A housing association employs people who handle complaints for a living. They know the Complaint Handling Code. They know the statutory deadlines. They know which language appears in a successful response and which triggers an escalation. They have case management systems, legal advisors on speed dial, and institutional memory stretching back years.
The tenant has a damp wall, a broken boiler, or a landlord who has stopped responding to calls. They have never made a formal housing complaint in their life.
The asymmetry is not an accident.
The gap between grievance and case
Tenants do not know the first stage of a formal complaint requires a response within a defined statutory window. They do not know Awaab's Law, in force since October 2025, places specific obligations on their landlord regarding hazards including damp and mould. They do not know the claims management firm offering to help them for free is planning to take a significant share of any compensation they receive.
They know they have a problem. They do not know how to turn it into a case.
This gap, between holding a legitimate grievance and being able to articulate it inside a system designed by the other party, is the accumulated result of decades of process design optimised for the provider's operational convenience rather than the tenant's ability to participate. I am not suggesting malice. I am suggesting inertia. When the people designing the complaint process are also the people responding to complaints, the outcome is predictable.
What the determinations show
The Housing Ombudsman publishes every formal determination it issues. The pattern is not subtle. Cases where tenants present clear, structured timelines with documented evidence tend to produce better outcomes. Cases where tenants have been bounced between phone calls, lost emails, and informal conversations produce outcomes hinging on whose account of events is more credible. Landlords almost always have better records.
This is not a failure of the Ombudsman process. The Ombudsman adjudicates on the evidence presented. The failure sits upstream, where tenants are expected to navigate a complex formal process with no training, no template, and no system to record what happened and when.
The claims management industry exists because this gap is commercially exploitable. An entire sector has grown up to fill the function, which tells you something important. The unmet need is real, and people will accept deeply unfavourable terms to have it met.
What changes when the tool sits on the other side of the table
I built a working response to this in under two weeks, in a series of command-line sessions with Claude Code.
A conversational interface walks the tenant through the details of their issue in plain English, identifies the relevant legislation, and explains what their landlord is required to do. The tenant describes the problem. The system produces a structured, correctly referenced complaint letter ready to submit. It uses the right language, cites the right provisions, and includes the right dates. Every stage of the formal process has a statutory timeline. The system tracks them in real time and surfaces them when they matter, not weeks later when frustration has built to the point of giving up.
The provider side mirrors this. A case management dashboard shows every active complaint, its current status, and which cases are approaching or past their deadline. Not adversarial. A provider who sees its own compliance position clearly is better placed to manage it.
Why the build itself is the point
The relevance of the two-week timeline is not the speed. It is what the speed makes economically viable.
A platform like this, built at conventional enterprise rates, would cost more than any housing sector charity or tenant advocacy organisation would accept. The arithmetic does not work. So the tool does not get built, the asymmetry remains, and the claims management industry continues to do what it does. The economics protect the status quo.
Built this way, the arithmetic inverts. Twenty-five years across architecture, data governance, product management, user research, and testing turns out to be a strong multiplier when collaborating with AI. The breadth I once worried marked me out as a generalist is the breadth letting me hold the whole problem in my head while AI does the typing. I am not a software developer in the traditional sense. I do not need to be. The cross-domain context is the multiplier.
This matters beyond one housing tool. There is a long tail of structurally important problems left unsolved because the cost of solving them exceeded the willingness of anyone to pay. The user base is too small, too poor, or too dispersed. The commercial case never closed. AI changes the floor of what is buildable, and the people best placed to spot which problems now become tractable are those with enough domain breadth to recognise them when they see them.
What the board should be reading into this
The wider point sits one level up.
The economics of building software for unprofitable user groups have shifted. The Consumer Standards regime, in force since April 2024, places explicit expectations on registered providers around complaint handling and tenant engagement. Awaab's Law adds statutory obligations around hazard response timelines which did not previously exist in this form.
Boards across the regulated public-facing sectors should be asking themselves a different question now. Not whether someone will eventually build the tool which closes the asymmetry between provider and end user. They will. The question is whether you would prefer to be ahead of the wave or behind it.