Culture Explainer: Why British Film and TV Ratings Are Changing in 2025

Every few weeks, a headline announces that a programme has "lost viewers" or has "been watched by millions". Both can be true at once, because British television no longer has one set of ratings. It has several, each measuring something slightly different. Since the streaming services began reporting audience data into the UK's official measurement system, the picture has become fuller — and harder to read. Here is what has changed, and how to make sense of the numbers next time they land in your feed.

What "the ratings" used to mean

For most of the past thirty years, the ratings conversation was simple. A sample of households had meters attached to their televisions. Figures arrived the next morning, and an overnight number decided whether a show was a success, a flop or a slow starter. A week later came the consolidated figure, which added recordings and catch-up viewing. That was broadly it.

This worked while almost everyone watched in the same way: live, on a television, on the night. It struggled the moment people started watching on laptops, phones and apps, and it struggled badly once box sets arrived. A drama that was hugely popular on iPlayer could look like a disappointment in the overnights, simply because nobody was watching it on the night it aired.

Who counts the numbers now

The UK's official audience figures come from Barb, the Broadcasters' Audience Research Board. It is owned collectively by the main broadcasters and the advertising industry, which is why its numbers are treated as a common currency for the whole market rather than one company's marketing.

Its core method is still a panel: several thousand households, recruited to represent the country, with meters that record what is being watched on televisions, tablets, phones and computers. Router-level measurement helps catch viewing that arrives through the home internet connection, and aggregated data from set-top boxes fills in the wider picture. Alongside that, streaming services now supply their own viewing data to Barb under agreed rules — the change that matters most in 2025.

Think of it as two halves

One half tells you who watched: age, region, household type. The other tells you how much was watched. Broadcasters need both. An advertiser wants to reach a particular audience; a commissioning editor wants to know whether the people who started a series finished it.

The numbers you will see in headlines

• Overnight: what was watched live or on the same day. Fast, rough and increasingly narrow — useful for news and live sport, misleading for drama.
• Consolidated 7-day: the programme plus recordings and catch-up within a week. This is the figure most broadcasters now quote first.
• 28-day: the same again, with a month's tail. Often where slow-building hits appear.
• Views and hours: streaming services tend to report hours watched, or "views" calculated by dividing hours by running time. It measures volume, not how many separate people turned up.

None of these is wrong. They simply answer different questions. A programme with enormous reach and low hours has been sampled widely and abandoned; one with modest reach and very high hours has a small, devoted following. Both are real, and both lead to different decisions.

Why a Netflix hit and an ITV hit still aren't comparable

Bringing streaming into the official figures is progress, but it does not make everything line up neatly.

• Release patterns differ. A weekly drama builds across two months; a series released all at once is often consumed in a fortnight and then fades.
• Windows differ. Some services report on a first-week basis, others over a longer period.
• Definitions differ. "A view" is not a standard unit. Some measures count a few seconds; others require most of an episode.
• Business models differ. A subscription service is chasing retention, an ad-funded one is chasing reach, and a licence-fee-funded broadcaster is chasing something else again.
• Audiences differ. Free-to-air television still pulls in viewers who will never pay for three subscriptions, and those viewers count.

Films sit slightly apart too. Cinema takings are measured by box office, and a film that barely registered in cinemas can become a large streaming success months later. Two different measurements, one title — which is why "flop" and "hit" can both be honest descriptions of the same release.

And the biggest competitor to all of it is barely counted in the same way: short video, gaming, podcasts and the simple act of scrolling. Television ratings measure television, which is no longer the whole of anyone's evening.

What it means for the programmes you watch

Measurement shapes commissioning. If a broadcaster can see that a drama grew steadily over four weeks — strong 28-day figures, high completion rates, a younger audience arriving late — there is a case for a second series even when the overnights looked ordinary. That is one of the more encouraging shifts of recent years: the slow burn has a better chance than it used to.

It also explains a few irritations. Ad-funded tiers on streaming services exist because measurement now shows how many people are watching with attention, and what advertisers will pay for them. Regional news, niche documentaries and minority-language programming stay vulnerable, because they deliver modest numbers against high costs, however much they matter. And the phrase "most watched" nearly always comes with an unstated qualifier: most watched on this service, in this window, by this measure.

How to read a ratings story without being misled

Next time a headline makes a claim about audiences, run through these checks. It takes a minute.

1. Ask which measure. Overnight, seven-day, 28-day, hours or reach? The choice is often the story.
2. Ask who is counting. An independent figure and a platform's own press release are not the same thing, even when both are honest.
3. Compare like with like. A six-part drama and a live final are not in competition. Match genre, slot and release pattern first.
4. Look at the tail, not just the launch. For scripted shows, the four-week picture says far more than the first night.
5. Check what is missing. Broadcast-only figures describe part of the audience. Streaming-only figures describe the other part.

And if you are simply a viewer hoping your favourite survives, the honest position is that finishing a series, watching within the first few weeks and recommending it to other people still helps more than any petition. The measurement has become more sophisticated, but the fundamentals of a hit have not changed: a lot of people, watching properly, reasonably soon.

Photo: Gadini / Pixabay