Begin with one fixed measurement contract
Record the timezone, reporting delay, comparison periods, search type, country, device, keyword set, Maps grid centres and lead definition. Keep those settings stable so a week-to-week change describes the market rather than a changed dashboard filter.
When a keyword, location or page set is added, give it a separate baseline date. Do not rewrite the original baseline or combine incompatible samples into one trend.
Separate implementation from provider outcomes
Report what was built, tested and deployed, then separately report crawl, index, impression, click, local-visibility and enquiry evidence. A submitted sitemap, indexing request or completed task is not acceptance by a search engine.
Keep a release annotation with the changed pages and date. Without it, later movement cannot be compared responsibly with the action that might have influenced it.
Track technical acquisition health
Count priority pages returning correctly, indexed pages where provider evidence exists, canonical mismatches, sitemap errors, broken internal links, Core Web Vitals groups and real enquiry-path failures. Show the affected commercial pages, not only a site-wide percentage.
A crawler issue count is not a performance score. Separate confirmed defects, accepted conditions, investigations and fixed items awaiting provider recrawl.
Read Search Console as a funnel
Report non-brand impressions, clicks, CTR and relevant page or query movement over a comparable lag-adjusted period. Use page and query tables to find emerging clusters rather than treating average position as the primary result.
If impressions rise but clicks do not, inspect query fit, titles and competing results. If clicks rise but stored enquiries do not, inspect intent, proof, offer clarity and friction.
Keep Maps visibility spatial
For each core service, show the percentage of fixed grid points found, the distribution of positions and every not-found point. Retain the same centre and geometry so proximity changes are not hidden by moving the sample.
Do not blend Maps grid results with organic web rankings. They are different surfaces with different location effects and need separate diagnoses.
Maintain an authority ledger
For every citation or referring domain, record the source, destination, anchor or profile field, acquisition method, submission date, cost, link qualification, approval state and live verification date.
Count a reference as live only after public read-back. Separate submitted, pending, secured, removed and rejected opportunities. Directory account creation alone is not a backlink.
Report reviews without turning them into a quota
Track genuine requests approved and sent, published reviews, response status and policy issues. Never incentivise reviews or ask only customers expected to leave a positive rating.
Review volume is a prominence input and decision aid, not a purchased rank. Keep the available client pool and consent boundary visible when setting expectations.
Define a lead from the stored record
A lead is a genuine enquiry successfully stored in the business system. Report form starts, CTA clicks, phone taps and email clicks as separate interactions; do not add them together and call the total conversions.
Review qualification, service, geography, source landing page and spam status without exposing private message content in the scorecard. Match analytics events back to storage with controlled tests.
Put cost beside each tactic
Show provider usage, paid tools, placements, memberships and content-production costs beside the outcome they support. Distinguish reserved, settled, refunded and recurring amounts and retain approval evidence.
A free tactic still consumes time. Stop low-quality submissions and content that create maintenance without visibility, authority or enquiry signals.
Use decision rules instead of commentary
Pages not indexed after 7–14 days trigger checks of response, canonical, sitemap, internal links and value. Flat impressions across two reviews trigger intent and competitor analysis. Rising impressions with weak CTR trigger snippet and fit work.
Clicks without enquiries trigger conversion diagnosis. No matched movement by day 30 shifts effort toward stronger commercial long-tail needs and authority gaps. A tactic without a leading indicator across two cycles is redesigned or stopped.
End the weekly review with allocation
Classify every active tactic as scale, maintain, revise or stop. Name the page or cluster receiving the next week’s extra content, internal links and promotion, and name the activity losing effort.
Record unknowns and evidence gaps rather than smoothing them into a score. A useful scorecard makes the next decision obvious; it is not a decoration for completed tasks.