Agree the measurement before the work starts
Choose the relevant market, page group and business goal. Confirm that a successful contact or purchase is recorded correctly, and record the baseline period and any known tracking limits.
Search Console measures Google search performance; GA4 measures the visits and events it can observe under your setup and consent conditions. Their totals are not expected to match exactly. CRM or sales records are needed to judge inquiry quality and closed business.
Where useful, keep identifiable AI referral visits as a separate segment. A brand mention in an AI answer is not a visit, a lead or revenue. See the AI visibility measurement guide.
Use a short chain of business measures
| Stage | What to record | Important limitation |
|---|---|---|
| Visibility | Relevant non-brand queries and landing pages | Impressions are not visitors |
| Visits | Organic landing sessions under the chosen reporting setup | Consent and attribution affect observation |
| Inquiries | Successful contacts or agreed conversion events | Not every inquiry is suitable |
| Opportunities | Qualified inquiries and their sales progress | Requires a consistent business definition |
| Value | Attributable sales and contribution margin | A sale may involve several channels |
Define “qualified” in terms your team can use—for example, the right service, service area and plausible project—not a vague lead score added after the results arrive.
Calculate return using the right value
For a contribution-based estimate, use: (attributable contribution before SEO costs − SEO costs) ÷ SEO costs × 100. Contribution here means revenue after the agreed variable costs; it is not the invoice total or necessarily accounting profit.
Illustrative example: attributed sales revenue of €12,000, a 40% contribution margin and €3,000 in SEO costs gives €4,800 contribution before SEO costs. The estimated return is (€4,800 − €3,000) ÷ €3,000 = 60%. These are example figures, not my project results.
Using the full revenue instead would produce a much larger percentage while ignoring delivery costs. If margin or attribution is unknown, report revenue-to-cost or lead indicators with that limitation—do not label them confirmed profit ROI.
Do not treat attribution as proof of causation
Organic search may introduce a customer who later returns through another channel. A brand search may reflect existing demand created elsewhere. The reporting model decides which channel receives credit; it does not prove what would have happened without SEO.
Use consistent definitions and comparison windows. Note seasonality, stock, promotions, price changes and website releases. For low volumes, show the actual measurement limits internally rather than promoting a dramatic percentage from a tiny base.
A monthly report should explain a decision
A useful review can fit on a few pages: work completed and checked, relevant search movement, lead quality, known limitations and the next priorities. Name unresolved dependencies as well as successes.
Agree review dates and operational deliverables, not guaranteed traffic or sales increases. SEO timelines and pricing models help define a sensible engagement. You can also send your website and goal if you want help putting the measurement and work plan together.
Further reading: the GA4 traffic acquisition report and recommended event definitions.