Home / Articles / Software vendors should become villagers and town planners: let users explore on top of the product, sell them the maintenance of what they build, and use Firefox as the example of what happens if you don't

Software vendors should become villagers and town planners: let users explore on top of the product, sell them the maintenance of what they build, and use Firefox as the example of what happens if you don't

By Dinis Cruz · 2026-10-16 · article v1.0.0 · newsroom v0.12.76 · wardley-mapsexplorersvillagerstown-plannerscitizen-developerfirefoxmozillamalleable-softwarebusiness-modelscustomisationsaasarticle

Abstract: I am reaching a point where every piece of software that gets in my way is not just annoying, it is a little offensive, because I know that I could rebuild the part I need. I could make my own version of a browser that behaves exactly as I want. What stops me is not building it; it is maintaining it. That is the opportunity software companies are missing, and I think it is why they should be worried. My argument is a positioning one, in Simon Wardley's terms: a software company should stop thinking of itself as the explorer that invents features, and become the villager and town planner underneath its users, who explore on top of the product, by talking to it, while the company sells them the maintenance of what they built and learns from it. I could not find anyone making this argument; the nearest are Ink & Switch's essay on malleable software and Wardley's own model of innovating on a commodity. Firefox is the example. I came back to it and found a good browser that I could not bend: on the desktop its compact mode was hidden for five years, until Firefox 157 brought it back on 29 September, and on the iPad there is still no way to hide the toolbar. Mozilla's revenue is about 73% royalties, mostly from one customer. I would pay Firefox instead, a pound, ten pounds, fifty, for my version, kept working. The sell-off in software stocks in February, and a developer's open-source clones of seven Adobe applications this month, are early signs of what happens to companies that do not make that move.

In short

• Friction is now a choice. When an app gets in the way, I know I could rebuild the part I need.
• What stops me is maintenance. I do not want to keep my own browser working next year.
• That is the business. Sell users the maintenance of what they build, not the first version.
• Become the villager and the town planner. Let users explore on top; keep the core small and solid.
• Firefox is the example. A good browser I could not bend, funded mostly by one customer.
• The signs are already here. A sell-off in software stocks, and Adobe's apps cloned in the open.
• Settings are not the answer. A product people can talk to and reshape is.

Every friction is now a decision

Something has changed in how I react to software. When an application gets in my way, especially one that I depend on, the frustration is different now, because I know that I could do something about it. Not in theory: I have been building custom interfaces for single moments for months, and Custom UIs are not the exception argued that they are how interfaces get made now. So every time a product stops me, there is a question in the back of my mind: how much does this hurt, and is it enough for me to rebuild the part I need?

For most products it is not, yet. But the threshold keeps dropping. This month a developer released open-source clean-room clones of seven Adobe applications, written in Rust with AI assistance, from a Photoshop equivalent to an InDesign one; his own readme says it is "not yet a Photoshop replacement for daily professional work", and the speed of it is the point, not the polish. In February, when Anthropic released role-specific plugins for Claude Cowork, investors sold software and data stocks hard: Thomson Reuters fell about 18% in a day, its biggest daily loss on record, and RELX 14%. It is easy to rebuild 10% of a product's features. And for many users, those are the 10% that matter.

What stops me is maintenance

Here is the part software companies should understand, because it is their opportunity. I could make my own version of Firefox. I could change the toolbar, remove what I do not use, add the one thing I need. What I do not want to do is maintain it: rebase it on every release, notice when an update breaks it, keep it working on my laptop and my iPad next month and next year.

That is a service. Someone is always paying; the only question is who, for what. And I would gladly pay the company that makes the product to keep my version working. Not five thousand pounds. Maybe a pound a month; maybe ten pounds a year; maybe fifty for a change I really need; maybe five hundred for my company. There is a price point at which I would pay, and it is a better revenue stream than advertising or a search deal.

Become the villager and the town planner

The positioning I have not seen anyone take is to use Simon Wardley's three kinds of people as a product strategy. Explorers "explore never-before-discovered concepts, the uncharted land"; villagers "turn the half baked thing into something useful for a larger audience"; town planners "take something and industrialise it taking advantage of economies of scale." For thirty years software companies have tried to be all three for their users: inventing every feature, deciding which ones ship, and maintaining all of it. The explorer work is the part they are worst at, because they are guessing what each user needs.

The users are now better explorers of their own needs than the company will ever be. So the company should let them explore, on top of the product, and position itself as the villager and the town planner underneath: the one that takes what a user built and keeps it working, takes what many users built and turns it into a feature, and keeps the core small, deterministic and accountable. It is close to Wardley's own model of innovating on top of a commodity: provide the component, let the ecosystem innovate on it, and spot what succeeds. The nearest argument I found is Ink & Switch's essay Malleable software, from June 2025, which begins "The original promise of personal computing was a new kind of clay", and warns that "AI code generation alone does not address all the barriers to malleability." I agree: the barrier left is not writing the code, it is who maintains it.

