Loading page…
A direct-to-consumer air filter manufacturer doing $22M a month — built by drop shipping other people's filters at almost no margin until the customer acquisition proved it repeated, then buying the factory.
about 1,000 employees · began as a near-zero-margin drop shipping operation with no inventory
"I had customers before I started our air filter factory." He drop shipped at a margin he describes as almost nothing — not to make money, but to find out whether the acquisition system repeated before he spent a dollar on machinery.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
David Heacock opens his own video by stating what his company does: "I'm David. I'm CEO of Filterbuy, an air filter company that makes $22 million a month." Everything below comes from that video, so read it as a founder describing his own business on his own channel — the credibility note sets out exactly what that means. The business sells residential air filters and manufactures them itself. Heacock started it after moving back to Talladega, Alabama, a town where, as he puts it, he looked around at the local industry and "there wasn't much." The company now runs multiple factories — he describes opening a plant in Salt Lake City, Utah — and employs around a thousand people. The part worth studying is not the $22 million a month. It is the order in which he did things. He did not build a factory and then go looking for demand. He says it plainly: "I had customers before I started our air filter factory." The first version of Filterbuy was a drop shipping operation — filters he did not make, shipped out of someone else's stock, at margins he describes as almost nothing. "Very low margin. Made almost no money early days doing that. But I was doing that to keep my risk low in order to prove out my business model." Only once he was convinced he had a system that repeatably brought in new customers did he spend money on manufacturing. His framing of the trade: you lose a couple of months getting to market, and in exchange your risk stays "effectively zero." The second thread is vertical knowledge. "We now have a thousand employees, but there is not a job in any of our factories or in any of our processes that I have not personally done." He ties that directly to systems design — having worked every station, he knows which pain points the systems have to solve, which is not something a manager infers from a dashboard. He has since started a second business on the same playbook, an HVAC service company called Filterbuy HVAC Solutions, with a friend he met at Goldman Sachs, and he writes cheques into other operators' boring businesses.
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.
Data credibility: 🗣 Founder-Reported, in a promotional video on the founder's own channel. David Heacock states "$22 million a month" about his own company in the opening seconds, and the video description repeats it ("My boring business makes about $22m per month"). There is no filing, dashboard, screenshot, third-party estimate or independent source anywhere in the material, and no breakdown of any kind — no margin, no profit, no customer count, no growth rate, no channel split. The auto-generated captions render the company name phonetically as "Filterby" and "filter by"; the description confirms the spelling Filterbuy and links filterbuy.com. The town is rendered "Tallaladega" in the captions and is Talladega, Alabama. The description also links filterbuyventures.com, so the video functions as a credential for an investment vehicle, which is worth weighing against an unverifiable revenue claim. The drop shipping origin, the thousand employees, the Salt Lake City plant and the $500-a-day compressor rental are all the founder's own recollections, unsupported by documents. The claim that he has personally done every job in every factory is not verifiable. Nothing in this entry about search channels, freight economics or the subscription-like nature of the category comes from the video — those are marked as our read.
Heacock had paying customers before he had a factory. The first Filterbuy was a drop shipping operation: filters he did not manufacture, shipped out of someone else's inventory, sold at margins he describes as almost nothing.
The first dollars arrived from resold filters at close to zero profit, and that was the design. Heacock treated early revenue as a measurement rather than an income: the number he watched was not what a sale earned, but what it cost to make a stranger buy, and whether that cost held as volume rose.
The turn came when he stopped paying somebody else to make the product. Once the acquisition system proved repeatable, Heacock invested in manufacturing, and a reseller earning almost nothing per order became a manufacturer who keeps the factory margin as well as the customer.
The drop shipping period was the stretch with nothing to show for it. Heacock describes making almost no money throughout, and he chose that on purpose, because the alternative was spending on a factory before knowing whether anyone would buy.
A residential air filter: a pleated panel of media in a cardboard frame that slides into a home's HVAC return and gets thrown away every one to three months. Filterbuy manufactures them and sells them direct. Heacock gives almost no product detail in the video. He names the company, states the revenue figure and the town it came from, then spends the rest of the run time on other people's businesses. There is nothing on sizes, MERV ratings, price points, channel mix, subscription versus one-off purchase, or the split between household and commercial buyers. What he does describe is the physical footprint: multiple factories, roughly a thousand employees, and a plant opening in Salt Lake City that stalled when an air compressor failed during the pandemic with a three-week lead time on a replacement. Our read: the product is best understood as a subscription the customer administers by hand. The replacement cycle is printed on the packaging, the size is fixed by the house and never changes, and repeating the purchase takes almost no thought. That makes the category unusually forgiving of a weak brand and unusually punishing of a weak supply chain — you win by being in stock, in the right size, delivered fast, at a price nobody bothers to shop a second time. Every structural decision here, from owning the factory to scattering plants near demand, falls out of that one fact.
