An SEO report has one job: tell the person reading it what was done in search during a period, and what changed. Everything else, every chart and every table, is there to support those two answers or should be cut.
Most SEO reports fail that job in one of two ways. They are dashboards exported to PDF, full of numbers with no story about the work. Or they are stories with no evidence, where every number is up and nothing is ever flat. This post covers what a report should contain, how the common report types differ, and the rules that keep it honest, with a real one-page report for our own site.
What is an SEO report for?
Three readers, three uses.
- The person paying (a founder, a CMO, a client) wants to know whether the money and time produced anything, and whether to keep going.
- The person steering (a head of marketing, an account lead) wants to know which work to do more of and which to stop.
- The people doing the work want to know what is left and what is blocked.
One report rarely serves all three. The common solution is two formats: a short board-level report for the first reader, and a detailed report for the second and third.
What sections does an SEO report need?
In order of what the reader asks first:
- Summary. Two or three sentences: what shipped, what changed, what is next. If the reader stops here, they should still know the state of things.
- Goals and pace. Each goal with its metric, where it stands now, the target, the deadline, and whether it is ahead of, on or behind the straight line from start to target. A goal with no reading yet says "not measured yet", never zero.
- Work shipped. What reached the live site or the web in the period, by kind: pages rewritten, articles published, technical fixes merged, listings and placements. Also what is waiting and what failed.
- Visibility in Google. Impressions, clicks and positions for the query groups that matter, from Search Console, compared with the previous equal period.
- Visibility in AI answers. How many answers named you and how many cited your pages, out of how many answers read, per engine, for the tracked questions.
- Competitors. Who is gaining on the questions that matter, and which pages they win with.
- Spend. What the period cost, in money, and where it went.
- Next period. The work planned, and why.
How do audit, competitor and search marketing reports differ?
The same reader, different questions. Each has its own format.
SEO report
- Question it answers
- What did we do and what changed?
- Covers
- A closed period
- Frequency
- Monthly or quarterly
SEO audit report
- Question it answers
- What is wrong with the site, and in what order should we fix it?
- Covers
- A snapshot
- Frequency
- Once, then after big changes
Competitor report
- Question it answers
- Who wins the questions we want, and with which pages?
- Covers
- The current state of the market
- Frequency
- Quarterly, or when something shifts
Search marketing (SEM) report
- Question it answers
- What did paid and organic search deliver together?
- Covers
- A closed period
- Frequency
- Monthly
An audit report is a prioritised list, not a score: findings grouped into what blocks access, what loses signal and what improves pages, each with an owner. Our site audit checklist gives the checks.
A competitor report ends every finding with an action. "Competitor X ranks first for Y" is a fact; "Competitor X wins Y with a comparison page, and three roundups engines cite name them and not us" is a work item.
A search marketing report puts paid and organic side by side, because they compete for the same results page. Report them separately, then note where one covers a gap in the other, for example ads on queries where an AI Overview took the organic clicks.
What rules keep a report honest?
These are the rules we hold our own reports to.
- Every number names its window and its sample. "Named in 12 of 40 answers, 1 to 30 September" can be checked. "Visibility 30" cannot.
- No movement words without proof. "Increased", "improved" and "dropped" only where a before-and-after comparison on the same questions cleared a significance test. Otherwise write the two readings side by side and say the difference is within the noise.
- Missing is not zero. A metric nobody measured is shown as not measured, with the reason, never as 0.
- Work and outcome are separate. What shipped is a fact you control. What the engines did afterwards is a measurement. A report that blurs them overclaims.
- "Nothing changed" is a result. Google itself says changes can take weeks to months, and advises waiting a few weeks before judging. A report that is always up is not measuring.
- No revenue claims without evidence. Show organic sign-ups and pipeline beside the work; do not claim one caused the other.

