WEBSITES

CRO and landing page optimisation based on data, not taste.

CRO starts with knowing where visitors drop off and why. We combine GA4, heatmaps, session replays and customer interviews into hypotheses that we A/B test. No button-colour tests because a management team member wants one. Every test is a grounded learning or a winner with a downstream effect on SQL and deal velocity. Under the same service we also build the campaign landing pages the tests run on, per ad group and per ICP.

Book a CRO conversation Take the Maturity Scan

What you get

+20%
Conversion lift in 6 months, on average
2-4
Tests per month on high-traffic pages
100%
Evidence-based hypotheses, no guesswork
1
Landing page template, unlimited variants per ad group

The observation

Most CRO programmes test random variants, not hypotheses.

Swap the button colour, reposition the header, pit two CTAs against each other. No quantitative audit, no customer interview, no linked business metric. We do it differently: we start with research within the broader websites offering, tie every test to SQL or deal velocity, and treat a losing test as just as valuable as a winner. For campaigns we also deliver landing page builds, because a test only works when the page it runs on matches the ad promise. These are the five patterns we see in almost every CRO and landing page audit.

No quantitative foundation

Starting tests without a GA4 funnel analysis, heatmaps or session replays. Hypotheses are more hoping than knowing where visitors get stuck.

No qualitative input

No customer interviews, no sales call transcripts, no exit survey. Tests guess at objections rather than hearing them.

Too little traffic for statistical power

Tests on pages with 200 visits per month, running for 4 weeks. Results are noise. This week's winner is next month's loser.

No link to deal or revenue

Tests win on micro-conversions such as scroll depth or click rate, but the effect on SQL, deal velocity or MRR is never measured.

No dedicated landing pages per campaign

Ad traffic lands on the homepage or a generic service page with four nav options and six proposition blocks. The ad promises problem A, the page talks about B, C and D. Bounce rate above 70 per cent, no meaningful A/B test possible.

Approach

From guesswork to a repeatable cycle, including landing page builds.

A fixed flow of research, hypothesis, test, learn. First audit and research take 3 to 4 weeks. After that a continuous rhythm of 2 to 4 tests per month, with a monthly report your management team can act on, linked to your CRM architecture. For campaigns we also deliver landing page templates with variants per ad group and ICP, so the page your tests run on matches the ad promise.

01 01

Audit

GA4 funnels, heatmaps, session replays and Search Console. Where is it leaking, on which pages and at which step.

02 02

Research

5 to 8 customer interviews, sales call transcripts and an exit survey. We want to hear the objection in the customer's own words, not invent it ourselves.

03 03

Hypotheses

For each hypothesis: an expected effect, a primary business metric and a power calculation. No test without a minimum threshold.

04 04

Test or landing page build

Preferably in HubSpot Content Hub experiments, in VWO or Optimizely when needed. For campaigns we build 1 HubSpot landing page template with editable modules per ad group or ICP, in 10 working days after a sharp brief. Tests run 2 to 4 weeks on pages with sufficient traffic.

05 05

Learn and scale

Implement the winner, document the loser as a learning. For each new ad group a variant in 1 to 2 weeks without starting from scratch. Monthly reporting with conversion lift and effect on SQL and deal velocity.

Who helps you

Three specialists on your CRO programme.

Data, content and build in one team. CRO is not a standalone activity but a continuous programme that touches web design, copy and development.

Sander Adriaanse

Webdesign / Frontend lead

Designs and builds the variants in HubSpot or the test tool. Ensures that A and B differ only on the variable being tested, not also on layout or imagery.
Dante Zwanenburg

Content & Ads specialist (lead)

Leads the CRO programme. Runs the funnel analysis, writes the hypotheses and delivers the monthly report. Feeds CRO results back into ad campaigns and lifecycle.
Carsten Huiskamp

CMS Architect

Sets up the experiments, GA4 events and HubSpot lifecycle connection. Makes tests measurable down to deal level. No tracking, no test.

Case · B2B mid-market

From random tests to a CRO programme that fills the management report

A B2B SaaS client had previously run A/B tests on the homepage themselves. Button colour, hero image, two CTA variants. Eight tests in a year, no measurable lift, no learnings recorded. Marketing consistently faced the question in management meetings of why CRO was not producing pipeline.

In the first sprint we ran a GA4 audit, five customer interviews and built a hypothesis backlog of fifteen tests prioritised by impact and effort, and built dedicated landing page templates with variants per ICP for the two heaviest ad campaigns. After six months the programme was running at 2 to 4 tests per month with a monthly report in management report format. Conversion lift came out at +22%, SQL volume increased by 18%.

What clients usually ask.

Honest answers to the questions we hear before every CRO programme.

Do we have enough traffic for A/B tests?

For a classic A/B test you need roughly 5,000 unique visits per month on the page template to reach a statistically reliable outcome within 2 to 4 weeks. With less traffic we work with sequential tests, multi-armed bandit approaches or lean more heavily on qualitative research. Sometimes the honest conclusion is: build traffic through marketing first, then test.

Which A/B tool do you use?

Preferably HubSpot Content Hub experiments, because results connect directly to lifecycle, contacts and deals within our HubSpot expertise. For more advanced setups (multi-page tests, personalisation rules outside HubSpot) we work with VWO or Optimizely. We choose based on scope, not on tool preference.

How often do you report?

Monthly. A report with completed tests, learnings, conversion lift and effect on SQL and deal velocity. Plus a forward-looking section with the upcoming hypothesis backlog, prioritised by expected effect and effort.

What if a test loses?

A losing test is also a learning. We document what the hypothesis was, why it failed and what it tells us about the target audience. Often more valuable than a winner, because it gives direction to future tests and to ad copy.

How do we know if we are ready for CRO?

Three checks. At least 5,000 unique visits per month on the pages that matter. Tracking in GA4 and HubSpot that you trust. At least 50 conversions per month on your primary goal. If not, the gain is first in tracking and traffic. Our maturity scan gives a first picture of that.

Do you also work on WordPress or other CMSs?

Yes. Tests run preferably in HubSpot Content Hub experiments via our HubSpot CMS development, but VWO and Optimizely work just as well on WordPress, Webflow or a headless setup. Hypotheses and programme discipline are platform-agnostic.

Do you also build the landing pages we run tests on?

Yes. Landing page builds have been part of the same service as CRO and A/B testing since 2026. Per campaign: 1 HubSpot landing page template with editable modules, variants per ad group or ICP. First template including copy research takes 3 to 4 weeks, every subsequent variant 1 to 2 weeks. Conversion rates of 4 to 8 per cent on B2B campaigns are achievable. Templates, copy framework and tracking are in the same scope, so the page your tests run on matches the ad promise.

How quickly can we have a new landing page live?

With an existing template and a sharp brief: 10 working days. The first landing page for a new campaign takes three to four weeks including copy research, template build and tracking setup. After that every subsequent variant is much faster, because we are not building from scratch but composing a variant within the template per ad group, ICP or intent level.

What does a CRO programme cost?

Our services are bespoke. A research phase with audit and initial hypothesis backlog is a one-off block. An ongoing programme usually sits within a RevOps-as-a-Service retainer in which research, design, build and reporting come together. Book a conversation and we will put together a quote based on your traffic and scope.

Ready to start?

Ready for CRO based on data?

Start with a conversation. We look at your traffic volume, conversion data and tracking. After that you get an honest scope. Sometimes the first step is not a test, but cleaning up tracking or building more traffic. We will say so honestly.