Curated by Machine, Defined by Whom? The Personalisation Dilemma Facing British Broadcasters

When Netflix introduced its recommendation engine to mainstream audiences, it did so with a singular purpose: keep subscribers watching for as long as possible. The algorithm cares nothing for cultural mission, public service obligations, or the kind of editorial coherence that has historically defined British broadcasting. And yet, as ITVX, BBC iPlayer, and Channel 4 streaming continue to compete for attention on the same devices, the pressure to emulate that model has become impossible to ignore.

The result is an industry caught between two distinct visions of itself — one rooted in the traditions of scheduled, editorially curated broadcasting, and another that aspires to the frictionless, data-optimised experience audiences increasingly expect from their streaming platforms.

What Personalisation Actually Means in a British Context

It is worth being precise about what is being discussed here. Personalisation in broadcasting is not a single technology but a spectrum of interventions — from simple watch-history recommendations to sophisticated behavioural profiling that adjusts not just what content is surfaced, but how it is presented, in what order, and with which thumbnail image.

Netflix's system, arguably the most mature in the consumer market, operates on the basis of billions of data points harvested across its global subscriber base. It is extraordinarily effective at predicting what a given viewer will engage with next. What it does not do — and was never designed to do — is reflect any particular editorial perspective, champion emerging voices, or balance entertainment with public interest content.

For the BBC, that distinction matters enormously. The corporation's royal charter explicitly mandates that it serves the public interest across all ages, communities, and regions of the United Kingdom. A pure recommendation engine, left unchecked, would almost certainly direct the majority of users towards the most broadly popular content — drama, sport, and established entertainment formats — at the expense of documentary, arts programming, and regional output. The algorithm, in other words, would gradually hollow out the very things that justify the licence fee.

The Quiet Compromise Already Under Way

None of the major British broadcasters would characterise their current approach as wholesale adoption of the Netflix model. Publicly, each maintains that editorial judgment remains central to how content is surfaced on their respective platforms. In practice, the picture is considerably more nuanced.

BBC iPlayer has introduced personalised rails — those horizontal rows of recommended content — that are increasingly shaped by individual viewing behaviour. The corporation is careful to ensure that certain categories of content, including news and regional programming, retain prominent placement regardless of an individual user's apparent preferences. But the tension between those editorial guardrails and the commercial logic of engagement metrics is visible to anyone who spends time on the platform.

ITVX faces a different set of pressures. Without the BBC's public service framework, the commercial broadcaster is more openly incentivised to maximise viewing hours and advertising revenue. Its investment in data infrastructure has been substantial, and its recommendation systems have grown noticeably more sophisticated over the past two years. Yet executives are equally aware that ITVX's identity — rooted in popular British drama, daytime programming, and live events — is itself a form of editorial curation that risks dilution if the algorithm is given too much latitude.

Channel 4's position is perhaps the most philosophically interesting. Its founding remit specifically requires it to be distinctive, innovative, and representative of minority interests. An algorithm optimised purely for engagement would, almost by definition, undermine that mission. Channel 4's digital team has spoken publicly about the challenge of building recommendation systems that actively surface challenging or unfamiliar content — a task that runs counter to the fundamental logic of most commercial personalisation tools.

The Identity Question No One Wants to Answer Directly

Beneath the technical discussion lies a more uncomfortable question: if a broadcaster allows an algorithm to determine what most viewers see most of the time, does it retain a meaningful editorial identity at all?

This is not an abstract concern. Editorial identity — the sense that a broadcaster stands for something, makes deliberate choices, and takes responsibility for its output — is part of what distinguishes British public broadcasting from a content aggregation platform. It is the reason a BBC commission carries different cultural weight than an equivalent production surfaced by an algorithm. Strip away that curatorial voice, and what remains is infrastructure.

Some within the industry argue that this framing is itself outdated. Audiences, they contend, no longer relate to broadcasters as institutions with coherent editorial personalities. They relate to individual programmes, talent, and genres. On this view, the editorial identity of a broadcaster is a legacy concept that serves industry insiders more than it serves viewers.

There is something to that argument. Viewing behaviour data consistently shows that most audiences arrive at content through search, social media, or word of mouth rather than through trust in a particular broadcaster's editorial judgment. The brand loyalty that once directed viewers to BBC Two on a Thursday evening has largely dissipated.

A Middle Ground That May Not Hold

The approach most British broadcasters are currently pursuing might be described as supervised personalisation — algorithmic recommendation operating within editorially defined constraints. Certain content is protected from purely algorithmic logic; certain values are encoded into the system's parameters. It is a compromise, and like most compromises, it satisfies neither side of the argument entirely.

The risk is that these editorial guardrails gradually erode under commercial pressure. As streaming competition intensifies and advertising revenues remain volatile, the temptation to optimise more aggressively for engagement — and to loosen the editorial constraints that limit that optimisation — will only grow.

British broadcasting has navigated similar tensions before. The introduction of multichannel television, the rise of catch-up services, and the shift to on-demand viewing all forced the industry to renegotiate the relationship between editorial purpose and audience behaviour. Each time, something was lost and something was preserved.

The current moment feels different in one important respect: the tools now available to broadcasters are capable of making editorial decisions at a scale and speed that no human commissioning team can match. That is precisely what makes them valuable — and precisely what makes them dangerous to the kind of broadcasting Britain has spent seven decades building.