How to review a monthly SEO report

Check the period and the source of every number

Most bad calls in SEO reporting start with an unfair comparison. Before you look at any chart, find three things on the first page: the date range, the period it's compared against and the day the data was pulled. If the report doesn't state all three, ask for them before the call.

Month against last month shows direction. Month against the same month last year shows whether that direction is real or just the season. Search Console keeps 16 months of data, and its help pages suggest weekly or monthly granularity when comparing two ranges, to cancel out day-of-the-week swings. If a drop lines up with a season, Google's guide to debugging traffic drops suggests checking the search in Google Trends to see whether interest fell everywhere or only for your site. The guide to local SEO seasonality covers busy and quiet months in more depth.

Watch for a window that moves. A report that compares 28 days one month, a calendar month the next and a quarter after that can make almost any trend look good. Ask for the same window every time, and a line on the first page whenever it changes.

Next, sort each figure by where it came from. Some numbers come from your own Google accounts, where you can open the same figure yourself. Others are a third-party tool's model on its own scale, and two tools will give two different answers. Both have a use, but only the first kind should carry the headline. The first page of a good report reads like a short memo built from your own accounts: what changed, which actions likely caused it, what didn't go to plan and what happens next.

Check the edges of the period first, then weigh each figure by its source using the table below:

• Pull date. Search Console says its newest data can be preliminary, and Search Console data reaches Google Analytics 48 hours after collection. A report pulled on the 1st can undercount the last day or two.
• Time zones. Search Console dates run on Pacific Time, while Google Analytics uses the reporting time zone set on the property. Month edges won't line up exactly between the two tools, so small gaps there aren't worth chasing.
• Profile searches. Google Business Profile Help says the searches metric updates monthly and can take up to five days to appear.
• Analytics retention. On a standard Google Analytics property, data retention can be set to 2 or 14 months. It only limits explorations and funnel reports, but a year-on-year exploration needs the longer setting.
• Known glitches. Search Console's data anomalies page lists logging errors and reporting changes. It records, for example, that FAQ rich results stopped appearing in Google Search from 7 May 2026, which cut reported FAQ impressions.

Split the totals into cohorts

A total can sit flat while its parts move in opposite directions. Split the data into cohorts, groups of queries or pages that behave alike, and compare each one with its own history.

Start with branded against non-branded searches. Search Console Help describes branded queries as ones that include your brand name, your domain or a product or service specific to your brand, and notes that non-branded searches are often more general. So growth in non-branded clicks is the clearer sign that new customers are finding you. Ask for the two lines side by side in every report. If your report doesn't split them yet, choosing SEO goals for a service business covers the setup, including a workaround for smaller sites.

Say an electrician in Maroochydore sees clicks hold steady for the month. Split by cohort, branded clicks rose after a letterbox drop while non-branded clicks to the switchboard and EV charger pages fell. The total said all was well. The cohorts say the service page work isn't landing yet, and that the letterbox drop propped up the month.

Keep the cohorts fixed from month to month. A cohort that is redefined every report can't show a trend. If the agency adds a cohort, ask for it to be backdated so you can see its history rather than a single month.

Cohorts to ask for, and the question each one answers in a monthly review:

Read the queries, rankings and AI figures with Google's definitions

Search Console's headline numbers are narrower than they look. An impression means a link to your site was seen, or potentially seen, in results. A click counts when someone leaves Google for your page. CTR is clicks divided by impressions. Average position is the topmost spot your site held, averaged across every query it appeared for, and all the links in an AI Overview share that overview's single position.

That last rule trips people up. Average position can improve simply because the site stopped showing for broad searches where it sat far down the page. So read position with impressions beside it, every time. The query table won't add up to the total either: Search Console leaves out anonymised queries to protect privacy, though they still count in the chart totals, and the table only keeps the most important rows.

Most reports also carry a rank tracker table: a fixed list of keywords and the position a tool found for each, often grouped into bands such as the top three. Treat it as a spot check. Google says it uses location, past searches and settings to tailor results, and Google Business Profile Help names distance, how far a business is from the person searching, as one of the main factors in local results. A tracker's position is what one simulated searcher saw from one point. Before you read the movements, ask who chose the tracked list, whether it covers every money service and the suburbs you serve, and whether any keywords were added or removed this month. A list that quietly drops the terms that fell can only ever go up.

AI figures need the same care. Google's documentation on AI features says appearances in AI Overviews and AI Mode are counted inside the Web search type of the main Performance report. Search Console's Generative AI performance report, which Google says reached all websites on 31 August 2026, adds a dedicated view: impressions only, grouped by page, country, device and date, with no queries and no clicks. Third-party AI trackers that list prompts and mentions run their own panels, so read their counts as estimates. How to interpret Google AI impressions covers that report in depth, and measuring ChatGPT referrals covers visits from AI assistants in Analytics.

In the query view, look for:

• Queries with climbing impressions but few clicks, where the title or description isn't earning the visit
• Queries matched to the wrong page, such as a suburb search landing on the homepage
• Queries for services you don't offer or places you don't serve: traffic that looks good and books nothing
• Your short list of most profitable jobs, tracked month to month as their own group, alongside the same terms in the rank tracker

Read the pages that bring in work

Switch to the Pages tab and start with the pages that earn money: each core service page and each area page. This view groups data by page, while the chart stays grouped by the whole site, so page rows won't sum to the headline figure. Clicks on a duplicate URL are credited to the canonical version, which matters when two versions of a page exist.

