Your Driving Will Be Monitored

Let us begin with the declaration, since everything after it depends on it. The product in this story is mine. For a few days, MoneySupermarket’s results page presented my motor policy as a telematics product, beneath a strapline assuring the shopper that his driving would be monitored. It was not a telematics product. Nobody’s driving was going to be monitored by anyone. The label was an error, and I have no grievance about the error itself: a system listing thousands of products will occasionally list one wrongly, and the people who run it have my sympathy. What interests me is what happened while the mistake was live.

Clicks fell to a small fraction of their normal daily level. Sales fell further still, to a figure a statistician would decline to call a rate.

The caveats first, because they are important. A few days is a few days. The numbers are small, there was no control group, and other things might have moved in the same window: a competitor’s pricing, the calendar, the ordinary weather of a comparison market. I would not submit this to a journal. But the collapse was far too large for noise to be a satisfying explanation, and it had a shape. Fewer people clicked, and of those who did, a smaller share bought. That is two leaks in one pipe, and they may not have the same cause. I have complained elsewhere in these pages that the industry models the second leak and merely infers the first; here, for once, I could see both of them running, and could account for neither.

To understand them, we need a shopper. Fortunately, we have one.

Regular readers will know Derek: nine years’ no-claims discount, a plush meerkat on the mantelpiece, and the settled conviction that he understands the insurance market better than the people who run it. Derek shops around. Derek reads the small print, or at least tells you he does. Derek is, by his own account, the best driver on the Kingston bypass, a claim the bin lorry was not invited to contest. Let us put Derek in front of my mislabelled listing and ask what the research says he will do. The research describes populations, not individuals, and Derek is fictional, so treat what follows as illustration. But it is illustration with footnotes.

I. The Word, or the Sentence?

The first suspect is the word telematics itself. It is a curious candidate, because a great many shoppers do not know what it means. A Consumer Intelligence survey of 742 UK drivers, run through its Viewsbank panel in October 2025, found that 38 per cent were unfamiliar with the term. The industry has been saying the word for two decades and more than a third of motorists still hear static.

Derek is not among the 38 per cent. Derek knows exactly what telematics is, because he explained it to everyone at Christmas, at length, with reference to a Formula One documentary. But for the large minority who hear static, the word probably did nothing on its own. The strapline did the work. “Your driving will be monitored” is admirably plain; it is also, from a conversion standpoint, a small act of commercial sabotage. It names a loss, specifies the lost thing, and offers nothing in return in the same breath. It has the cadence of a sign on a prison fence.

Which of the two did the damage? I do not know, and I cannot find anyone who does. I have found no published controlled test separating the word “telematics” from the monitoring disclosure in a UK shopping context. That is a statement about what I found, not about what exists. If such a test sits in some insurer’s drawer, I should very much like to read it.

II. The Price of Being Watched, or Derek Consults His Self-Image

Suppose it was the sentence. Why should five words cost so much?

The most direct evidence comes from Christian Biener, Martin Eling and Martin Lehmann, who in 2020 ran a framed behavioural experiment on exactly this question. Introducing digital monitoring into an insurance contract cut participants’ willingness to pay by somewhere between a quarter and a half of the expected loss. Grade it honestly: this is a laboratory with real monetary stakes, not a comparison site at eight in the evening. But the direction is unambiguous, and the size is not small. Translated into Derek: a watched policy is worth markedly less to him than an unwatched one, and he will expect the price to reflect it before he so much as hovers.

The authors draw on Armin Falk and Michael Kosfeld’s 2006 paper The Hidden Costs of Control, in which imposing control lowered the effort of those controlled, because control reads as distrust and distrust is reciprocated. That was a lab game, not a motor policy, and I apply it by analogy only. Still, the analogy has teeth. An insurer announcing that it will watch Derek is announcing that it does not take his word for it. And Derek’s word, in Derek’s opinion, is gilt-edged.

