A good review earns its rank; a summary borrows it

On the 8th of this month Google announced a ranking change it called the product reviews update, and it finished rolling out around two weeks later. It is narrower than a core update and Google was explicit that it is separate from one, though it noted that the quality advice it gives for core updates applies here too. What it targets is specifically product review content — single-product reviews, comparison roundups, best-of lists, buying guides — and for now it only involves English-language reviews. The framing Google chose is worth holding onto, because it shapes the right response: the update is designed to better reward reviews that share, in its words, insightful analysis and original research, written by experts or enthusiasts who know the topic well. It is a lift for strong reviews rather than a strike against weak ones. Google said it learned through user testing what most people already suspected from their own searches: that readers appreciate reviews sharing in-depth research far more than thin content that simply summarises a bunch of products. That last phrase is the whole target in five words. An enormous amount of review content online is exactly that — a summary of several products, assembled from manufacturer descriptions and specifications, carrying the right keywords and little else — and it had ranked well enough for years on volume and on matching the query’s words. The update is Google’s move to separate that thin summarisation from the genuine article, and the impact bore it out: for many sites in the review and affiliate space the volatility was comparable to a core update, with some pages dropping hard and others, the ones that had actually done the work, rising. Nothing about this is a punishment. It is a re-rating that finally rewards the reviews that earned it and stops lending rank to the ones that did not.

Google did not leave reviewers guessing about what it now rewards; it published a set of questions to ask of your own content, and every one of them probes the same thing — whether you actually engaged with the product or merely described it. Does the review express genuine expert knowledge about the product where you have it? Does it provide unique content about how the product works and is used, beyond what the manufacturer already states? Does it discuss the benefits and drawbacks of the product based on real research, rather than a one-sided summary? Does it explain how the product has evolved from previous models to improve or fix things? And does it identify the key decision-making factors for the product’s category and rate how the product performs on them — Google’s own example being a car review that weighs fuel economy, safety, and handling? Each question is really a test of first-hand engagement and original analysis. A review written by someone who has spent real time with the product can answer them naturally, because the answers are just what that person knows; a review assembled from the spec sheet cannot, because there is nothing behind it to draw on. The clearest illustration in the field is the kind of review Wirecutter publishes — explaining how it picked the products, how it tested them, what the trade-offs are, who wrote it, with the affiliate relationship disclosed openly. You do not have to match that scale, but it shows the shape of the thing: a review whose value is the work behind it, not the words on the page. The single test that captures all of Google’s questions is whether your review contains anything a reader could not have got from the manufacturer’s own page. If it does, you have a review; if it does not, you have a summary.

Step back from the checklist and the update makes a single argument that runs through everything we believe about ranking durably: the authority of a review is something you earn by knowing the product, and a summary that skips that work is only borrowing an authority it does not have. Think about where a thin roundup’s credibility actually comes from. It has none of its own — it has not tested anything, measured anything, or formed a judgement — so whatever authority it projects is lifted wholesale from the manufacturer’s descriptions and from the surrounding signals it managed to accumulate. That is borrowed authority, and borrowed authority is fragile precisely because it is not anchored in anything the site itself did. A review built on first-hand expertise is the opposite: its credibility is its own, generated by the testing and analysis and honest judgement that went into it, and that cannot be lifted from a spec sheet or faked at scale. The product reviews update is, in effect, Google getting better at telling the two apart and re-pricing them accordingly — devaluing the borrowed, rewarding the earned. This is the same principle that governs links, citations, and reputation across the rest of search: authority that you generated yourself, through genuine work, holds its value, while authority you assembled cheaply from others’ material is exposed the moment the engine learns to look past the surface. A review is just a particularly clear case of it, because the difference between having used a product and having summarised its box is so stark once anyone looks for it. The lesson generalises well beyond reviews: build credibility you own, not credibility you borrow, because only the first kind survives an engine that keeps getting better at telling them apart.

If your reviews lost ground, the response is not defensive but constructive, because there is no penalty to lift — only better reviewing to do, and the work is concrete. First, actually engage with what you review: test it, use it, measure it, and write down what you found that the manufacturer would never tell a buyer — the real-world performance, the irritations, the way it stacks up against the obvious alternatives. That first-hand material is the substance the update rewards and the one thing a competitor cannot copy from the same spec sheet you both have. Second, make your expertise visible rather than implied: state who wrote the review and why they are qualified to judge this category, because both readers and Google weigh a review more heavily when the experience behind it is legible. Third, cover the decision-making factors that genuinely matter for the category and rate the product honestly on each, weaknesses included, since a review that only praises reads as a sales page and is treated as one. And resist the instinct that built the thin-content problem in the first place: do not widen to capture more keywords by adding more shallow reviews of more products. Breadth of summary is the model the update devalues; depth of genuine assessment is the model it rewards. The encouraging truth underneath all of this is that the work and the reward finally point the same way — the review that best serves a buyer is now the review that ranks, so there is no tension between doing right by the reader and doing well in search. Building review content on first-hand expertise and original research, the authority a summary can never borrow, is the earned-authority work the AC Group has done for 27 years.