The short test: ask whether another analyst could repeat the measurement from the report. If the query set, location, device, dates, and conversion rules are missing, the result cannot be independently checked.
The Four-Part Proof Chain
SEO reporting often jumps from a list of completed tasks to a claim of growth. The missing piece is a controlled comparison. A defensible claim follows four connected steps, and each step leaves evidence.
Make the Baseline and Remeasurement Match
A ranking report is not a fair before-and-after comparison if the agency changes the query list, checks from a different location, expands the grid, or mixes mobile and desktop results. The matching rules matter most in local SEO because Google Maps results can change block by block.
| Control | What a matched comparison requires |
|---|---|
| Query set | The same tracked queries, with additions reported separately |
| Search surface | Organic results compared with organic, map results compared with map results |
| Geography | The same city, search coordinates, and geo-grid boundaries |
| Device | Mobile compared with mobile, desktop compared with desktop |
| Grid settings | The same grid shape, point count, spacing, and map depth |
| Measurement window | Dated captures and a stated comparison period |
If a report adds valuable new keywords, that is useful, but those terms belong in a new coverage section. They should not be mixed into the matched set used to calculate change. Our geo-location analysis uses a fixed grid for this reason, and the Edmonton SEO audit establishes the dated baseline before recommendations are made.
Separate Activity Reports From Outcome Reports
Activity report
- Pages published or revised
- Technical issues corrected
- Listings updated
- Links or mentions earned
- Tests launched
Outcome report
- Matched query visibility change
- Map coverage across the same grid
- Organic clicks and qualified visits
- Qualified calls and form submissions
- Pipeline or revenue under a stated attribution model
Activity is evidence that work occurred. It is not evidence that search visibility improved. A good report includes both columns and does not use one as a substitute for the other.
Read Percentages Beside Absolute Values
Percentages can make a small change look large. Moving from one lead to two is a 100 percent increase, but the absolute gain is one lead. The same rule applies to rankings, clicks, calls, and map coverage. Require the starting value, ending value, absolute difference, and percentage change.
Matched map visibility comparison
Both checks use the same 49 grid points, query, device, business location, and map depth. A visible point means the business appeared in the top three map results.
Absolute gain: 9 visible grid points · Coverage gain: 18.4 percentage points · Relative increase: 75%
The accurate claim is that top-three map coverage increased from 12 to 21 of 49 matched points during the measurement period. The chart alone does not prove which change caused the increase.
Weight Ranking Gains by Demand and Business Value
Ten new number-one rankings can be commercially weaker than one move from position eight to position three on a high-demand service query. Ask for the keyword-level result, monthly demand estimate, intent category, landing page, and conversion path. Branded and non-branded terms should be reported separately because a rise in searches for the company name tells a different story from broader market discovery.
Demand estimates are directional, not exact traffic forecasts. Their job is to prevent a low-demand term from receiving the same strategic weight as a query used by many qualified buyers. Search Console clicks and impressions should then show whether real exposure moved in the same direction.
Define a Conversion Before Counting It
A lead count is meaningful only when the counting rules stay fixed. Write the definition before the campaign begins. For a local service business, a qualified lead might be a new prospective customer in the service area who calls for at least 30 seconds or submits a completed estimate form. State how duplicate contacts, spam, job applicants, vendors, existing customers, and test submissions are treated.
| Metric | Minimum evidence | Common limitation |
|---|---|---|
| Calls | Call log, source, duration, and qualification rule | Missed calls and repeat callers can distort totals |
| Forms | Timestamped submissions with spam and test filtering | Form source can be lost without tagged sessions |
| Booked work | CRM stage or invoice tied to the original lead | Sales follow-up also affects close rate |
| Revenue | Named attribution model and reconciled closed value | Search may assist a sale without being the final click |
State Attribution Limits in Plain Language
SEO rarely operates in isolation. Seasonality, pricing, promotions, news coverage, competitors, review growth, website changes, sales capacity, and Google updates can all affect the same outcome. A responsible report records major changes and avoids claiming sole credit when the evidence only shows that results moved after a set of interventions.