$22 million a month, stated by the founder in the first fifteen seconds of his own video and repeated in the description underneath it. No breakdown comes with it — no margin, no average order value, no customer count, no channel split. Structurally, what can be said is that Filterbuy is a manufacturer that owns its own demand. The margin available to a business that both makes the product and acquires the customer is the sum of two margins usually divided between a factory and a retailer, and that combined pool is what funds advertising the pure-play filter brands cannot match. He runs a second business on the same thesis, Filterbuy HVAC Solutions, started with a friend from his Goldman Sachs days, and he backs other operators — he says he is flooded with boring businesses to look at and has written million-dollar cheques into a few. No figures are attached to either. Our read: the revenue number is the least useful thing in the story and the most likely reason the video exists. He links a ventures site in the description and sells no course, which makes the $22 million a credential rather than a disclosure. Treat it as context for the genuinely instructive part — drop shipping before manufacturing — rather than as an audited result.
There is no software stack worth describing, and that is the honest summary. At a thousand employees the cost structure is media, labour, freight and capital equipment, and the nearest thing to a technology decision is where the plants sit. The one operating detail the video does give is a good one. While opening the Salt Lake City plant, the air compressor broke during the pandemic and a new one was three weeks out — which in a plant means three weeks of nothing. He found a local rental company with a diesel unit, had it running within a couple of hours, and paid $500 a day for it. That is the real texture of manufacturing: fixed costs keep running whether the line does or not, and the gap between a lost afternoon and a lost month is knowing who nearby owns the machine you need. Nothing is disclosed about unit economics — cost to make a filter, retail price, contribution margin, acquisition cost, or how much of the $22 million survives to the bottom line. Our read: a solo operator should read this section as a warning rather than a template. The drop shipping phase is reproducible on a laptop this weekend. The current company is a capital-intensive manufacturer with a payroll of a thousand people, plants in several states and a freight bill that shapes every decision. Those are two different businesses, and only the first one is startable now.
Copy the sequencing before anything else. Sell the product before you can make it. Drop shipping, white-labelling or reselling at a margin near zero is not a failure mode here, it is the instrument — it answers the only question that matters, which is whether you can get strangers to buy repeatably, and it answers it without capital at risk. Heacock's generalisation of it is the cheapest experiment in the video: landing page, paid traffic, count the clicks to buy, and if people buy, go find somebody to deliver what you sold. Copy the category shape rather than the category. The transferable property of an air filter is not air — it is a consumable with a forced replacement cycle, a part number the customer can read off the old one, and a purchase nobody enjoys researching. Printer consumables, water filters, pool chemicals, vacuum bags, replacement blades and gaskets all share it. Copy working the process yourself before systematising it. His claim is that there is no job in any of his factories he has not personally done, and that this is the source of the systems, not a piece of founder theatre. What to change: do not copy the factory. Vertical integration was a decision he made from a position of proven demand and real cash, and a solo operator copying it at the start has inverted the exact sequencing that made the business work. Stay a reseller until the margin you are handing away is larger than the cost of making the thing yourself — that crossover, not ambition, is the trigger. Also change the channel assumption. He started before a marketplace or ad platform could turn off a business overnight. Build the reorder relationship — email, SMS, an account with the sizes already saved — from the first order, not once the channel starts squeezing.
Commodity competition is the structural one. A pleated filter in a cardboard frame has no technical moat, and the big-box retailers that sell them also sell private label versions of them. The defence is cost, availability and delivery speed, which are exactly the things capital buys — so an under-capitalised copy competes with the wrong weapons. Operating leverage cuts in both directions. A thousand employees and several plants are a wonderful thing while volume is rising and a very expensive thing when it is not. Heacock's own compressor story shows how little it takes to idle a plant that still has to make payroll. Input cost and trade policy. He argues elsewhere in the video that tariffs are pushing manufacturing back toward the United States, which is a tailwind for a domestic manufacturer and a headwind on whatever it buys from abroad. Media, board and freight prices move independently of what a customer will pay for a filter. Our read: the largest unstated risk is channel concentration. The video never says where the orders come from, and a direct-to-consumer commodity at this volume is almost always leaning on one search or marketplace channel whose rules it does not control. And the evidence. The $22 million a month is self-reported by the founder in a promotional video on his own channel, with a ventures link in the description. There is no filing, dashboard, third-party estimate or independent source anywhere in the material. Everything else about the business — margin, growth rate, profitability, how much of it is the filters — is undisclosed.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,403 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