Most SEO reports are written to be survived rather than read. Here is the reporting discipline we settled on after nine years of getting it wrong in interesting ways.
Start from the money, work backwards
A report that opens with sessions is already lost. Ours open with the number of qualified enquiries attributable to organic search, the cost of producing them, and the resulting cost per enquiry. Everything else — rankings, impressions, click-through rate — sits below that as diagnostic detail explaining why the top-line number moved.
This sounds obvious. It is not how most reports are structured, because the top-line number is the one that can embarrass an agency, and the diagnostics are the ones that reliably look good.
Define a qualified lead before you measure anything
The single most common failure we inherit is a tracking setup that counts every form submission equally. A recruitment spam message and a $40,000 commercial enquiry both fire the same conversion event, and the bidding algorithm dutifully optimises toward whichever is more common. Before instrumenting anything, we sit down with whoever answers the phone and write a one-paragraph definition of a qualified lead. Then we build tracking that can distinguish it.
Three metrics we no longer report
- Average position. Averaged across thousands of queries it is meaningless, and it moves for reasons unrelated to performance.
- Keyword counts. "Now ranking for 4,200 keywords" tells you nothing about whether any of them are worth having.
- Domain authority scores. Third-party estimates of a third-party estimate. Useful as a rough competitive screen, useless as a KPI.
Attribution windows and the honesty problem
Search-driven purchases in professional services routinely take sixty to ninety days from first touch to signature. A thirty-day attribution window will systematically under-credit organic search and over-credit whatever channel happened to be last in the chain. We set client windows to match observed sales-cycle length, measured from their own CRM, and we state the window on every report so no one is comparing incompatible numbers month to month.
Report the failures
Every Harborline monthly report has a section headed "What did not work." A landing page test that lost, a content cluster that never gained traction, a paid campaign we paused. It costs nothing to include, it builds more trust than any chart, and it forces us to actually review our own decisions instead of quietly abandoning them.
If your current report does not contain a single thing that went badly, that is not because nothing did.
Related reading
Not sure where your search traffic is leaking?
Send us your domain. We will return a 12-point technical and commercial audit within three business days — no fee, no obligation, and no automated PDF dump.