This does not make measurement useless. It makes the conclusion narrower and more honest. A matched geo-grid can demonstrate visibility change. Search Console can demonstrate search exposure and clicks. A call log can demonstrate qualified conversations. Connecting those layers supports a stronger business case, but each connection should remain inspectable.
Printable SEO Evidence Checklist
Use this at the next reporting meeting
- The report identifies the exact website, Google Business Profile, and locations being measured.
- The baseline and remeasurement are dated.
- The same queries, device, geography, grid, and search surface are used in both periods.
- New or removed keywords are separated from the matched comparison set.
- Rankings are shown as absolute values, not percentages alone.
- Keyword gains are weighted by demand and commercial relevance.
- Map visibility is reported across a grid, not from one office computer.
- Work completed is separated from outcomes observed.
- Conversions have a written definition and a named source of record.
- Duplicate, spam, test, and existing-customer leads are handled consistently.
- Revenue claims state the attribution model and its limitations.
- The underlying export, grid, or Search Console evidence can be inspected.
Print this checklist with Ctrl+P or your browser print menu.
For examples of inspectable market evidence, review our Edmonton research and evidence. Those pages are labeled as market studies because they measure a category and do not claim client outcomes. That distinction is part of the proof standard.
How do I verify local SEO leads, qualified enquiries and booked jobs?
Define each stage before accepting an SEO outcome report. A phone-link click is an interaction, a connected call is a contact, a qualified enquiry meets the business’s rules and a booking has a confirmed business record. Those counts should not be presented interchangeably.
| Stage | Evidence | Acceptance check |
|---|---|---|
| Phone-link click | Analytics event from the link | Event describes a click, not a completed call |
| Form enquiry | Successful server-confirmed submission | Validation failures and duplicate clicks excluded |
| Qualified lead | Agreed service, territory and contact criteria | A responsible person or documented process assigns status |
| Booked job | Scheduling or customer system record | Booking exists and duplicate contacts are reconciled |
| Revenue | Appropriate business record and reporting definition | Revenue is not inferred from a click or unconfirmed booking |
Which tracking checks should be completed before reporting results?
Keep an event dictionary with the event name, trigger, source, owner and known limits. Test valid and invalid form submissions, repeated clicks and mobile contact actions. A successful form event should follow confirmed submission, not merely pressing the button. Use an appropriate test environment so verification does not create fake production enquiries.
Document campaign tags on relevant Business Profile links, attribution rules and the reporting window. Search Console describes search performance; analytics describes recorded website activity; business records confirm qualification and bookings. Consent, cross-device activity, delayed calls and direct returns can prevent a complete join. Report the gap instead of forcing every total to match.
How are qualification and booking rates calculated?
The following numbers are an illustrative funnel, not DecodeLocal client results or Edmonton benchmarks. Of 100 received enquiries, 40 qualify and 12 become booked jobs. Qualification is 40/100 = 40%. The qualified-lead-to-booking rate is 12/40 = 30%. The enquiry-to-booking rate is 12/100 = 12%. State the denominator every time.
Deduplicate repeated contacts at the lead stage while preserving legitimate later jobs. Compare matched periods and annotate service changes, staffing and seasonality. With low counts, show the absolute numbers beside percentages. A higher ranking and more bookings in the same period do not isolate SEO as the only cause.
What should a complete lead-measurement handoff contain?
Ask for the event dictionary, qualification rules, tested scenarios, source-of-record mapping and a report showing counts, rates and limitations. A configuration screenshot alone does not demonstrate that a form success, failed submission and duplicate are counted correctly.
Use the service-page SEO guide when enquiries expose unclear scope or an unsuitable offer. Use an Edmonton SEO audit to establish the local search baseline, with tracking implementation scoped separately. Google’s key-event documentation explains how configured events are used in Analytics.