Capabilities

How the work runs.

Eight areas that usually sit in separate teams. None of them is the starting point. The starting point is the number the business needs to move.

Financial modelling

Acquisition cost, customer value, marginal return and forecast models, built before the budget is committed.

Commercial digital

Acquisition strategy and investment cases, worked backwards from the target.

Data and measurement

Tracking, attribution and the pipelines under them. BigQuery, SQL, GA4, GTM, Looker Studio.

Performance marketing

Google Ads and paid media at scale, optimised for the outcome the business pays for.

Websites, search and experience

Sites organised around what the business needs them to do. UX, CMS, conversion and SEO.

Marketing technology

CRM, platforms and operational systems wired together, so a customer keeps their history.

Automation and AI

Internal tools, reporting systems and AI-assisted workflows that take repetitive work out of the week.

Digital strategy

A commercial objective turned into a plan the available budget and team can actually run.

The method

The same loop, whatever the channel.

  1. Start with the target. What is the business trying to achieve, in a number?
  2. Understand the economics. What is another customer worth, and what can it cost to get one?
  3. Find the constraint. What is stopping the number moving? It is often not in marketing.
  4. Use the data. If you can’t say where a number came from, don’t build a strategy on it.
  5. Build what is needed. Use a platform where it works. Build where it doesn’t.
  6. Move faster. Technology and AI shorten the gap between insight and shipped change.
  7. Measure the result. The handful of numbers that change a decision, not thirty that don’t.
  8. Reinvest in what works. Keep it, stop the rest, and put the next dollar where the return is.

In practice

Commercial digital

Testing starts with a commercial question

There is an industry habit of testing to demonstrate activity. Every test I run starts from a question the data raised, and results are documented so the same test never gets run twice. Over time that becomes a library of what works.

Performance marketing

Campaigns are managed, not launched

The common model is to set a campaign up, let it run, and report at the end. By then the budget is gone. Every lever should stay visible and adjustable while the money is still in play, with attribution mapped before the campaign starts, not retrofitted after it.

Websites, search and experience

Evidence changes minds, argument rarely does

Telling a team their site isn’t working rarely lands. Watching a real user struggle through it, live, lands differently. I run UX sessions against defined personas while stakeholders watch. The output is a plan ranked by effort against outcome, and it gets actioned because the team saw the evidence.

SEO that earns its traffic

Most content strategies fail on volume over value. I justify each page against a commercial or audience intent before it gets written. Fewer pages, each doing real work, structured so they strengthen each other rather than compete.

Automation and AI

Build where it makes sense

Where an existing platform does the job, use it. Where it doesn’t, the usual answer is to log the gap and wait for a developer or an analyst. I would rather build it: the dashboard, the pipeline, the plugin, the page.

AI is what makes that practical now. It speeds up code, analysis and repetitive production. Deciding what is worth building stays with me.

See it applied