In 1849 the reliable way to make money in California was not gold. It was selling something to the people looking for gold. Levi Strauss sold work trousers. Sam Brannan bought every shovel and pan in San Francisco and sold them at a markup to men sprinting past his store. Most of the miners went home with less than they arrived with. The suppliers stayed. That pattern has repeated in every rush since, and 2026 has several running at once.
The 1849 lesson, briefly
The gold rush produced a small number of spectacular outcomes and an enormous number of people who broke even at best. The suppliers had a different distribution entirely: lower ceiling, dramatically higher floor, and revenue that did not depend on any individual claim panning out. They were selling into the enthusiasm rather than betting on it.
The lesson is not that speculation never works. It is that the shovel-seller gets paid whether the miner strikes gold or not, and gets paid again next week when the next miner arrives. Same rush, completely different risk profile.
The 2026 rushes and their shovel plays
Three are running loudly right now. In each one, the speculative layer gets the attention and the service layer gets the recurring revenue.
- The AI rush. Everyone is building a model or an app on top of one. The shovels are the physical layer underneath: data centers, the power to run them, the cooling, the fiber, the electricians and HVAC contractors who service the buildings. None of that is glamorous and all of it gets paid regardless of which model wins.
- The creator rush. Millions of people are trying to build an audience. The shovels are the tools and the services: editing, thumbnails, scheduling software, course platforms, the agencies that clip long video into short. The creator economy has always paid its suppliers more reliably than it pays its median creator.
- The crypto rush. The speculative layer is loud and cyclical. The shovels are exchanges, custody, compliance, tax software, and the payment rails — infrastructure that collects a fee on activity in both directions of the market.
No stock picks here and none intended. The point is structural, not a recommendation: in each case the boring layer sells to everyone in the rush, and it does not need the rush to end well.
Picture the machines paying you while you sleep
That’s the real promise of vending — income that doesn’t cost you your time, and a life on your own terms. VendBuddy turns this guide into a step-by-step plan so you actually build it instead of just reading about it. Start free today.
Start building free →Why the boring layer compounds while the speculators churn
Three reasons, and they are the same three every time.
Revenue does not depend on the outcome. A shovel-seller is paid on activity, not on success. That decouples income from the thing everyone else is gambling on.
Churn is the customer, not the supplier. Rushes have enormous participant turnover. When one miner quits, another arrives and buys the same equipment. The supplier sees a steady stream where the participants see a lottery.
Boring is a moat. Nobody makes a video about installing a card reader. Attention flows to the exciting layer, which means the unexciting layer stays under-competed for far longer than its economics would justify. Difficulty and tedium are barriers, and barriers are the only thing that keeps a business from being arbitraged flat.
Screenshots versus spreadsheets
There is an entire genre of internet content built on losing money in public — the account balance screenshot, the liquidation post, the caption about buying the dip on the way to zero. It is genuinely funny and it entertains millions of people, and we are not above enjoying it.
It is also worth noticing what the genre is made of: an enormous volume of individually small losses, posted by people who were early to the excitement and late to the exit. The chaos is the joke, not any particular person in it.
Meanwhile the people selling into that same rush post nothing. There is no viral format for a spreadsheet with a slightly better gross margin than last quarter. The shovel-sellers are not more clever than the speculators — they are just holding a different instrument, one where a bad month is a bad month rather than a screenshot.
Unattended retail is the picks-and-shovels of everyday places
Here is the bridge, and it is smaller than the historical framing implies. You do not need a gold rush to sell shovels. You need a place where a lot of people are stuck doing something for a while.
Every gym, warehouse, apartment lobby, laundromat, medical building, and manufacturing floor is a small rush of foot traffic. People are there for their own reasons, on their own schedule, for a predictable amount of time. They get thirsty, hungry, bored, or short of something. The operator does not participate in why they are there — the operator sells the shovels: a drink, a snack, a charging cable, thirty seconds of convenience at a moment when convenience is worth paying for.
The economics behave like the historical pattern. Revenue is tied to foot traffic rather than to any individual customer succeeding at anything. Turnover in the building is fine, because the next shift also gets thirsty. And the whole category is unglamorous enough that most locations have simply never been asked. A typical dialed-in machine in an ordinary captive placement nets somewhere in the $150 to $400 a month range — a boring number, repeated across a route, which is exactly the shape of a shovel business.
The nonconventional shovels
Snacks and drinks are the default, not the boundary. Unattended retail has quietly expanded into categories that behave much more like the shovel play than the classic snack machine does, and they map onto rushes of their own.
- Bitcoin ATMs. A physical on-ramp that collects a fee on the transaction regardless of which way the price goes afterward. Higher capital and compliance requirements than a snack machine, and a genuinely different business — the placement economics are in the ATM placement guide.
- Trading card machines. The collectibles rush has its own shovel layer, and a card machine in a comic shop or game store is squarely in it. Supply volatility is the real operational risk. Full breakdown in Pokemon and TCG card vending machines.
- Claw and prize machines. Entertainment revenue in venues where people are already waiting, with unit economics that look nothing like snack vending. Covered in claw machine profit and locations.
- Card readers and payment hardware. The least visible shovel of all: cashless payment is now the majority of transaction volume in modern machines, which makes the reader less an accessory than a prerequisite.
All of these now sit inside the same location-finding workflow, which is the practical version of this whole thesis: pick the shovel, then find the rush it fits. The ranked comparison across every model is in alternative vending machine businesses, ranked.
How to actually start
The honest sequence has three steps and none of them are exciting.
One: decide whether the shovel business fits you at all. It is unglamorous work with a slow ramp, and the reasons not to do it are worth reading before the reasons to do it — both the balanced version in is vending a good business and the deliberately discouraging version in do not start a vending machine business.
Two: find the rush before you buy the shovel. Secure the location first. A machine in your garage is a depreciating asset with a storage cost. The screening criteria are in how to judge a vending location.
Three: match the hardware to the place. The right machine for a night-shift warehouse is not the right machine for a game store, and buying the wrong one is the expensive version of this mistake.
Two minutes of questions about your capital, your time, and the places you already have access to — and an honest answer about which unattended retail model fits, including when none of them do.
Take the readiness quiz →FAQ
What does picks and shovels mean in business?
It means selling the tools, supplies, or services that everyone chasing an opportunity needs, rather than chasing the opportunity yourself. The phrase comes from the 1849 California gold rush, where suppliers of equipment and clothing had far more reliable outcomes than the median miner. The structural advantage is that revenue depends on activity in the market rather than on any individual participant succeeding.
What are boring businesses that make money?
The pattern to look for is recurring demand, low glamour, and a service layer underneath something more exciting. Unattended retail, equipment servicing, logistics, trades, and infrastructure maintenance all fit. In every case the appeal is the same: the work is unappealing enough that competition is thinner than the economics would justify.
Is vending a picks-and-shovels business?
Structurally, yes. A vending operator sells convenience to people who are already somewhere for their own reasons, and gets paid on foot traffic rather than on any customer outcome. Turnover in the building does not hurt revenue, because the next shift has the same needs. Typical net per machine in an ordinary captive placement runs in the range of a few hundred dollars a month, which is a boring number by design.
Related reading: every unattended retail model, ranked, the macro case in vending versus rental real estate on cash-on-cash, and the filter post do not start a vending machine business. When you know which shovel you want, the free Machine Finder matches it to the property type you have access to.