Give new pages time. Every change takes a while to show up on Google's end, according to its SEO Starter Guide: a few hours for some, several months for others. The guide's general advice is to allow a few weeks before you assess whether the work helped. A page published on the 20th that shows nothing by the 31st hasn't failed yet.

Then open Google Analytics for the same pages. Analytics Help defines an engaged session as one that lasts longer than 10 seconds, has a key event, or has two or more page or screen views. Engagement rate is the percentage of sessions that were engaged, and bounce rate is the share that weren't. A low engagement rate from organic visitors to a service page is worth a look: the page may not match the search that brought them, or the phone number may sit too far down. Judge it against the page's job, though. A visitor who reads your number and rings from another phone leaves no trace in Analytics.

Read the technical section by asking which issues touch money pages. A site health score is a crawler's own scale, so a score moving a few points says little on its own. Google's checks live in Search Console: the Page indexing report, and the Core Web Vitals report, which Google says uses field data from real visitors and groups similar pages as Good, Need improvement or Poor. If the report's speed findings and Search Console disagree, ask why. For what a full technical review should cover, see what an SEO audit should deliver.

Judge the backlink section on the quality of each link. Google's spam policies class buying or selling links for ranking purposes as link spam, so a sudden jump in links with no explanation is a question to put to the agency. Search Console's Links report shows top linked pages, top linking sites and top linking text, but Google says it is a sample, may include links that have since been removed and is capped at 1,000 rows. Worth reading: new links from relevant local or industry sites, links that point at service pages, and an explanation for any important link that was lost.

For each money page, check:

• Clicks and impressions against the same month last year
• Indexing status. Search Console's Page indexing report says all of a site's important pages should be indexed, so confirm each money page is
• CTR against impressions. Search Console Help suggests sorting pages by CTR, lowest first, to find ones people see in results but don't click
• Engagement rate from organic visitors, compared with your other service pages
• Overlap. If two pages keep swapping places for the same searches, read when to merge overlapping SEO pages

Check which leads became booked jobs

A report that counts leads without judging them can reward the wrong work. In Google Analytics, a key event is an event marked as particularly important to the business, and any event can be marked. So ask exactly which events your report counts as leads. A completed form is close to a lead. A tap on a phone link only shows someone pressed the number. Whether the call connected, and whether it was a real enquiry, has to come from your phone records or call tracking.

Business Profile numbers need the same care. Google Business Profile Help says performance data includes views, searches and actions from both organic results and Google Ads, so profile actions in an SEO report can include people who came through an ad. The guide to paid and organic lead attribution covers how to separate them. Each profile metric records a single action someone took on the profile, as the table below shows.

Then check the leads against your own job records. If calls come through tracking numbers, call tracking and conversion reconciliation explains how to match them to real calls. Say a landscaping business in Coolum sees form leads jump after publishing a guide to retaining walls. If most came from outside its service area, the guide found readers and did little for bookings. Sort each organic enquiry into one of these groups:

• Booked job or accepted quote
• Quoted, still open
• Wrong service or outside your area
• Existing customer who searched your name
• Spam, sales pitches and duplicates

Hold every action to a date and a check

The last part of the report should say what was done, why, and what comes next. Google's guide to hiring an SEO suggests asking whether they'll share every change they make to your site, with detailed reasoning behind each recommendation.

Read the work-completed list separately from the results. A useful line names the page or profile changed, the date, the finding it answers, the number expected to move and when it'll be checked. Lines such as ongoing optimisation, technical improvements or a count of hours can't be tested, because nothing ties them to a page or a date. Work listed three months running with no visible result deserves a direct question.

Take a short set of questions to the monthly call. What period is this, and what is it compared with? How did non-branded searches do apart from branded ones? Which money pages gained or lost clicks, and what caused it? Which leads became booked work, and which pages did they start on? What changed on the site or profile, and on what date? What will be checked next month to know whether this month's work helped?

If you haven't agreed what the numbers should point towards, start with choosing SEO goals for a service business, then ask how the local SEO work maps to them. If you'd like a second opinion on a report you've been sent, you can request a strategy review.

Warning signs that a report is steering you away from the full picture:

• Only good news, month after month, with no line on what didn't work
• A date range or comparison period that changes from one report to the next
• A tracked keyword list that grows or shrinks without a note
• Impressions or visibility rising while enquiries stay flat, with no explanation offered
• Third-party scores as the headline, ahead of your own clicks and leads
• No dated record of what was changed on the site or the profile
• Screenshots in place of access to your own Search Console, Analytics and Business Profile

Worked example: a Caloundra plumber's monthly review

Say a plumbing business in Caloundra receives its September report. The business and every detail below are hypothetical, there to show the method rather than to set a benchmark. The summary page says organic clicks rose against August and that rankings for blocked drain searches improved. Before the call, the owner checks the period, source, cohorts, queries, pages and leads in turn.

The period comes first. The report compares September with August only, so the owner asks for September last year as well. The same rise shows up a year earlier, so most of the headline gain follows the business's own seasonal pattern. Next comes the source. The rankings claim comes from a rank tracker, while Search Console shows the blocked drain page with more impressions, fewer clicks and an average position that barely moved: a page seen more often and chosen less.

The cohorts and queries change the story again. Branded clicks carry most of the growth, while non-branded clicks to the hot water page fell, and several hot water searches are landing on the homepage. In Analytics, the hot water page's engagement rate from organic visitors trails the other service pages, and on a phone its call button sits below several screens of general copy. The leads step is the last surprise: the report counts every phone tap as a lead, but the job diary shows a share of profile calls came during an ad campaign, and a few form leads were outside the service area.

The owner's notes from the review, step by step: