Client satisfaction with monthly SEO reporting runs low across the industry, and the pattern is consistent enough that agencies hear it directly. A working client, quoted in an SEO community thread, said it plainly: “As a client of a few SEO agencies, I hate most monthly reports. It’s almost always data without insights.” That is why so many report templates rank at position one and earn no clicks. The framing is agency-side. What follows is a six-section spine, a concrete example month, and a short test you can run against the PDF sitting on your desk right now.
The six sections a monthly SEO report actually needs
Deliver the template before explaining it. That is the discipline the article is arguing for, so the article should model it. Below is the spine we run every monthly report on, in the reading order clients actually use. Definitions are one line each. The payloads follow in the next section.
- Summary: the verdict in three sentences, the top two wins with numbers, one honest miss, the next step.
- What changed: month-over-month movement across impressions, position, clicks, and sessions, with one sentence of meaning per chart.
- Results by page group: pages clustered by service, category, or funnel stage so movement ties to business lines.
- Learnings: variance explained without excuses, including bad news and the mechanism behind it.
- Next 30 days: phased plan with impact/effort split and named owners.
- Dependencies and open questions: what stalls this if nobody decides, from dev bandwidth to legal approval.
The order is not aesthetic. Clients read reports backward, they scroll to the conversion number first, so the summary has to answer the conversion question before they get there.
What goes in each section, and why the order matters
Six subsections follow, one per report section, in the reading order above. Each is short enough that a working consultant can run it Monday morning.
Summary: the verdict, not the recap
The summary is a verdict, not a preamble. One sentence naming what happened, two lines carrying the wins with baseline and window, one line naming a miss, one line naming the next step. What it is not: a paragraph that begins “This month we continued to focus on…” A verdict sentence looks like this: “Non-branded impressions rose 18% over 30 days on the /resources templates after the canonical cleanup deployed on the 4th; clicks lag by roughly two crawl cycles and should follow in the next report.”
What changed: the diff, not the dashboard
Month-over-month movement across the four metrics that carry the story: impressions, average position, clicks, and sessions. Split branded and non-branded when the mix moves materially. One sentence of meaning per chart, no exceptions. A quick compression that matters: bounce rate is gone in GA4. Engagement rate replaced it, and it should be reported as a supporting signal, not a headline.
Results by page group: cluster to business lines
Group pages by service, category, or funnel stage before you look at the numbers. A URL-by-URL ranked list is not a report, it is a list. On the SITKA Gear engagement, product listing pages and product detail pages were clustered by hunting category, and every month’s report tied movement back to those categories rather than to individual URLs. The rule generalizes: for a B2B site, cluster by service line; for a local services site, cluster by service area; for an ecommerce site, cluster by category or collection. Movement inside a cluster tells the client what is working. Movement across clusters tells them what to invest in next.
Learnings: variance explained, without excuses
If a launch cannibalized rankings, say so. If a competitor’s redesign shifted the SERP, name it. If we do not yet know why a category page dropped, write “our read is” and name what we are testing. The field treats explaining a bad month as a soft skill. It is a hard rule. A report that hides a miss teaches the client not to trust the wins.
Next 30 days: phased and dependency-aware
Split the plan into Phase 1 and Phase 2 with an impact-versus-effort read behind each item. Phase 1 is what ships this cycle regardless of blockers. Phase 2 is what unlocks larger movement once a decision or a dependency clears. We run impact and effort in spreadsheets before any of this reaches the report, which is how the phasing stays honest. Every item gets a date and a named owner. “Ongoing” is not a plan.
Dependencies and open questions: what stalls this if nobody decides
Name the blockers by name. Dev bandwidth for the schema deploy, content workload for the two /pricing expansions, legal review for the medical claim on the /how-it-works page, tracking accuracy on the demo form. Most agencies bury dependencies in a footer because naming them looks like blaming the client. Naming them is the fastest way to unblock work.
The 80/20 rule, applied to what belongs on page one
What is the 80/20 rule in SEO? Roughly 20% of your pages and queries drive 80% of the movement worth explaining. Applied to reporting, it means those pages get named in the summary and contextualized in the body. Everything else gets a line in an appendix, if it earns space at all.
On a mid-sized B2B site, five URLs might drive 70% of a given month’s non-branded impression gain. Those five get named in the summary with their baseline and window. The next twenty pages, taken together, might drive another 20%. They get a paragraph in the “Results by page group” section, clustered by service. The remaining eighty-plus pages get a line in the appendix table, and only if the client asked to see the full set.
The discipline behind the framing is not judgment. It is arithmetic. We run impact and effort in a prioritization sheet before writing a word of commentary, so the summary reflects a decision, not a preference. Reports that show everything at equal weight are not being comprehensive. They are refusing to make a call, and the client is paying for the call.
The rule to steal: if a page is not in your top 20% and is not a leading indicator of one that will be, it goes below the fold.
Metrics that earn their place, and the ones that don’t
The metrics that belong in a monthly report are the ones that either explain movement or measure outcomes. Impressions and average position show up first because they move first, often weeks before clicks. Clicks and conversions show up because they are what the client is buying. Branded versus non-branded split earns space when the mix is material, because a 15% traffic gain from branded queries carries a different implication than the same gain from non-branded ones. Engagement rate belongs as a supporting signal, hedged as a signal not a headline.
Ranking position matters, but only alongside impressions and clicks. A move from position 14 to position 8 is meaningful when the query has volume the client cares about, meaningless otherwise. Shorthand bands treating positions 1 to 3 as excellent, 4 to 10 as first-page, 11 to 20 as on the cusp, and 21 and beyond as low are fine as a mental model. Do not build a slide around it.
Metrics that do not belong on page one: rankings for terms the client does not sell against. Authority Score as a headline number. Anything called “bounce rate” in a GA4 report, since the metric no longer exists in the platform. Traffic totals with no query-mix context. Backlinks acquired this month with no note on which pages they landed. Authority Score in a client report is a tell. It moves slowly, correlates loosely with anything the client sells, and belongs in an internal working document, not on the first page of a monthly.
How to write the commentary: one sentence per chart
Here is a live caption. “Impressions rose 22% on the /resources templates over 30 days after canonical fixes deployed on the 4th. Google is surfacing these pages for more queries, not higher on the same ones, so clicks will lag by roughly two crawl cycles.” Now here is the version most reports ship with: “Impressions up 22%.”
The difference is the mechanism sentence. The rule is one sentence per chart, and that sentence names what changed, why (or our read of why), and whether it matters to the business. The four-part sequence we use across every finding, recommendation, and caption is the same: impact first, mechanism second, fix third, proof condition fourth. At caption scale it collapses to one sentence, but the shape holds.
The register matters as much as the sequence. We recommend, the data suggests, this typically improves, in our experience. Not “Google will reward,” not “guaranteed,” not “dominate.” Confident restraint reads as senior. Overreach reads as vendor. Charts without commentary are not neutral. They force the client to interpret, and that interpretation is the work they hired us to do.
A month, shown in full
An anonymized SMB B2B site, one month, walked through all six sections. The numbers are illustrative but structured the way the real ones would be.
Summary. Non-branded impressions rose 18% over 30 days on the /resources templates after canonical fixes shipped on the 4th; two /pricing pages moved from position 14 to 8 on internal linking cleanup, and one category page dropped four positions after a competitor’s site relaunch. Next: finish the canonical work on the remaining eight templates.
What changed. Impressions up 18% MoM on /resources (baseline 41,200; 30 days). Average position held at 9.2 for that group. Clicks flat because the query set that grew is top-of-funnel and CTR at position 9 sits near 2%. Sessions up 4% overall, sourced from /pricing, not /resources.
Results by page group. Two page clusters carried the month. The /resources templates picked up impressions on informational queries around implementation and integration, a mechanism sentence worth writing down: canonical alignment let Google index the primary URLs instead of the paginated variants, so eligibility expanded. The /pricing cluster moved because internal links from three high-authority blog posts were pointed at the primary URL rather than a legacy variant. The /solutions cluster underperformed, one page dropped after a competitor’s redesign added feature comparison tables we do not yet match.
Learnings. The /solutions drop is not a Google penalty read. Our read is the competitor’s new comparison content better satisfies a query we were ranking for on thinner content. We are testing an expansion to that page in the next cycle before assuming template-level work is needed. The /resources gain confirms that the canonical hypothesis was correct, which lets us finish the remaining eight templates on the same pattern with more confidence.
Next 30 days. Phase 1: finish canonical work on the remaining eight /resources templates and deploy schema on the two winning /pricing pages. Phase 2: topical expansion on the two /pricing wins, one longer piece each targeting the query cluster GSC now shows they are eligible for. Impact-effort read is in the shared sheet.
Dependencies and open questions. Waiting on dev to deploy schema on /pricing (owner: Priya, target: 14th). Waiting on marketing lead to approve the /solutions expansion brief. Open question: is the /solutions competitor testing that comparison page or is it live sitewide? We will confirm in next month’s SERP audit.
Reporting AI visibility without overclaiming it
Here are the facts worth carrying. Google Search Console added a Search Generative AI tab in June 2026, initially in the UK; GA4 added an AI Assistant channel in May 2026. Vendor tools that measure AI visibility exist, including SE Ranking’s AI Results Tracker and Semrush’s AI Visibility Toolkit, which reports a 1 to 100 score plus brand mentions and citations across the major answer surfaces.
The honest share: AI-referral traffic grew roughly 16x from 2024 to 2026 across a large sample of sites, and it still sits near 0.32% of total website traffic. For most SMBs, that is a rounding error. A dedicated AI slide in a monthly report at that share is a marketing artifact, not a report.
When AI visibility does earn a place in the monthly, and it will earn one sooner for some verticals than others, this is what belongs on it: a baseline brand-mention count in the major AI answer surfaces, citation share versus named competitors, and a query set the client actually cares about, tested monthly. Hedge the commentary. “AI-referral share is under 1% of sessions this month” is an honest opening; “AI is the new SEO” is a slogan.
Our position on AI reporting matches our position on AI in general: AI-assisted analysis, human-reviewed outputs, no auto-published claims. Track the numbers internally until the share earns the space in the client-facing report.
Cadence, length, and when to deliver
Monthly for most engagements. Weekly only during migrations, active recoveries, or launches where waiting a month costs meaningful revenue. Quarterly reads as absent to most marketing leads and should be reserved for engagements structured around large seasonal cycles.
Length matches the audience. A marketing lead reading on a phone gets six focused pages with an appendix link. An enterprise stakeholder review might take twelve pages plus a data appendix. One credible convention is eight to twelve pages delivered within five business days of month-end. Use it as a starting point, not a law.
The delivery day matters more than most agencies admit. A report that lands the same Monday every month is a report the client learns to read. A report that lands “sometime in the second week” is a report that gets filed. Pick a day, hold it, and if a month slips, name the slip in the next month’s summary.
Six questions to ask of the report on your desk right now
If a report fails four or more of these, the work behind it may be fine and the reporting discipline is what is broken.
- Is there a summary that names a verdict? If the first paragraph starts with “This month we continued to…”, the report has no verdict.
- Do rankings appear alongside impressions and clicks? If rankings sit on their own slide, the report is measuring position without measuring whether anyone is finding those pages.
- Is bad news present and explained? A month with only wins is a month with hidden misses.
- Are next steps dated and phased? “Ongoing optimization” is not a plan. Dates and named owners are a plan.
- Are dependencies named? If nothing is waiting on you, the report is not honest about what actually moves work forward.
- Can you tell what changed and why in one read? If you have to open a second tab to interpret the charts, the commentary is missing.
Send us the last report you received
If the last report you received made it through this article without you flinching, you are probably in good hands. If it did not, that is fixable. Send us the PDF and we will read it the way we read our own, then tell you what is missing, what is honest, and what the next monthly should look like if you kept the same agency and asked for a different report. No sales call, no pitch. A reporting review, delivered in writing. Send it here.
Common questions about monthly SEO reports
What should a strong monthly SEO report include?
Six sections in this order: summary, what changed, results by page group, learnings, next 30 days, dependencies. What makes it strong is not the sections themselves. Variance gets explained, next steps are dated with named owners, dependencies are called out on the page rather than buried in a footer, and every chart carries one sentence of meaning.
How do you build a monthly SEO report?
Four steps. Pull the data from GSC, GA4, and your SEO suite of choice. Cluster pages by service line, category, or funnel stage before you look at the numbers. Comment on every chart in one sentence that names what changed and why. Close by naming the next 30 days as a phased plan with dependencies. The order matters; commenting before clustering produces the URL-by-URL reports clients skim past.
What does monthly SEO work actually include, and how is it different from the report?
Work and reporting are separate things most agencies muddle. The work is audits, on-page fixes, technical remediation, content production, internal linking, link acquisition, and competitive gap-closing. The report is the evidence that work happened plus the business impact it produced. If your report describes activities without tying them to movement, you are seeing a timesheet, not a report.
How long should a monthly SEO report be?
Match the audience. Six focused pages for a marketing lead reading on a phone. Up to twelve pages with a data appendix for an enterprise stakeholder review. Eight to twelve pages within five business days of month-end is a workable convention.
