An SEO report a founder can actually act on
You need to report on SEO to someone who will reasonably ask why the numbers moved.
Short answer
Report non-brand clicks, distinct queries producing at least one click, distinct pages producing at least one click, and conversions from organic. Those four describe whether the channel is broadening or narrowing. Average position and site-wide click-through rate belong in the appendix, because both routinely move in the wrong direction while everything improves.
Split brand out, always
Brand searches convert best and reflect work other teams did. Leaving them in the headline number means a good month of PR reads as a good month of SEO, and a rebrand reads as a catastrophe. One split fixes both.
Breadth beats total
Distinct queries with at least one click, and distinct pages with at least one click. These say whether you are winning more things or squeezing the same few harder. A total that rises while breadth is flat means one page got lucky, and that is fragile in a way the total conceals.
Explain the lag once, then keep explaining it
Search results reflect work done a quarter ago. Say this in every report, because the alternative is being asked to explain a number that was determined before the period began. A month that shipped nothing can grow substantially, and that needs to be normal rather than surprising.
Two numbers that mislead by design
Average position is an impression-weighted average over a changing query set: rank for a hundred new terms at position 40 and it gets worse while clicks rise. Site-wide click-through rate falls whenever you broaden, for the same reason. Neither belongs in a headline.
Report against the right comparison
Same period last year where you have it, trailing periods in whole weeks where you do not. Month against previous month imports both month length and seasonality, which is why it produces the most dramatic and least reliable narratives.
Name the noise
Every report should say what the metric’s normal variation is. Without it, every fluctuation demands an explanation and someone invents one. With it, most fluctuations are correctly identified as nothing.
End with a decision
What you will do next period and why, tied to a number in the report. A report that ends with a chart puts the burden of deciding on the reader, which is how SEO reporting becomes a ritual rather than a management tool.
Questions
- How often should I report?
- Monthly is right for most teams. Weekly reporting on a channel that moves quarterly generates noise and invites reactions to variance.
- What single number should I lead with?
- Non-brand clicks, compared against the same period last year. It is the closest single figure to "is this channel growing", and it excludes the traffic you would have had anyway.
- Should I report rankings?
- Rarely, and never as a list of positions. Rankings are personalised, localised and volatile, and a position without impressions attached says nothing about whether anyone saw it.
Measured, not asserted
The lag is why this matters. On ihatepdf.cv, one month shipped zero commits and grew 61%, from 4,409 to 7,115 clicks. Reported month on month with no explanation of lag, that month looks like a mystery and the following month, which shipped plenty, looks like a failure.
Free tool for this: Search Console Analyzer. No account, nothing uploaded.
Where this goes deeper
Every number on this page comes from one complete dataset: one product taken from zero to 100K+ users on search alone, with nothing spent on advertising. The full argument is Chapter 14 and Chapter 20 of the book. Five chapters are free to read.