The Google Business Profile reporting dashboard for local SEO agencies

The Keywords tab lists the search terms Google attributes to the listing as monthly impression counts, each tagged with a branded or discovery pill. A segmented control filters to one group or the other, a second control sorts by impressions or alphabetically, and the share column recomputes against whichever segment is on screen. Branded terms are rendered as bracketed placeholders rather than the client name, because the shape of the term is the reportable fact: the name on its own, the name plus hours, the name plus reviews, the name plus directions.

Discovery share is the number a local retainer is judged on. Branded impressions grow when the brand grows; discovery impressions grow when the listing starts winning category and near-me searches it was not winning before, which is the work an agency is being paid for. The tab states both shares as headline tiles, and it adds a third the industry usually hides: what percentage of profile impressions the keyword table actually names. Google never attributes every impression to a named search, so the table sits under the impressions KPI instead of pretending to reconcile with it.

Impressions are a means; actions are the point. The Actions tab breaks the window into the five things a Business Profile can record — website clicks, direction requests, calls, messages and bookings — as a card each and a row each. Every row carries the raw total, its whole-percent share of all actions, a per-day figure, and the rate that matters for comparison: actions per thousand impressions. Direction requests and calls usually dominate a storefront while website clicks lead for a service business, and seeing the mix stops a report from treating every listing the same way.

Because the rate is derived from the counters rather than stored beside them, it stays true when the date range changes. A quiet month with a better action rate is a different conversation from a busy month with a worse one, and the report lets you have both. A calls-by-weekday chart sits beside the rate: phone volume follows opening hours rather than browsing, so its peak often lands on a different day from the impressions peak — useful evidence when a client is deciding when to staff the counter or when to run a promotion.

The Reviews tab starts with the distribution rather than the average: five bars from one star to five, each carrying its count and a whole percent, and whole percents that add up to exactly one hundred. The average is stated above them and re-derives from those same bars, so a client who adds the numbers up lands on the figure printed on the card. Four tiles carry the flow through the window — new reviews, replies published, reviews still waiting and the lifetime count the average rests on. The recent list shows each review with a star pill, a sentiment tag and its reply state, with reviewers reduced to initials. Average rating moves in points rather than percent, and the KPI card reports it that way: 4.5 to 4.6 is six hundredths, not two percent.

You cannot make a client's customers leave five stars, but you can answer every review and answer it quickly, and that is a deliverable an agency can be held to. The Responses tab states it as three tiles: response rate across the reviews received in the window, median response time in hours, and the size of the unanswered queue. It says in words that a lower median is the better outcome rather than leaving that to a colour. Beneath the tiles sits the working queue, every review still without an owner reply, oldest first, with the days it has been waiting and a chip that escalates as that wait grows. A rating is a level rather than a total, so it never multiplies when the date range widens.

Local clients rarely have one listing. The Locations tab opens with a card per location and a comparison table beneath it: impressions, the window-on-window trend, customer actions, action rate, average rating, new reviews and response rate, side by side. Clicking a row re-bases the entire report on that location — the KPI band, the trend, the impression mix, the keyword table, the reviews and the response queue all narrow to it, and a location chip in the toolbar does the same job from anywhere in the report. A back control returns the rollup.

The rollup does the arithmetic properly. A combined average rating is weighted by review volume rather than averaging the location averages, so a branch with forty reviews cannot outvote one with six hundred. Impressions, actions and posts add; ratings, review counts and open questions are levels that stay put when the date range changes. And because a real portfolio is uneven, the comparison is built to surface the weak location rather than smooth it away: a falling trend and an unanswered review pile are exactly what the table is for.

A local retainer is judged on three questions in order: is the listing being seen, is being seen producing calls and directions, and is the rating holding up while that happens. The tabs are ordered to answer them in that sequence, the KPI band summarizes all three, and the export control names the location scope and date range it was taken under so a deck slide can be traced back. Agencies still ask for a Google My Business report by name years after the rename, and this is that report under the product's current name.

It is as important to say what is not here. There are no photo view metrics, because Google retired the legacy Insights API along with them and any dashboard still showing photo views is showing a figure nobody can source. There are no third-party benchmarks either — the report compares a client with its own previous period, never against a made-up sector figure. Connecting the client's Business Profile account is a separate setup step in the workspace; until then the report renders a deterministic sample profile so you can see the structure before committing to it.