The problem with he-said, she-said

When a review appears online, it becomes the only record most people ever see. It is published in seconds, ranked by an algorithm, and read as settled fact long before the business it describes has any chance to answer. Two accounts of the same afternoon then sit side by side in flat contradiction, and the reader is left to pick a side on instinct alone. That is not a fair test of anything. It is a coin toss wearing the costume of evidence.

Most disputes of this kind have no umpire, and so they resolve on tone — on who sounds more aggrieved, more wronged. This one has an umpire. The trick is simply to call it.

A witness that keeps its own records

There is a witness present on every outdoor job that has ever been done, and it holds a quality no human witness can match: it cannot be talked round after the event. What most people never realise is how meticulously that witness is recorded. The United Kingdom keeps one of the longest continuous weather records on Earth. The Central England Temperature series runs unbroken back to 1659. The England and Wales rainfall record stretches over two hundred and fifty years. The nation-wide rainfall series begins in 1836, the temperature series in 1884. And every year since 2017 the Met Office has drawn these threads together into a single formal audit — the State of the UK Climate, published in the International Journal of Climatology — peer-reviewed, public, and written by an institution with precisely no stake in the reputation of one small firm.

The conditions on any given working day, then, are not a matter of memory, and they are not a matter of opinion. They were measured, and they were written down, by people with no reason to take anyone's side.

What the record shows about the weather we work in

The record of recent British weather is not a quiet one. The six months from October 2023 to March 2024 were the wettest winter half-year in the England and Wales series in more than two hundred and fifty years, with much of the country seeing over 150 per cent of the average rainfall and repeated flooding. Six of the ten wettest such half-years in that entire quarter-millennium record have fallen in the twenty-first century alone. Nine named storms struck the UK in 2024. In late September that year, the Radcliffe Observatory in Oxford — a short distance, in climate terms, from this region — recorded its wettest two-day period in a near-two-hundred-year record dating to 1827, and its wettest calendar month for two hundred and fifty years. Storm Eunice in February 2022 brought gusts over one hundred miles an hour; that same July, the country passed 40°C for the first time on record; and in January 2025 Storm Eowyn became the most powerful windstorm to strike these islands in over a decade.

This is the backdrop against which every window is cleaned and every gutter cleared in this country. It is not a trade reaching for an excuse. It is the measured, published condition of the island we work on.

The line the law draws, not us

Exterior cleaning and maintenance happen at height and in the open, and working at height is governed by law. The Work at Height Regulations 2005 make it unlawful to work at height where weather conditions place the safety of the operative, or the property beneath, at risk. In practice, four conditions close a job down. Sustained wind above roughly twenty-five miles an hour, or gusts above thirty-five, make a water-fed pole impossible to hold accurately and unsafe to extend. Active rainfall makes streak-free exterior cleaning physically impossible. Frost and temperatures below about two degrees turn every surface into a slip hazard. And any day on which the Met Office issues a named-storm warning is an automatic stand-down.

So when a review claims that a service was carried out, or ought to have been, on a day the published record shows to have been unworkable, the claim has quietly left the territory of opinion. It has become a statement about physics and about law — and both of those can be checked.

Putting a number on the days the sky closes the job

It is one thing to say the weather interrupts the work; it is another to measure how often. So we did. Taking the legal standard above, we scored every month from January 2020 to March 2026 — sixty-three months in all — against it. The answer, averaged across the whole period, came out at roughly 53 per cent. That figure is a demonstrative model, not a surveyor's certificate — your own street in a given month would need its own reading — but the order of it is not in dispute. A little over half. Which means a mobile trade working at height in this region can expect to lose somewhere between one hundred and sixty and one hundred and eighty working days a year to weather alone. That is not a failing. That is the climate, counted.

Of those sixty-three months, twelve were severe, twenty-seven poor, and only four reached the standard we would call excellent. Twenty-seven named storms crossed the region in that time. And the disruption does not fall evenly: February 2022 delivered three named storms in eight days, while four months later July offered the best working conditions in the entire audit. Excellent months sit directly beside catastrophic ones, and no diary can be built that the weather will honour.

A wet day is not a lazy day

The reason we stand a job down in the wrong weather is not comfort; it is the standard of the result and the safety of the person delivering it. Water on glass in active rain dries to streaks, and the honest outcome is a job that must be done again in full. Coatings and sealants applied to a damp or freezing surface will not cure and will not bond; they fail quietly, months later, long after the invoice is forgotten. Read properly, a rescheduled visit is not the mark of an unreliable business. It is the mark of one that would rather be right than merely convenient.

From opinion to evidence: the method

For every disputed date, we retrieve the conditions for that day, in this region, from the Met Office archive, and we publish them beside the review. Where a review alleges that a service was not delivered, or was cancelled without cause, and the date falls inside a named-storm window or a month the record classes as severe, the conditions on that date answer the claim directly. The standard is simple: if the weather did not permit the work to be done safely and to a professional standard, then the assertion that it should have been is refuted by atmospheric fact, not by our argument. It is not clever. It is meant to be checkable.

The record cuts both ways — which is the point

It would be convenient to present the weather as a shield that only ever defends this business. It is not, and pretending otherwise would forfeit the one quality that makes it worth anything: its indifference. On a calm, dry, workable day, the record offers a bad review no meteorological defence whatsoever, and the dispute must then be settled on booking records, photographs and correspondence — exactly as it should be. A witness that agreed with us on every occasion would be worthless. A witness that answers to no one, and will on a fine day say so plainly against us, is worth everything precisely because it cannot be bought.

Why we lead with the weather

We open our whole reputation with the weather for a single reason. It is the one part of this story that no one — not a disgruntled customer, not a competitor, not this business, not a platform, not an algorithm — can go back and edit. Everything else in a reputation can be argued or drowned out. The sky, on a given day, over a given town, cannot. It was what it was, it was measured, and it is published for anyone on Earth to read.

That is the end of opinion. It is the beginning of atmospheric fact.

Sources and further reading. Met Office — M. Kendon and colleagues, State of the UK Climate (annual, 2017–2024), International Journal of Climatology, Royal Meteorological Society; Met Office HadUK-Grid dataset; Met Office UK Storm Centre named-storm archive; the Work at Height Regulations 2005. Compiled in the Shining Windows WAHR Weather Audit, January 2020 – March 2026. Workability figures are Demonstrative Models; records, storms and temperatures are Met Office primary.