There was a time when reputation was something that accumulated slowly, through years of consistent behaviour witnessed by a relatively stable community. A shopkeeper's good name was built on decades of fair dealing with customers who knew them personally; a professional's standing rested on the judgements of colleagues who had observed their work at close quarters over a long career. Reputation, in other words, was local, gradual, and grounded in direct experience. The digital revolution has upended every one of these characteristics, creating a system of reputation that is global, instantaneous, and mediated almost entirely by strangers.
The most visible manifestation of this transformation is the online review. Platforms from Amazon to TripAdvisor, from Google Maps to Trustpilot, have placed an extraordinary degree of power in the hands of consumers, who can now broadcast their satisfaction or displeasure to a potentially limitless audience with a few keystrokes. For businesses, the consequences are profound. A single negative review, if it gains traction, can inflict damage that would once have required a sustained campaign of word-of-mouth criticism. Conversely, a stream of positive reviews can elevate an obscure enterprise to prominence with a speed that traditional advertising could never have achieved. The asymmetry is striking: it takes far less effort to destroy a reputation online than to build one.
The implications extend well beyond commerce. In the professional sphere, platforms such as LinkedIn and Glassdoor have made individual reputations searchable, quantifiable, and permanent in ways that earlier generations would have found deeply unsettling. A remark made in haste, a disagreement taken out of context, or an accusation that may or may not be justified can follow an individual indefinitely, surfacing in every future employer's search results. The legal scholar Viktor Mayer-Schönberger has argued that the internet's inability to forget constitutes a fundamental threat to the human capacity for personal reinvention — the possibility of leaving past mistakes behind and starting afresh, which has historically been one of the most valued features of open societies.
The reliability of online reputation systems is itself a matter of growing concern. Research has demonstrated that review platforms are vulnerable to manipulation on a significant scale. Businesses routinely solicit fake positive reviews, sometimes through specialised agencies that operate in a grey area between marketing and fraud. Competitors post fabricated negative reviews to damage rivals. Algorithms designed to detect such manipulation are engaged in a perpetual arms race with those seeking to circumvent them, and the platforms themselves face a conflict of interest: their commercial model depends on the perception that reviews are trustworthy, which creates an incentive to understate the extent of the problem rather than confront it openly.
There is also a subtler distortion at work. Online reviews are not a representative sample of customer experience. People are disproportionately motivated to leave reviews when they are either extremely satisfied or extremely dissatisfied, producing a distribution that clusters at the extremes and underrepresents the moderate middle. The resulting picture is systematically skewed, yet it is treated by most consumers — and by the algorithms that aggregate and rank businesses — as though it were an accurate reflection of reality. The consequences for businesses whose actual performance is competent but unremarkable can be severe: in a system that amplifies extremes, adequacy becomes invisible.
What is needed is not an abandonment of online reputation systems, which have brought genuine benefits in terms of consumer empowerment and market transparency, but a more sophisticated public understanding of their limitations. Users should approach ratings with the same critical scepticism they would apply to any other source of information, recognising that a star rating is not a fact but an aggregation of subjective judgements, each shaped by individual expectations, moods, and circumstances. Platforms, for their part, should invest more seriously in combating manipulation and should present their data in ways that acknowledge, rather than conceal, its inherent imperfections. The alternative — a world in which reputations are made and unmade by systems that few people understand and nobody fully controls — is one that should give us considerable pause.