What a monthly local SEO report should actually tell you

Most local SEO reports are built to survive a meeting, not to inform a decision. Twenty pages, a lot of green arrows, impressions up, keywords ranked up, and no answer to the only question the owner has, which is whether this is producing customers.

I ran restaurants before this. A report you cannot act on is a report you stop reading, and once the owner stops reading it, the relationship is running on faith.

What data actually exists to report on?

Less than the page count suggests, and knowing the real list is the fastest way to spot padding.

Google's Business Profile performance report publishes a fixed set of metrics: interactions, searches, views, directions, calls, website clicks, messages, bookings, booking clicks, products, menus and offers. That is the whole inventory from the platform itself.

Note what is not on that list. There is no rank. There is no position. Google exposes no local pack position anywhere in the performance report, and Search Console has no local pack dimension either. Any position number in your report came from a third party tool sampling searches from chosen coordinates, which is legitimate and useful, and it is a measurement of that sample rather than a figure Google published.

Google also changed what some of these count. Views is now a deduplicated unique visitor number, and Google states plainly that "a person can only be counted once a day." If your report compares this year's views to a figure from an older tool or an older methodology, it is comparing two different quantities.

What are the only numbers worth putting on page one?

Actions, not exposure.

Page one holds calls, form submissions, direction requests, and booking starts, each with last month and the same month a year ago beside it. Those are the things that turn into revenue, and they are the only numbers an owner should have to read to know whether the month was good.

Everything else lives behind them. Impressions and views tell you the top of the funnel moved, which is worth knowing only when actions did not move with it. That gap is the interesting signal: more people saw you and fewer contacted you means the problem is the listing or the site, not the ranking.

A report that leads with impressions is a report hiding the number that matters.

Add one line of context per number. Not "calls up 14," but "calls up, and the increase is on weekdays before 10am, which matches the hours change we made on the fourth." A number without a cause is trivia.

Split calls from website clicks rather than reporting a combined action count, because the two are moving in different directions right now. Sterling Sky's 2026 field data, compiled from Jepto across 179 profiles at 34 US law firms over two years, shows clicks to call from Business Profiles declining even for profiles whose rankings held steady. Website clicks did not fall the same way, because website icons appear on desktop while call buttons dominate mobile. A report that adds the two together will show you a flat line over a real problem.

How should rankings be reported?

Geographically, for a short list of named terms, with the same method every month.

A single position number is close to meaningless in local search, because your position is different at every point in your service area. The reporting standard is a grid, and why one ranking number cannot describe a local result is worth understanding before you accept any rank report at all.

Three rules keep a rank section honest:

• Name the terms. Three to five terms you chose because they produce customers, not a count of keywords ranked.
• Fix the points. The same grid, the same spread, every month. Changing the measurement changes the result.
• Show the map, not the average. An average across a grid hides that you dominate your own block and disappear four miles east.

If the report shows a term improving, it should also show which neighborhoods improved. That is where the decision lives, because a gain in a part of town you cannot serve profitably is not a gain.

How do I audit the report I already get?

What belongs in the report that most agencies leave out?

The work log, the review comparison, and the honest list of what did not happen.

The work log is one line per change with a date: category changed, twelve services added, hours corrected, four pages published. Without it, you cannot attribute anything. With it, you can look back three months later and see what actually moved the number.

The review comparison should be your new reviews this month next to your top two competitors' new reviews this month, plus one number almost nobody reports: days since your most recent review. Total review count is a vanity number that hides a flat lining profile. The recency figure moves before the ranking does. In Whitespark's 2026 Local Search Ranking Factors report, in which 47 local search experts scored 187 factors, recency of reviews ranks eleventh at 164 points and a sustained flow rather than bursts ranks fourteenth at 154, while total review quantity with text sits ninth at 170. Three separate mechanisms, and most reports show only the third. It should also name where the reviews came from, because reviews from customers who loved you two months ago come from a different motion than reviews from this week's customers.

The honest list is the one nobody includes. What was planned, what was not done, and why. An agency that never reports a miss is either not planning or not telling you.

If your program includes location pages, the report should show them individually rather than as a group. That is how you catch the point where local landing pages start hurting instead of helping, which shows as a set of pages with impressions and no actions.

What should I ignore in the report I get now?

Anything that cannot change a decision.

Cut these without guilt: total keywords ranked, domain authority scores, backlink counts from tools nobody at Google uses, social impressions, a screenshot of one search from the agency's office, and any chart with no baseline.

Cut industry benchmarks hardest of all, because there is no current one worth quoting. The only large study of Business Profile performance is BrightLocal's, published July 2019 on 45,264 businesses across 36 industries, drawn from a data window running September 2017 to December 2018. It reported medians of 1,009 searches, 1,260 views and 59 actions per listing per month, and a view to action conversion rate of 4.68%. Three things disqualify it for your report today. It is drawn from users of the publisher's own audit tool, so it skews toward businesses that already care about this. It has never been refreshed. And Google has since retired or redefined several of the metrics it was built on, including the direct versus discovery breakdown, which is no longer reportable at all, and views, which is now a deduplicated unique visitor count. The 4.68% cannot be reproduced on a 2026 account.

The newer numbers in circulation are worse. Birdeye's state of Business Profiles page publishes an action mix with no sample size and no methodology stated anywhere on it. WebFX discloses only in-house performance data with no sample and no date range. Neither is citable, and both get pasted into agency reports as an industry average.

Benchmark against your own trailing 90 days and against your named local competitors. Nothing else. Say that sentence out loud in the meeting, because every competing report will contain a benchmark that does not exist.

Be equally skeptical of technical items presented as wins. Structured data added is a task, not a result, and structured data is not required for AI features despite how often it appears on a report as a headline.

Apply the same test to any AI search section. Google's Search Console generative AI performance reports, launched June 2026, give impressions by page, country, device and date. Google states explicitly that clicks, click-through rate, position, and the user's prompt are not available, and that AI Mode and AI Overviews are not separated from each other. If your report shows an AI ranking, an AI click count or the prompts customers typed, those numbers did not come from Google.

Also cut anything explaining strategy in the same document as results. Those are two different meetings. The monthly report answers what happened. The quarterly conversation answers what to do next.

If your report shows movement on a term you do not sell, that is worth a question rather than a celebration. It usually means the profile is eligible for searches that are not yours, which points back at choosing the primary category when several fit.

What to do this week

Run the page audit above on last month's file and write down the two numbers: pages total, pages that could change a decision. Send the decoration list back and ask for it removed.

Write down the three to five search terms you actually want. If your report does not track exactly those, that is the first fix.

Ask for a work log with dates for the last three months. If nobody can produce one, you have learned something more useful than any chart in the file.

Then set your own baseline for calls, forms, and direction requests, so next month's report has something to be measured against that you own.

Questions about the report you are getting?

Email me at eric@seod.com with last month's report, from whoever produces it. I will mark it up and send it back: which numbers are real, which are decoration, which claims cannot be supported by the data shown, and the two things I would ask them to add.

I read these myself and I will not use it as a pitch. If the report is solid and the agency is doing good work, I would rather tell you that than take a client who is already being served well.

Hold whatever report you get to the four questions above. Ours arrives weekly rather than monthly, with keyword position, calls, and the single next fix, which is the reporting standard inside Search visibility. There is more on local search in the library.