic-wand button to click. The company is at about $3.5M ARR, growing roughly 400% year over year and, in his words, accelerating. There are over 25 full-time people, of whom three are engineers plus Turner himself — everything else is distribution. It sells exclusively door-to-door, spends nothing on ads, and has no inbound funnel at all. Growth has been funded with angel and family-office money, then revenue factoring through Stripe, then $450K from Founderpath, all while running roughly at EBITDA break-even. Turner is open about how he got here: "I failed twice with this company. And failed hard. I mean, failed hard." The current shape is the third attempt, and he has term sheets on the table he is in no hurry to sign.
Swipe.by opens every market the same way, and the motion is copyable from day one. A rep lands in a city, works about seventy touch points a week, and books demos that a market lead runs over a screen share while the rep sits with the owner.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself. The exact words are quoted below, with a timestamp that jumps to the second each figure is said.
“And so, So, in your case here, you like there's 35 grand a month that's fixed, you know you can plan it.”
9:24 · hear it in the video“Your expectation is Nathan will pay you 30 to 40% of any new sales you drive, which average 6K ACVs.”
15:45 · hear it in the videoData credibility: Founder interview. Every figure — $3.5M ARR, roughly 400% year-over-year growth, 25+ staff, ~$6K ACV, 300–400 demos a month, EBITDA break-even — is stated by Carl Turner on camera and is unaudited; no dashboard, filing or third-party source is shown. The interviewer runs Founderpath, which lent the company the $450K discussed in the interview, so an outside party has underwritten the revenue, but none of that underwriting data appears on camera and the conversation is promotional for both sides. Churn, net revenue retention, customer count and gross margin were never disclosed, and the valuation talk ($15M declined, term sheets at close to double, a $60–75M target) is negotiation in progress rather than a closed round. Our video metadata for this upload is incomplete: the channel name and exact title were not captured, and we identify the host only as Nathan of Founderpath, as addressed in the recording. The country is inferred from what is said — Miami as a named market, W-2 employment, Indeed recruiting — rather than from a stated headquarters.
There is no free trial in this business and there never has been.
Two things compound here. The first is the giveaway.
The quiet years are the most honest part of this story. "I failed twice with this company," Turner says.
Underneath the marketing language, Swipe.by is three layers. The bottom layer is a set of commoditized restaurant tools given away at no charge: the website, online ordering, loyalty and rewards, with reservations and staff scheduling on the roadmap. Turner's view is blunt — "anyone who charges you to design your website is just insanity in today's world" — and the giveaway is how he displaces the incumbents whose entire business is one of those tools. The middle layer is data. POS integrations pull in-store sales, payments and sometimes financing data, and the AI phone system is built to capture express verbal consent on a call so the customer can be marketed to later. That is the actual differentiator he describes: "Can we get an express verbal consent from a phone call to turn a phone call for a burger into 20 more burger transactions down the line?" Most competitors, he argues, stop at the transaction — an AI voice that helps someone buy a burger, which is what the phone already did. The top layer is the paid one: social media, remarketing, reputation management, AI phone answering that can clone the owner's voice, and printed direct mail on the way. Our read: the phrase that matters is "fully autopilot." He is scathing about rivals whose "automated" email still needs someone to click a magic wand, and for an owner with no time that distinction is the entire product.
Average contract value is about $6,000 a location per year, and the majority of customers pay for the year up front. Nothing is sold self-serve: a demo is mandatory, partly because owners who hear "we'll help with email" immediately think "I already have Mailchimp," and partly because the company wants the relationship before the money. The pricing architecture is the interesting part. Free website, free online ordering, free loyalty — all monetized indirectly by making the AI services on top the only line item. Turner frames it as cutting the point solutions out entirely: "There is other companies where that's their bread and butter. And we can come in and cut them out entirely. We then monetize with the AI layer on top." Our read: this is a land-grab structure that only works if the AI layer retains. Giving away four products to sell one means the gross margin on the free tier is negative and the payback lives entirely in renewals — which is exactly why annual prepay and the no-free-trial rule are not pricing quirks but survival mechanics. Note also what is missing from the interview: no churn number, no net revenue retention, no gross margin. At $3.5M ARR with roughly 400% growth, those are the numbers that decide whether this is a $60M company or a treadmill, and they were not disclosed.
