Skip to content

Marketing & Growth

Analytics & CRO

Tracking, dashboards and conversion testing

Every marketing decision you make is only as good as the data behind it, and in our experience most companies' data is quietly wrong. Conversions counted twice, a form submission that fires on page load, purchases missing because the payment gateway redirects away before the tag runs, three tools reporting three different revenue figures. Everyone knows the numbers do not agree; nobody has time to find out which one is right.

Fixing that is the least exciting and highest-value work in marketing, because it is what every other decision is built on. Then conversion optimisation is what you do with data you can trust: making more of the traffic you already have turn into customers, which is almost always cheaper than buying more traffic.

What's included

Analytics and measurement

  • A measurement plan first.

    What the business needs to know, which events represent that, and what each is named — written down before anything is implemented, so the tracking answers questions rather than collecting whatever was easy.

  • Implementation and repair.

    Analytics, tag management, conversion tracking and platform pixels, implemented correctly and verified end to end rather than assumed.

  • Server-side tracking

    where accuracy matters, since browser-based tracking now loses a significant and growing share of events to privacy features and blockers.

  • Consent management.

    Cookie consent that actually controls what fires, and analytics configured to stay useful under it — a legal requirement in most jurisdictions and a technical exercise in practice.

  • E-commerce and revenue tracking

    reconciled against your own back office, because a platform's self-reported revenue is a claim, not a fact.

  • Attribution modelling

    that is honest about its limits. We show the platform-reported view and the reconciled view, and explain where they diverge, rather than presenting one number as truth.

  • Dashboards for the people who use them.

    One page per audience: the number, the trend, and what changed. Not a forty-widget report that gets opened once.

  • CRM and offline conversion integration

    so a lead that closed three months later is connected back to the campaign that produced it — which is the point at which lead generation becomes measurable at all.

Selected clients in Marketing & Growth

4 clients

  • Tramontina
  • Pedia Sure
  • CFI
  • HTU

Conversion rate optimisation

  • Funnel analysis. Where people leave, on which device, in which language, and what they did just before. This alone routinely finds problems nobody knew existed.
  • Qualitative evidence. Session recordings, heatmaps, on-site surveys and usability testing — the "why" behind a drop that the numbers only describe.
  • A prioritised hypothesis backlog, ranked by expected impact, confidence and effort, so the work order is a decision rather than a preference.
  • A/B and multivariate testing, run to a predetermined sample size and duration, with the result reported honestly — including the tests that show no difference, which are the majority and are still information.
  • Landing page and checkout optimisation, where the highest-value changes usually are.
  • Form optimisation — field count, sequencing, validation and error handling, which is often the single cheapest improvement available.
  • Performance work as conversion work. Page speed is a conversion variable, not a technical vanity metric, and it is one we can actually fix.
  • Personalisation, where the traffic volume justifies it — including simply showing the right language and currency, which many bilingual sites still get wrong.

Why testing needs discipline

Most conversion "wins" reported in this industry are noise. A test called after three days because the numbers looked good will be reversed the following week, and a programme run that way produces a series of confident changes that cancel each other out.

So we fix the sample size and the duration before starting, run for full business cycles to avoid weekday effects, test one meaningful change at a time, and record every result — including the failures — in a shared log. Over a year that log becomes genuinely valuable: a written record of what your specific customers respond to, which no agency's general best practice can substitute for.

Where traffic volume is too low for statistical testing — which is common — we say so, and we use qualitative research and expert review instead of pretending a test can conclude. Running an underpowered test is worse than running none, because it produces a confident wrong answer.

On working with Codigoo

We had been blaming regulation for two years. They spent a week watching recordings and showed us it was question order.
Mohsin Rahman, Head of Product · Tazkara Pay

Questions we are usually asked

Where should we start?

Almost always with tracking accuracy, because everything else is being judged against it. It is usually a short piece of work, and it frequently changes the picture of what is working enough to redirect a budget immediately.

How much can conversion rate improve?

It depends entirely on where you start and what is broken, and any specific figure quoted before looking at your funnel is invented. What we can say is that the first round of work on a site that has never been optimised typically finds obvious problems — a broken step on mobile, a form nobody can complete, a payment method your market expects and you do not offer.

Do we need expensive tools?

Rarely. Most of what matters can be done with well-configured standard analytics and a testing tool. Companies routinely pay for enterprise platforms while their basic conversion tracking is broken, which is spending in the wrong order.

Can you fix what you find, or only report it?

We can fix it. This is the reason this service sits inside a development house: a recommendation that needs engineering time it will never receive is a document, not an improvement. We change the code, measure the result, and keep the record.

If your analytics and your accounts disagree about how many sales you made, start there — a consultation will tell you how far off the numbers are.

Talk to us about Analytics & CRO

Thirty minutes with an engineer and a strategist who do this work — not a sales team. You will get an honest answer about whether it is the right thing to buy, and what it would realistically cost.