Nearly 30 years of the Harden’s survey
Harden’s uses market research methodology to find out what ‘ordinary’ diners-out think about restaurants.
In a given year, the total number of reporters in our combined London/UK survey, conducted mainly online, numbers 8,000, and, between them, they contribute 50,000 individual reports.
At a time when a studies suggest that as many as 1/3 of the reviews on TripAdvisor are paid for by the restaurants they cover, we believe there is an ever-greater need for trusted sources such as the Harden’s annual diner survey.
For while obviously folks can attempt to stuff the Harden’s ballot too, the high degree of editorial oversight plus the historical data we have both about the restaurants and also about those commenting makes it much harder to succeed.
In this way Harden’s can socially source restaurant feedback, but – vitally – curate it fully as we do so. It is this careful curation which provides extra ‘value-added’ for diners.
How we determine the ratings
In the great majority of cases, ratings are arrived at statistically. This essentially involves ‘ranking’ the average survey rating each restaurant achieves in the survey – for food, service and ambience – against the average ratings of the other establishments in the same price-bracket.
(This is essentially like football leagues, with the most expensive restaurants going in the top league and the cheaper ones in lower leagues. The restaurant’s ranking within its own particular league determines its ratings.)
How we write the reviews
The tone of each review and the ratings are largely determined by the ranking of the establishment concerned, which we derive as described above.
At the margin, we may also pay some regard to the proportion of positive nominations (such as for ‘favourite restaurant’) compared to negative nominations (such as for ‘most overpriced’).
To explain why a restaurant has been rated as it has, we extract snippets from survey comments (“enclosed in double quotes”). On well-known restaurants, we receive several hundred reports, and a short summary cannot possibly do individual justice to all of them.
What we seek to do – without any regard to our own personal opinions – is to illustrate the key themes which have emerged in our analysis of the collective view.