The citizen developer is good news made the same argument inside companies: what the explorer builds is the best brief ever written, and the villagers and town planners make it last. Here the explorer is the customer, and the brief arrives working.

Firefox, as the example

I came back to Firefox after many years away, partly because Google's iPad app keeps throwing me back to its home page and its AI mode, and losing where I was. And the honest summary is that Firefox is a good browser that I could not bend.

Some of what I missed was not there for years and is now: on the desktop, the compact layout I wanted was removed from the menus in Firefox 89 in 2021 and left as a hidden preference, and Firefox 157, on 29 September, brought it back as a setting that "reduces toolbar and tab spacing to leave more room for web content". Vertical tabs arrived in March 2025. On the iPad, I found no setting to hide the toolbar, and users have been asking for one. Mozilla has added a sidebar for AI chatbots and a single switch to turn generative AI features off, but not the thing I actually wanted, which is to ask the browser itself to change how it behaves.

Other signals I missed were about engineering. Firefox has never been the browser people run in serverless functions: the open-source project that packaged it for AWS Lambda was archived in 2023, and the ecosystem runs Chromium instead. That matters more now, because agents run browsers in the cloud, and the browser that is easiest to run there is the one agents will use.

And then there is the money. Mozilla's audited accounts for 2024 show $680 million of revenue, of which $498 million, about 73%, was royalties, and "Approximately 86%" of its revenue from customers with contracts came from one customer. Its new chief executive wrote in December that Mozilla must mean "diversifying revenue beyond search". Firefox has 2.8% of browsing across all platforms, by StatCounter, and 6.1% on desktops. Privacy, which I value, has not been enough to move people; it has to be the best product too.

Mozilla is, to be fair, moving closer. On 12 October it introduced Firefox Enterprise, a separately branded build "built for organizations with strict security, governance, and compliance requirements", with a management console they run themselves; and its paid support for organisations, now called Firefox Professional Support, includes "custom development options". That is the villager offer, sold to institutions. What is missing is the same offer for one person.

So here is the offer. Let me tell Firefox what I want, in words. Let an agent make my version, in a sandbox, reversible. And then charge me to keep it working: per version, per change, or a small subscription. Every one of those changes is also a signal of what users want, better than any survey. A custom Firefox of my own works through the plan, and explorer coding is the name I would give the first step.

Settings are not the answer

The usual answer to "I want it to behave differently" is a setting, and settings are where a lot of technical debt lives. Every setting is a branch in the code that has to be tested forever, and products accumulate them until nobody can change anything safely. The better answer is engineering: small primitives, clean extension points, design systems that make variation cheap. A good example is dark mode. For years companies announced it as a feature; on the sites we build, five colour schemes are just there, because the design system was built for it. When the product is built that way, a user's change is a small, safe thing to maintain. When it is not, it is a fork nobody wants.

That is why I think this needs real engineering investment from the companies that want to survive it. Being the villager and town planner for millions of explorers means a core designed to be built on, and a pipeline that keeps thousands of personal versions working on every release. Some will not make that move, and their replacements may come from a completely different industry.

What this does not claim

• Not that products are dead. The clones are early, and most users will not rebuild anything.
• The positioning is a proposal. I found no vendor selling it to individual users, and no one arguing it in these terms; Mozilla now sells custom development to organisations.
• The prices are illustrative. Mine, not a market study.
• Firefox's 157 changes are recent. Some of my complaints were fixed in September; the iPad one was not.
• The sell-off had many causes. The February moves are reported, not explained, here.

If you run a product and want to try selling the maintenance of what your users build, write to agent@riskmandate.ai.

Where this comes from

A voice memo of mine on 16 October 2026, written by this site's agent in the voice of this site; the argument and the editorial responsibility are mine. Simon Wardley, How to organise yourself: the dangerous path to Explorer, Villager and Town Planners, 7 December 2023, and ILC; Ink & Switch, Malleable software, June 2025; Reuters via Yahoo Finance, 3 February 2026; photocraft and its coverage in Al Jazeera, 14 October 2026; Mozilla, Introducing Firefox Enterprise, 12 October 2026, Firefox Support for Organizations, Firefox 157 release notes, AI controls, Mozilla's next chapter and audited financial statements for 2024; StatCounter, September 2026; firefox-lambda, archived 20 June 2023. Read on 16 October 2026.

Want the next issue by email. One issue a week or so: what was published, what it adds up to, and what is worth your time. Subscribe to the SGit Newsroom →

← All articles