PRACTICAL SEO GUIDE
An SEO report should connect the work completed to the decisions your business needs to make. Start with a small number of relevant measures and keep the evidence behind them. A ranking screenshot, an AI answer or a single good week is an observation, not proof that a campaign caused new revenue.
START HERE
Track completed work, discovery in search, useful website visits and qualified enquiries or purchases. Keep an initial baseline and a dated release record. For AEO, record observed citations with their context and use available referral data carefully; no single AI response represents every user or every query.
A STRONGER SEARCH FOUNDATION
A focused set of improvements, chosen for your market and your business.
For a local service, a qualified enquiry might be a real person requesting an offered service within the service area. For an online shop, a completed order is more meaningful than a click on the checkout button. Agree how spam, duplicates, cancellations and enquiries outside coverage will be handled.
Choose a comparison period that fits the business’s seasonality and volume. Record known outages, campaigns and opening-hour changes. Keep a release log with the page, change, intended customer benefit and publication date. When several things change together, explain that the data cannot isolate the effect of each one.
Use Search Console to investigate queries and pages, subject to the data available in the account. Separate searches for the business name from broader service discovery when practical. A higher average position with fewer relevant enquiries is not automatically success. Check landing pages and customer fit before deciding to repeat a tactic.
For a manual observation, record the tool, date, question and cited page. Treat that record as a sample. If a visit includes an identifiable AI referral, compare it with the landing page and subsequent customer action; missing referral information limits attribution. Google’s Search Console reporting includes its AI search features within overall Web search performance, so do not label every Web click an AI click.
Use the evidence to choose a next action: repair a page, clarify a service, improve the contact route, or gather more data. Give the action an owner and a completion check. When data is too sparse to support a conclusion, say what remains uncertain and what observation would change the decision.
| Measure | What it helps answer | Limitation to record |
|---|---|---|
| Published and verified changes | Was the agreed work completed? | Completed work alone does not prove an outcome |
| Relevant query and landing-page trends | Are intended customers discovering the site? | Demand, seasonality and reporting coverage may change |
| Qualified enquiries or orders | Is discovery leading to useful business activity? | Deduplication and attribution may be incomplete |
| Observed AI citations and referrals | Where have we seen the site used or visited? | Samples and referrals do not measure all AI exposure |
| Next action and owner | What will the evidence change? | Include dependencies and a check for completion |
GOOD QUESTIONS
No. A citation may be shown without a visit. Record observed mentions separately from identifiable referral visits and from confirmed enquiries or sales.
There is no universal observation period. Search processing, demand, competition and the amount of data affect what can be inferred. Compare suitable periods and disclose overlapping changes.
These sources support the platform-specific guidance above. Examples and worksheets on this site are educational planning tools.
EXPLORE THE SITE
LET’S MAKE THE NEXT STEP CLEAR
Get a practical conversation about your search visibility, your service area and what to tackle first.