He is not alone in that opinion of himself. In Ola Svenson’s much-cited 1981 study, a large majority of his student drivers placed themselves in the safer half of the group, which is arithmetically impossible and psychologically universal. Derek would place himself in the top decile and regard the question as mildly impertinent. Tell such a man that his driving will be monitored and you have not described a feature. You have questioned a self-image, over his own breakfast, on his own laptop. Jack Brehm’s reactance theory, dating from 1966, predicts what follows when a valued freedom appears threatened: people move away from whatever is doing the threatening, and tend to think less of it on the way.

Derek does not click. Derek scrolls, with a small noise of contempt that the laptop, fortunately, does not record.

III. The Mood of the Road, or the Tower in the Middle

There is a larger thing in the background, and I suspect Derek feels it more keenly than he could explain, which for Derek is saying something.

In 1791 Jeremy Bentham published his design for the Panopticon: a ring of cells around a central tower, arranged so that the inmate could always be seen and never know whether he was being looked at. The genius of the scheme, as Bentham saw it, was economy. The watcher need hardly watch at all. The possibility of observation did the disciplining. Michel Foucault, in 1975, took the design as the emblem of a whole society that governs by visibility.

Bentham meant it for prisons. The modern motorist may be forgiven for suspecting it has been adapted for roads. Average-speed cameras measure Derek between points; number-plate recognition logs him in and out of towns; his car, if it is new enough, reports home. In 2023 the Mozilla Foundation reviewed the privacy practices of 25 car brands, found every one wanting, and called cars the worst product category it had ever reviewed. In 2024 the New York Times reported that General Motors had been passing drivers’ behaviour data to data brokers used by insurers, and some of those drivers discovered it only when their premiums rose. The US Federal Trade Commission moved against the practice the following January. And when London extended its Ultra Low Emission Zone in 2023, a self-styled band of “Blade Runners” took to cutting down the cameras.

Derek did not cut down any cameras. Derek wrote to his MP about them, twice, and has the replies filed. But one need not admire the vandalism, or the letters, to recognise the temperature they both record.

None of this proves that shoppers were thinking about Bentham as they scrolled past my listing. It would be a stretch to say so, and I shan’t. But grievances accumulate, and a man who already feels observed from the gantry, the kerb and the dashboard is unlikely to volunteer for one more lens, still less pay for the privilege. The strapline did not invent that resentment. It merely gave it a target with a “buy” button.

IV. The Wrong Aisle, or the Sorting Machine That Worked

There is a second explanation, less psychological and more architectural, and I find it the most interesting of all.

The large comparison sites have for years run separate black-box pages, and telematics in the UK is overwhelmingly a young driver’s product. GlobalData’s 2024 UK Insurance Consumer Survey found 14.4 per cent of 18 to 24 year olds held a pay-as-you-go or usage-based policy, against 2.8 per cent of those aged 25 and over. Consumer Intelligence, analysing 6,868 quotes from November 2025, found that 17 to 19 year olds faced a median penalty of £2,172 for choosing traditional cover over telematics. Ian Hughes, its chief executive, has called the drivers who leave the moment they can afford to “telematics refugees.”

The young, in other words, do not so much choose the tower as get priced into it. Then they serve their sentence and walk out. Britain’s largest insurer has lately asked Parliament to finish the job and make the box compulsory for them, a proposal I have dealt with elsewhere and shall not deal with again here, beyond observing that the market had already achieved the compulsion by price and did not require the statute.

Derek is too old to have been a refugee; the black box arrived after his youth did. But he knows perfectly well what the label means, because every comparison site he consults has taught him. Black box is for the nephew. Black box is for the lad who passed his test in the spring and whose insurance costs more than his car. A black-box badge on a listing in Derek’s results does not say “monitoring” so much as “you are in the wrong queue, sir.” He was not, in that moment, reacting to surveillance. He was sorting, exactly as years of market architecture have taught him to sort, and the sorting machine worked perfectly on a wrong input.

