Reach Is Not Quality — Labproof Notes

Ask anyone which restaurant in a neighbourhood is doing well, and they’ll name the one they see most: the chain in the mall, the opening everyone posted, the brand with the billboards. Visibility reads as success.

In the venue censuses I’ve run across Kuala Lumpur and Bangkok, the same split keeps showing up. The most visible venues (chain outlets, mall counters, hyped openings) tend to rate at or below their neighbourhood’s norm. The quiet, well-run independents rate above it and build reach more slowly. Visibility and quality are different measurements, and a market read that only takes one of them gets the market wrong.

The outside evidence points the same way.

Chains get seen; independents get rated

A study of 21,841 restaurants across seven US metro areas found that chains rate lower than independents on Google Maps, 3.8 stars against 4.2 on average. Yet chains collect more ratings there: a median of 310 against 206 (Li & Hecht, 2020). More reach, lower rating.

That isn’t an accident. Google’s own help page used to say that well-known brands are likely to be prominent in local results because they’re prominent offline (2023 version). The current page says prominence reflects how well-known a business is, including links and review count, and that more reviews and positive ratings can help.

Ratings also do different work for the two. Michael Luca’s study of Seattle restaurants found that a one-star increase on Yelp lifted an independent restaurant’s revenue by roughly 5 to 9 percent, and had no measurable effect on chains (Luca, 2011). A brand arrives with its reputation. An independent has to earn one, a review at a time.

Visibility feeds itself

Once something is visible, it tends to get more visible, whatever its quality. In a large experiment with an artificial music market, showing people what others had downloaded made success more unequal and harder to predict. Quality set the bounds (the best songs rarely failed, the worst rarely won), but the middle was close to a lottery (Salganik, Dodds & Watts, 2006). In another experiment, a single arbitrary early upvote raised a comment’s final rating by a quarter on average (Muchnik, Aral & Taylor, 2013). Reach is partly momentum.

The star average is a weak read on its own

Don’t swing to the opposite mistake and trust the rating alone.

Ratings pile up at the extremes, because people with strong views are the ones who write reviews (Hu, Pavlou & Zhang, 2009; Schoenmueller, Netzer & Stahl, 2020).
Platforms disagree. Google Maps ratings run higher than Yelp’s for the same restaurants, by about a star for chains (Li & Hecht, 2020). Compare within one platform, never across.
Low volume is easy to fake. A study of Yelp found restaurants were more likely to commit review fraud when their reputation was weak, such as when they had few reviews (Luca & Zervas, 2016). A high rating on a handful of reviews is the least trustworthy number on the page.

Ratings still move real demand. An extra half-star on Yelp made San Francisco restaurants sell out far more often (Anderson & Magruder, 2012). The point isn’t that rating doesn’t count; it’s that rating and reach answer different questions.

Read the market on two axes

When I profile an area, every venue is placed on two scales. The first is reach: review count, recency, social following and ad activity. The second is rating relative to its neighbours, on one platform. Four groups fall out:

High reach, lower rating: usually chains, mall outlets and hype. Visible because of the brand or the moment, not the product.
High reach, high rating: the genuine leaders. Rare, and worth studying.
Low reach, high rating: the quiet quality players. Often the best value in the area, and usually the most useful finding in the report.
Low reach, low rating: the tail.

For an operator, the lesson runs both ways. If you’re the chain, your visibility isn’t evidence that your product is winning, so check your rating against the street. If you’re the independent, your rating is your reach engine: it’s the one thing that moves your numbers that a bigger brand can’t buy.

Most of the published research is American and from the earlier years of review sites; the Southeast Asian read here comes from my own census work. Test it on your own street before you stake a decision on it.

Sources

Li, H. & Hecht, B. (2020). 3 stars on Yelp, 4 stars on Google Maps: A cross-platform examination of restaurant ratings. Proceedings of the ACM on Human-Computer Interaction (CSCW).
Google Business Profile Help, Tips to improve your local ranking on Google (current page, and the 2023 version via the Internet Archive).
Luca, M. (2011, rev. 2016). Reviews, reputation, and revenue: The case of Yelp.com. Harvard Business School Working Paper 12-016.
Salganik, M. J., Dodds, P. S. & Watts, D. J. (2006). Experimental study of inequality and unpredictability in an artificial cultural market. Science, 311(5762), 854–856.
Muchnik, L., Aral, S. & Taylor, S. J. (2013). Social influence bias: A randomized experiment. Science, 341(6146), 647–651.
Hu, N., Pavlou, P. A. & Zhang, J. (2009). Overcoming the J-shaped distribution of product reviews. Communications of the ACM, 52(10), 144–147.
Schoenmueller, V., Netzer, O. & Stahl, F. (2020). The polarity of online reviews: Prevalence, drivers and implications. Journal of Marketing Research, 57(5), 853–877.
Luca, M. & Zervas, G. (2016). Fake it till you make it: Reputation, competition, and Yelp review fraud. Management Science, 62(12), 3412–3427.
Anderson, M. & Magruder, J. (2012). Learning from the crowd: Regression discontinuity estimates of the effects of an online review database. Economic Journal, 122(563), 957–989.