Services

Growth, UX & Analytics

Find out where you lose people, then prove that you fixed it.

What it is
Conversion work: UX audits, analytics you can trust, controlled experiments, and the research that tells you what to change.
Who it's for
Teams with traffic that does not convert, or with dashboards nobody believes.
Result
A prioritised list of what to fix, measurement that actually fires, and experiments that show whether the fix worked.

The problem

Two things are usually true at once: nobody is sure where users drop off, and nobody quite trusts the analytics that would tell them. So decisions get made from opinion, shipped without a control, and declared a success because the number went up during a month when three other things also changed. The fix is unglamorous. Audit what is actually being tracked, repair it, find the friction in the interface, then change one thing at a time and measure it properly.

What's included

  • UX audit with prioritised findings and before/after visual examples
  • Severity and expected-impact rating on every finding, so you can fix in order
  • Conversion rate optimisation across interface, product concept and content
  • Audit of existing analytics: what is configured, what actually fires, where the holes are
  • Analytics built from scratch: event taxonomy, funnels, attribution
  • A/B testing: hypothesis, experiment design, supervision, interpretation of results
  • Competitor analysis and market positioning review
  • User testing and customer development interviews
  • UX/UI design grounded in what the research found
  • Design system and component library, handed off for developers rather than for a portfolio

How we work on this

  1. 01

    Measurement before opinions

    We check what your analytics actually records before drawing conclusions from it. Broken tracking is common and it invalidates everything built on top.

  2. 02

    Findings come ranked

    An audit that lists forty problems without ordering them is a to-do list nobody starts. Every finding carries a severity and an expected impact.

  3. 03

    One change at a time

    If three things ship together and the number moves, you have learned nothing. Experiments are designed so the result is attributable.

  4. 04

    We report what happened

    Including the tests that lost. A negative result you can trust is worth more than a positive one you cannot.

Stack

  • Plausible
  • Umami
  • GA4
  • A/B testing
  • Funnel analysis
  • Attribution modelling
  • User testing
  • Competitor analysis
  • Figma
  • Design systems

Powered by Claude Code

We use AI-assisted development as part of the production process, not as a demo. In practice that means fewer billed hours for the same delivered scope — the saving lands on your invoice, not in our margin. Our team holds Anthropic certifications in Claude Code, the Model Context Protocol, Agent Skills and the Claude API.

Packages

UX Audit

Where your interface loses people, ranked by what it costs you.

Price on request

Scoped at estimate

What's included

  • Heuristic review of the core conversion flows
  • Prioritised findings with before/after visual examples
  • Severity and expected-impact rating per finding
  • Walkthrough call with your team

Not included

  • Implementation of the fixes
  • User research sessions

You receive

Audit document with prioritised findings and visual examples.

Get an estimate

Research & Redesign

Find out what users actually do, then redesign against that evidence.

Price on request

Scoped at estimate

What's included

  • Competitor analysis and market positioning review
  • User testing and customer development interviews
  • UX/UI design grounded in what the research found
  • Design system and component library
  • Handoff built for developers, not for a portfolio

Not included

  • Recruiting research participants
  • Front-end implementation

You receive

Research report, design files, design system, developer handoff.

Get an estimate

Questions about this direction

Our analytics is already set up. Why audit it?

Because 'set up' and 'working' are different states, and the gap is where bad decisions come from. The common failures are events that stopped firing after a release, duplicate tracking that doubles the numbers, funnels that skip a step users actually take, and attribution that credits the last click for work the first one did. The audit tells you which of your existing numbers you can trust before you act on any of them.

How long before we see results from CRO?

That depends on your traffic, not on us. An experiment needs enough visitors to reach significance, and on low-traffic pages that can take longer than the change is worth — in which case we will say so and recommend fixing the obvious friction directly instead of testing it. We tell you the required sample size before running anything.

This section of your site has no case studies. Why?

Because we do not have client-approved numbers for this work yet, and we do not publish metrics a client has not signed off. It is the thinnest evidence base of our five directions and we would rather say that than fill the space with something unverifiable.

Do you implement the fixes you recommend?

The audit itself is advisory and stops at the recommendation. Implementation goes through our Product Engineering direction, and the two are often bought together — but you are free to take the audit to your own team, and the recommendations are written to be actionable without us.

Think this is your problem?

Book a call or ask for an estimate. Both are free and neither commits you to anything.