Start with the employer decisions that matter commercially. Use verified proof, clear ownership and a fixed recheck rather than chasing a higher-looking score.
Recommended sequence: improve the UK employer hub first. Verify US domestic scope in parallel. The weakest numerical result does not automatically deserve the largest investment.
01
A02 · Start here
Turn the UK payroll service into an employer decision hub
Implement the supplied one-page blueprint, with operational versus leadership routes, realistic hiring triggers, what a buyer should provide and direct enquiry. Link from the homepage, team and London role pages.
Suggested owner
UK practice lead + copy/editor + developer
Planning window
First 10 working days after approval
Evidence, completion checks and measurement
Why this action
Ordinary hiring is a strength worth protecting: JGA appears in 25/32 Payroll Manager replacement observations and 27/32 Senior Payroll Specialist growth observations. Portfolio appears in all 32 in each. These results support improving employer evidence and conversion around existing relevance, not treating JGA as invisible everywhere.
Cited material is scattered across several JGA pages: Observed supporting citations reach the UK homepage in 37 answers, the team page in 28, London jobs in 28, the test-labelled page in 22 and the payroll service page in 16. These are answer-level counts and incomplete capture can only reduce observed totals. A clear employer hub can connect those existing surfaces.
Complete when
The page answers who it serves, what is recruited, geography, process, proof and next action. All new statements approved by JGA.
Measure separately
Employer enquiry completion and relevance, followed by a separately dated S07/S11 recheck.
Verify domestic US Payroll Director coverage, then add a dedicated section or page with entity, audience, role remit, multi-state operating context and one approved US search example. Link from the US homepage and payroll service.
Suggested owner
US practice lead + website editor
Planning window
First 10 working days after approval
Evidence, completion checks and measurement
Why this action
The US domestic director situation needs its own evidence: JGA is recommended in 4/32 US Payroll Director observations. Robert Half appears in 27/32 and Frazer Jones in 21/32. The US situation is a different buyer task from the ten UK situations, so this is not an isolated test of geographic effects.
Complete when
US practice signs off every capability claim; employer route works; geography and legal entity are unambiguous; no unsupported placement numbers.
Measure separately
Recheck the same four S03 prompts twice per tool as a separate dated follow-up. Track qualified US enquiries without promising uplift.
Create one confidential leadership/interim case and one systems/integration/governance case. Use anonymised clients if necessary, stating role, geography, hiring problem, JGA contribution, dates and defined outcome.
Suggested owner
Payroll practice leads + client relationship owners
Planning window
Weeks 2–3 after approval
Evidence, completion checks and measurement
Why this action
Leadership and transition briefs are less consistently connected to JGA: JGA appears in 13/32 confidential leadership, 15/32 systems transformation and 14/32 interim/succession observations. Relevant comparator pages make practice leads, hiring models and transformation responsibilities explicit. The study suggests testing more specific proof and scope; it does not establish why an individual model chose a firm.
More confident wording is not the same as stronger evidence: The collected answers include uncited capability claims, historical directories, live adverts, related service pages and publisher testimonials. The report distinguishes these from independent verification. The goal is to supply inspectable facts that match the buyer task, not to manufacture a higher-sounding AI score.
Complete when
Client permission or approved anonymisation recorded; no invented logos or outcomes; unavailable metrics omitted.
Measure separately
Whether follow-up answers use the relevant page for the relevant mandate; citations alone are not treated as sales attribution.
Review /recruitment-test/ against the service page before any consolidation. Align experience totals, named entities and dated award statements. Give the retention figure a denominator, period and method if retained.
Suggested owner
Content owner + developer
Planning window
Week 2 after approval
Evidence, completion checks and measurement
Why this action
Cited material is scattered across several JGA pages: Observed supporting citations reach the UK homepage in 37 answers, the team page in 28, London jobs in 28, the test-labelled page in 22 and the payroll service page in 16. These are answer-level counts and incomplete capture can only reduce observed totals. A clear employer hub can connect those existing surfaces.
More confident wording is not the same as stronger evidence: The collected answers include uncited capability claims, historical directories, live adverts, related service pages and publisher testimonials. The report distinguishes these from independent verification. The goal is to supply inspectable facts that match the buyer task, not to manufacture a higher-sounding AI score.
Complete when
One intended canonical destination per genuine duplicate, verified redirects/internal links if changed, retained originals or rollback plan, and no blind deindexing.
Measure separately
Search Console index/canonical checks by JGA; this public review cannot establish Google-selected canonicals.
Review existing professional-body, directory and public recruiter profiles that the tools actually cite. Correct entity, country, practice lead and current service scope through the normal owner/editor routes.
Suggested owner
Marketing + partnership owners
Planning window
Weeks 2–4 after approval
Evidence, completion checks and measurement
Why this action
More confident wording is not the same as stronger evidence: The collected answers include uncited capability claims, historical directories, live adverts, related service pages and publisher testimonials. The report distinguishes these from independent verification. The goal is to supply inspectable facts that match the buyer task, not to manufacture a higher-sounding AI score.
Complete when
Descriptions are factual, consistent and approved; no paid placement presented as an independent endorsement and no unsolicited mass submissions.
Measure separately
Record changed URLs and dates; monitor future citation use separately from recommendation changes.
06
A06 · Next workstream
Measure a fixed follow-up instead of selecting favourable examples
Freeze the changed-page list and rerun priority situations with identical prompts, country and declared product conditions. Preserve all answers, including omissions, and keep this separate from the baseline.
Suggested owner
Riseklix analyst + JGA marketing
Planning window
After agreed pages are live and discoverability is checked
Evidence, completion checks and measurement
Why this action
A recommendation can change between observations: Across 176 exact question/product pairs, JGA appears in both observations for 80, in one for 48 and in neither for 48. Preserve the two answers and their conditions. A single positive answer is insufficient evidence of stable coverage.
Complete when
Both runs saved, dates/modes disclosed, source-domain and recommendation changes shown together; no guaranteed ranking or revenue promise.
Measure separately
Priority-question recommendation counts, cited pages, relevant enquiries and implementation completion. A before/after difference alone does not establish causality.
Owner roles and windows are planning suggestions, not work already instructed or delivery commitments. JGA must approve factual claims and implementation.