This is the most replicable piece of the whole business, and Turner gave the full comp math on camera. A rep's quota is about six locations a month, which at roughly $6K ACV is about $36,000 of new ARR a month. The rep takes home 30 to 40% of that. There is no straight commission-only trap: there is an "activity guarantee" — knock the doors and set the appointments and you make roughly $36K to $40K a year even if you close nothing. Activity is measured at around 70 touch points a week, about 10 doors or demos a day, and touch points are not unique, so three visits to the same restaurant in a day counts as three. Do 10 demos in a week and the walk-in requirement drops. Promotion is fast and deliberate: strong reps move to W-2 with no commission within about six months and are sent to open a new market. Everyone who lasts two years vests a small equity sliver. Reps are recruited through Indeed ads, and 300 to 400 demos a month run through the market leads on screen shares. Our read: the reason this works against AI outreach is positional, not technical. When every competitor is running the same Clay-plus-voice-agent sequence into the same inboxes, the scarce thing is a human who shows up. "Every AI cold caller or human cold caller will hit a wall of us having a beer."
Turner has never taken venture money. He started with angels, family offices and high-net-worth individuals, then used revenue factoring through Stripe, then took $450K from Founderpath on a 16-month payback — about $35,000 a month, fixed. The switch away from Stripe is a useful lesson for anyone growing fast on revenue-based financing. Stripe took a percentage of every dollar collected, which is fine when growth is flat; at his growth rate the company was retiring the full loan inside three months. "Stripe has fairly good terms if you pay them back over a year and a half. If you pay them back within 3 months — time value of money — interest is kind of insane." A fixed monthly payment over 16 months was worth more to him than a lower headline rate, because it let him plan cash against commissions. On valuation he is candid to the point of being quotable. He would turn down $15M all cash. He has VC term sheets at close to double that, and is pushing for $60M to $75M on a $10M check, which he acknowledges is a 20x to 25x ARR multiple. What worries him is not a down round — "I don't need money to grow the business" — but liquidation preferences: at a 2x preference on a $10M check, a $20M exit returns the founder nothing.
Copy the positioning first. Picking one vertical inside SMB — not "local business" but independent restaurants — is what lets a four-person engineering team ship something that feels finished. Turner tried widening to hair and nail salons late last year and pulled back, because the product was "just missing 5% to be perfect for them." That 5% is the whole lesson about vertical software. Copy the free-tools wedge. Anything commoditized in 2026 — a site, online ordering, a loyalty punch card — is better used as a door opener than as a product, and it buys you the data that makes the paid layer work. Copy the no-free-trial rule if you sell to owners who are physically busy. Skin in the game is the difference between a signed account and a dead one. Change the sales model unless you can fund it. A 25-person company with three engineers is a sales organization with software attached; that takes commission cash before revenue arrives, which is precisely why Founderpath is in this story. Our read: a solo builder should copy the wedge and the verbal-consent data capture, then sell through the people who already walk into these restaurants weekly — food distributors, POS resellers, linen and produce reps — rather than trying to hire a door-knocking team from scratch.
Churn is the first one, and Turner names it himself: "There's crazy churn in SMB SaaS." Restaurants close. Owners are flaky and switch when someone undercuts on price, however much value you added. A company growing 400% is adding accounts faster than it loses them; the moment new-market ramp slows, retention becomes the only number that matters, and no retention figure was disclosed. Second, the cost structure is upside down for a software company. Most of year-one revenue goes to the rep who closed it, and four products are given away to sell one. That works at break-even EBITDA while growth is compounding; it is fragile if ACV has to come down. Third, competition from above. Toast and the POS incumbents already own the restaurant relationship and are shipping their own AI layers. Turner says "I don't want to be Toast," but he is selling into accounts Toast can bundle against for free. Fourth, the valuation he is chasing. Twenty to twenty-five times ARR on $3.5M means growing into a $60M-plus price. He is right that a down round is survivable when you are profitable, but a stacked liquidation preference on later rounds is the trap he flagged and the one most likely to bite. Our read: the fifth risk is key-person concentration. The moat is a sales culture, and sales cultures live or die on the person recruiting and promoting into them.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,435 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click