One detects a certain comedy in this. The comparison sites built a separate aisle for telematics; Derek learned to walk past it; the aisle then came to my product. Having spent an earlier instalment of this series complaining that the four great oracles rank for reasons of their own, I find it instructive to be undone by the one piece of sorting they perform faultlessly.

V. The Awkward Evidence, or Which Derek Turns Up

Here the tidy story begins to fray, and an honest essay should show where.

It would be convenient to claim that older, more affluent drivers are the most surveillance-averse segment in the market. Consumer Intelligence’s own 2016 Viewsbank survey, of 1,453 motorists, found the opposite on one measure: the older the driver, the less concerned about being watched. The firm’s explanation was that older drivers rate themselves safest and are probably right. A decade separates that survey from my few bad days, and the mood of the road has darkened since. Still, it gives pause.

And it describes a Derek we have all met. This is the Derek who says, with some satisfaction, “Let them watch. I’d score brilliantly.” The research, in short, predicts two Dereks: one who bristles at being watched, and one who would welcome the chance to prove himself, provided the proving were his idea. The puzzle is that both of them scrolled past. If the second Derek did not mind the watching, then his objection must have been to something else: the category, the implication, or the sheer oddity of a sensible product wearing somebody else’s badge.

Which brings us to the second leak. Suppose Derek, in a spirit of forensic curiosity, clicked anyway. Why would he then buy at a lower rate than usual? I can think of at least three answers and cannot choose between them. He may have arrived primed, so that every question on the form read as the first entry in a file. He may have found no mention of monitoring on our own pages, and a product whose listing and landing page disagree looks, reasonably enough, like a product that is lying somewhere. That is precisely the kind of discrepancy Derek lives to spot, and he would dine out on it until Easter. Or the clickers may simply have been a different crowd altogether: the curious, the young, the people who click on black-box listings for sport. Each story predicts something different. None has been tested.

I concluded the last instalment of this series with the observation that the label is never quite the thing, that the two are close enough that validation never complains, and that the gap between them is where the money sits. I had meant it as a complaint about vendors. It turns out to hold with equal force, and rather more expensively, when the label is attached by accident to a product of one’s own.

VI. What I Would Like to Know

So this is the request, and I mean it literally.

First, has anyone run the clean experiment? Same product, same price, same position, with and without the label; then the word without the sentence, and the sentence without the word. It would take a comparison site an afternoon to design and a fortnight to run. They hold the impressions, the clicks and the conversions across thousands of listings. My accident suggests an answer. Their data could give one.

Second, has this happened to you? If your product has been mislabelled on a comparison site, in either direction, I would like to compare notes. One accident is an anecdote. Several are a dataset.

Third, does the damage persist? Did the Dereks who saw my product labelled as monitored come back once the label had gone, or did they file it permanently under “not for me”? Derek, for one, does not forget. I suspect few firms have looked.

Fourth, and for researchers: is the mechanism reactance, category sorting, incoherence, or the slow-building resentment of people who feel increasingly observed and are running out of places to say no? The first three predict different things about who stops clicking and who stops buying. The fourth may be the soil in which all three grow. That is a tractable question, and the answer would be worth a great deal to an industry that keeps trying to sell monitored products to people who keep declining them.

In the meantime, for fellow providers, four lines for the beer mat. Look at your own listing daily, as a customer would see it, on every site that sells you. Track click-through and conversion separately, because the two leaks tell different stories. Keep dated screenshots of anything odd. And ask each comparison site for a log of changes to how your product is labelled, because you cannot diagnose what you did not know had happened.

Bentham’s tower worked because the inmates could not leave. The comparison site, whatever else it is, has an exit on every row, and Derek knows where every one of them is. The strapline promised that someone would be watching the driver. In the event it was Derek who was watching, and he drove on, under the steady button-eyed gaze of the one observer he has ever willingly let into the house.