Why group reporting hides the branches that are failing

Head office asks how search is doing. A number comes back. Visibility is up four points, or organic sessions are up eleven per cent, or the average position improved. Everyone nods and the meeting moves on.

That number is arithmetically correct and close to useless, because a group does not compete as a group. Each location competes in its own market, against a different set of rivals, for customers who will travel a few miles at most. Distance is one of the three things Google ranks local results on, measured from the customer who is searching.¹ Averaging across forty of those markets destroys the only thing worth knowing.

What the average conceals

Consider a group where the flagship sites are strong and getting stronger, and a tail of smaller branches has been slowly falling out of the map results for a year. The average holds up perfectly well, because the strong sites carry it. Nobody sees the tail until a regional manager asks why enquiries have dried up, and by then it has been happening for four quarters.

The same thing happens in reverse. A group number can drop because two large sites had a bad month, while thirty-eight locations quietly improved. The response is usually a review of the whole search programme, which is the wrong response to that data.

A group average is a summary of markets that never meet. It can move for reasons that apply to none of them.

What to report instead

Keep the mean if the board wants it, but stop treating it as the answer. The useful reporting is about distribution and movement.

The per-location data already exists. In Business Profile Manager you can tick several profiles and download their performance data, including calls, direction requests, website clicks and bookings, as one spreadsheet for comparing locations.²

The uncomfortable part

Reporting per location means somebody has to look at the list. Forty rows is more work than one number, and a list with red on it invites questions that a single green figure does not.

That is the point. The reason blended reporting survives in so many groups is that it is comfortable, not that anybody believes it. It is worth asking, at the next review, whether the person presenting could name the three worst performing locations. If they cannot, the reporting is not doing its job.

Where the variation usually comes from

• Profiles managed locally by whoever happened to have the login, with categories and service lists that drifted apart over years.
• Review volume concentrated in a handful of sites where a manager cared, and near zero everywhere else. More reviews and positive ratings can help a business's local ranking,¹ and each branch has its own profile and its own reviews.
• Location pages generated from one template with the place name swapped, which gives nothing for any individual market to rank on, and which can fall under the doorway rules in Google's spam policies.⁴
• Branches that opened, moved or rebranded without anybody updating the listings that point at them, on Google and elsewhere. Apple Business place cards, for one, carry their own hours and location details into Apple Maps, Safari and Spotlight.⁵
• Competitive pressure that is genuinely different by market, which no amount of central process will equalise.

Only the last of those is a real market problem. The other four are process problems, and process problems are fixable once you can see which locations have them. That visibility is the first thing worth setting up.

A reasonable first step

Take your last search report, break the headline number down by location, and sort ascending. Most groups find something in the bottom ten they did not know about. That exercise costs an afternoon and it usually changes what the next quarter is spent on.