Selected work and evidence

Evidence should make the next decision easier.

This page explains the work we deliver and the proof we expect to retain. Named results will only be published when the client has approved the account, the comparison period is clear and the underlying evidence can support the claim.

01

Technical SEO and assurance

Find the failure, make it reproducible and verify the fix

Our technical work connects crawl data, rendered-page evidence, Search Console and the underlying templates. The output is a prioritised repair plan or an implemented change with acceptance checks.

Evidence retained

  • Affected URLs and templates
  • Before and after technical checks
  • Release risks and validation record
See the related service
02

AI consultancy and implementation

Turn one business process into a controlled pilot

We define the task, approved inputs, model behaviour, human review and success measure before a pilot is treated as ready for wider use.

Evidence retained

  • Current and proposed workflow
  • Data and review boundaries
  • Pilot results and adoption decision
See the related service
03

Training and adoption

Build a method the team can repeat after the session

Training uses the organisation’s roles, documents and real decisions. Follow-through focuses on reusable workflows, quality checks and the next practical action.

Evidence retained

  • Role-based exercises
  • Reusable working materials
  • Action and follow-up plan
See the related service
04

Analytics and conversion

Trace the complete journey from visit to received enquiry

We test the visible page, form behaviour, stored submission, analytics event and notification route. Improvements are based on the full journey, not a dashboard metric in isolation.

Evidence retained

  • Cross-device journey tests
  • Event and notification records
  • Prioritised conversion changes
See the related service

Publication standard

What a future case study must contain

A publishable case study needs a clear starting point, dated comparison, agreed measure, description of the intervention and the factors that limit attribution.

Where confidentiality prevents a named account, we may describe the method without attaching an invented quote or unsupported uplift. An anonymous percentage is still a claim and needs the same evidence.

Until that permission and evidence exist, the service pages explain the process, deliverables and validation standard instead.

Bring us the process or search problem you need to improve.

We will suggest a focused starting point and the evidence needed to decide what happens next.

Start a conversation