Yesterday I wrote of Google now allowing user correction of unverified business listings in Google Maps. It was reported by Barry Schwartz and repeated elsewhere that I, in part, played a part in this outcome. Well it has been a sort of Charlie Brown moment…. You know, when Lucy holds out the football every fall and Charlie, in his trusting way, goes to kick it…

After further looking and some user reports, it turns out that this feature is much less widespread than I previously thought.

I assumed that “unverified” meant any record that had not been claimed by the owner or one of Google’s partners (like Superpages or Talking Phone Book); a record that had no “details” yet associated with it. Google’s actual definition is clearly much more narrow than that.

In fact, while I can replicate the results (so it wasn’t a temporary test on Google’s part) on a single search, I have only been able to find very few “unverified” records in any other industries or locations amongst the many that I have tried.
The issue of accuracy in local data is an important one (see Greg Sterling’s recent post: Data Quality: The Local Achilles Heel) and there is value in allowing the community to correct any errors. In fact as a tactic, the easier it is to correct an erroneous local record, the less Google will be criticized for (even obvious) errors in the data and the way they are assembled.

However, in this case, the essence of Google’s credibility is at stake. The undisputed king of the algorithm has an algorithm that is, on the very rare occasion, assigning a competitor’s web site to a business’s Map record. This error, while uncommon, shatters the perception that Google is infallible in the search arena. An error like this, creates the impression of unreliability and non-trust amongst users. Google is many users’ single most trusted URL on the web and there is implicit faith in its accuracy.

The problem, while not widespread, is serious and allowing users to correct the problem could be a step forward. Allowing correction of the few “unverified” records (please report if you find more) that I have stumbled upon out of the twenty five million U.S. business records extant is not an adequate solution.

Perhaps, the example that I have found, is but a first step in allowing community input. Perhaps an algorithm that fixes the problem is just around the corner. Perhaps I won’t have to feel like Charlie Brown after all.