SEO reporting has a credibility problem, and it is largely self-inflicted. Counting branded organic traffic as an SEO win is the equivalent of an ad agency invoicing you for people who typed your name into the address bar.
Split brand and non-brand, permanently
In Search Console, filter queries containing your brand and its common misspellings into one bucket and everything else into another. Report the two separately from the first month. Non-brand is the line SEO is accountable for; brand movement usually reflects paid media, PR or seasonality.
Value the outcome, not the session
Assign a value to the conversion, then work backwards. For ecommerce, use gross margin rather than revenue. For lead generation, use qualified-lead value: pipeline value multiplied by historical close rate, discounted for the leads sales rejects.
A model finance will accept
- 1.Non-brand organic conversions in the period
- 2.Multiplied by average conversion value at margin
- 3.Minus the pre-engagement baseline, projected forward with seasonality
- 4.Minus total programme cost including internal time
- 5.Divided by that cost, for a return figure with a defensible numerator
Handle incrementality honestly
Some organic conversions would have happened anyway through another channel. You cannot eliminate that ambiguity, but you can bound it: compare converting journeys that include an organic entry against those that do not, and run geo or category holdouts where the business allows.
Report leading indicators monthly, ROI quarterly
| Cadence | What to report |
|---|---|
| Monthly | Indexation, non-brand impressions, average position on target clusters, published output |
| Quarterly | Non-brand conversions, pipeline influenced, return against cost |
| Annually | Share of voice, brand demand growth, content asset value |
Monthly ROI reporting on a channel with a three-to-six-month lag produces panic in quarter one and complacency in quarter three. Match the reporting rhythm to the channel's actual physics.