The Commodity Generic Play (ONCE)
Find commodity software still charging luxury subscription prices, then sell a beautiful generic once
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 90%
37signals' ONCE line is a repeatable market-selection method: scan business software for categories that have become commodities (dozens of near-identical alternatives, no real price differentiation) yet still charge recurring luxury prices. Build the 80/20 core of that category — beautifully, simply, with two people in about three months — and sell it as a one-time download the customer installs and owns. No hosting, no marginal cost, no subscription.
Origin
Jason Fried and David Heinemeier Hansson's ONCE product line at 37signals, launching with Campfire (a group chat tool 37signals originally shipped in 2006, years before Slack). The framing borrows explicitly from generic goods in other industries — generic peanut butter, generic pencils.
Core principles
- 01Commoditized category + luxury recurring price = an opportunity, in software as in every other industry.
- 02Subscriptions ask you to pay again every month for the same thing you had last month.
- 03Get the core 80/20 right at a very high level of quality; explicitly do not match feature-for-feature.
- 04One-time software has essentially zero marginal cost to the maker — no hosting, no ops.
- 05An alternative does not have to replace the incumbent; it can sit alongside it.
- 06If it isn't an alternative — a different approach, price, or conception — it doesn't need to exist.
How to run it
- 1
Scan for commodities charging luxury prices
Look across business software for categories with fifty near-identical alternatives and no meaningful price differentiation, where customers are still paying tens, hundreds, or thousands of dollars a month, month after month.
Pro tip Group chat was the first target: Slack, Teams and a handful of others, all expensive, all monthly, all functionally similar.
- 2
Define the 80/20 core, not the feature set
Identify the ~90% of things people actually do every day in the category and build only those — exceptionally well, simply, easily. Explicitly do not do everything the incumbents do.
Pro tip Minimize the interface surface: the first ONCE product aims for roughly 30 words of UI copy so it works globally without localization packs.
- 3
Build it small: two people, ~three months
Constrain the build to a two-person team over roughly three months. The scope discipline is what keeps the generic a generic instead of a bloated clone.
- 4
Price it as a no-brainer, one time
Set a single one-time price low enough that buying it is obviously worth it — 37signals targeted under $1,000 for Campfire. Customers download it, install it on their own server, and own it. No recurring fees.
Pro tip Ship the source code too. Buyers can read, modify and tweak it (no resale). That is something subscription software structurally cannot offer.
- 5
Sell it as an addition, not just a replacement
Position the product for the jobs a subscription cannot economically serve: a backup for when the incumbent goes down, an air-gapped instance for an executive team or a secure facility, an experiment for one group. Nobody pays monthly for a backup; a one-time price makes it an insurance policy.
Pro tip It is a potential trojan horse — the groups running the cheap alternative may make identical progress, which raises the question of why you are paying so much elsewhere.
- 6
Version, don't subscribe
Ship all 1.x updates free and let customers choose whether to take them or skip a version. Only a genuine 2.0 that took another six-plus months of work justifies a repurchase, likely at a discounted upgrade rate.
In the wild
37signals built Campfire, a group chat tool, in 2006 — years before Slack — and it never caught on. Chat has since become a commodity with expensive monthly incumbents, so they rebuilt it as the first ONCE product: pay once, install on your own server, get 90% of what you do daily plus the source code.
→ A sub-$1,000 one-time product with zero marginal cost to 37signals, self-service (no sales team), rolled out to a first batch of customers in mid-December 2023.
Years earlier, the White House approached 37signals wanting to use Basecamp but needed to install it themselves — they could not host data on someone else's servers. SaaS made that impossible.
→ Fried cites it as the tiny-but-real market that installable, non-internet-facing software unlocks: a side effect, not the target, but proof the model serves buyers SaaS structurally cannot.
Common mistakes
Trying to match the incumbent's feature list
The whole economic advantage comes from doing 90% of the daily job extremely well with two people in three months. Feature parity destroys the cost structure and turns your generic back into a competitor you cannot afford to be.
Attacking a category that isn't a commodity yet
The play only works when there are already many near-identical alternatives at undifferentiated prices. Without that, there is no luxury-price anomaly to arbitrage and the buyer has no reason to accept a simpler tool.
Framing it as a direct replacement
Insisting customers rip out Slack narrows the market enormously. Selling it as a backup, an air-gapped instance, or a per-team experiment lowers the buying bar to near zero — and the replacement can happen later.
Is it for you?
Best for
Small software teams with existing margin who want to enter a crowded category without competing on features or burning cash on acquisition
Not ideal for
Categories that are still differentiating fast, products that fundamentally require hosted infrastructure, or teams that need predictable recurring revenue from day one
From the transcript
“so we're we're out there looking for Commodities that are still charging Lar prices that we can create really beautifully tight simple essential generics of…”
“the 8020 we kind of get that right at a high degree of quality and we can build these in typically three months each with…”
“but uh we're trying to find a price point where it's kind of becomes a no-brainer like you know what we should we should buy…”
“is that these don't necessarily need to replace things they can be in addition to like you might still run slack at your company but…”
From the episode
Jason Fried challenges your thinking on fundraising, goals, growth, and more