04 · CRO · 04/06

Conversion

Q: What's the difference between UX audit and A/B testing?

[UX audit](/en/conversion/audit-ux) answers « where it breaks and why », based on session replays and heatmaps. A/B testing answers « does this fix actually win ». The audit poses hypotheses, the test validates or discards them. The two don't replace each other, they build on each other.

Q: How do you measure CRO mission ROI?

On two axes. One, **conversion rate gained** (before/after, per funnel step). Two, **additional revenue** at constant traffic (typically +10 to +30 % over 6 to 12 months for a well-run CRO program). We compare against your acquisition cost to calculate net ROI: one conversion point gained is often more profitable than ten points of [paid acquisition](/en/acquisition-payante) budget added.

At a 0.8% conversion rate, moving to 1.2% means +50% of sales without spending an extra euro on ads. Conversion is that math.

UX audit A/B testing Landing pages

02 · Lived scene

Friday evening. You open your weekly leads dashboard :

"1,500 visitors. 47 form submissions. 12 qualified leads."

How many dropped before the 4th field?

When you lengthen the form, you collect more signal, but you lose 40 % of leads at the 5th field.

When you shorten it, you get more leads, but your sales rep wastes 2 h qualifying junk.

When you redo a landing, you start from intuition, not from session-replay. You may rebuild the same mistake.

Your form is a blind bet. Not a funnel.

Our method to transform your traffic

03 · The path we follow

Seven steps to transform your funnel.

01 / 0702 / 0703 / 0704 / 0705 / 0706 / 0707 / 07

001 · AUDIT

102 · TRACKING

203 · HYPOTHESES

304 · TESTS

405 · DECISION

506 · ITERATION

607 · OWNERSHIP

We look at where your funnel leaks, and by how much.

We audit what's in place: the tools (analytics, A/B testing, session replay, heatmaps), the tech (CMS, front-end framework, cache infra), the channels driving traffic, and the sales process handling conversion (qualification, support, returns). Without that complete picture, you'd test hypotheses without knowing the terrain.

No test without clean measurement.

If your tracking is broken, the test will be false. We check that conversion events fire cleanly: transaction, lead form, funnel step. If they don't, we switch to our server-side tracking service before any CRO testing.

We prioritize via PIE score, not by instinct.

Not all hypotheses are equal. We rate them via PIE score (Potential gain · Importance for the business · Effort required). We test first what combines high impact and reasonable effort. Not the trendy test, the one that pays.

A/B rigorous, 95 % threshold minimum.

We run tests on a strict protocol: control group, variant group, sample size calculated upfront, minimum 14 days to absorb seasonality. We don't stop a test because it's starting to win. We wait for 95 % statistical significance minimum.

We win, or we learn. Never "maybe it works".

By test end: either the variant wins clearly (significance reached), we deploy. Or it fails, we drop the hypothesis. Or the result isn't significant, we document the learning and move to the next test. No middle ground.

The loop runs, the rate climbs in steps.

Once the winning variant is deployed, it becomes your new baseline. The next test attacks the next prioritized drop zone. Over 6 months, that's 8 to 12 rigorous tests, of which 30 to 40 % win. Conversion rate climbs in steps, not in leaps.

Your team pilots the roadmap, we stay on backup.

Full documentation: prioritized hypotheses roadmap, A/B testing accounts with separate accesses, decisions runbook. Your team or your next agency can pick up the cadence without starting over. We stay on backup as long as you need it. We leave the moment you don't.

Shall we book a moment?

04 · Our conversion specialties

Three levers to transform your traffic.

Three sub-services that build on each other: we look at what slows visitors down (UX audit), we test what could win (A/B testing), we rewrite the pages where decisions happen (landing pages). None of them holds alone. An audit without a test is just an opinion; a test without an audit hits in the dark. We deploy them together, in that order.

UX audit

We record visitor sessions (without identifying them) and map their clicks on your pages. You see where they stop, what they can't find, and where they leave. Not a hunch. Recorded proof.

A/B testing

We compare two versions of your page (original vs new) and watch which one pays more. One test at a time, picked for its potential. We wait for the result to hold before deciding. No premature stops, no bullshit.

Landing pages

We rewrite the pages where decisions happen: your purchase flow, your quote form, your sign-up. Clear promise, smooth path, customer testimonials with numbers.

05 · Where it held

+17 % leads. Marketing budget unchanged.

Amecam · pro vehicle outfitting · B2B Wallonia · 3 years with HeySquad.

+0 %leads in one year2024 → 2025 comparison on the Amecam site contact form, at constant marketing budget.

0 €marketing budget addedGain achieved through conversion redesign + CRM integration, not additional paid spend.

Amecam outfits professional vehicles: mobile workshops, vans, food trucks, mobile offices. B2B and B2C combined market, multi-thousand-euro basket, long sales cycle. Before we took over, the contact form mixed every request type together, the sales team got noise, and lead follow-up was lost between email and CRM.

We rebuilt the landing pages by request type (mobile workshops, leisure vans, food trucks, accessories) with a clear promise and a contextual form. We tested the contact form in A/B (required fields, order, microcopy). In parallel, we wired the form directly into Odoo: each lead arrives with its context (request type, declared budget, geographic zone), the sales rep prioritizes without untangling emails.

2024 → 2025 results: +17 % leads incoming at constant marketing budget. The request type comes through clean in Odoo, the sales team knows which segment they're talking to before the first call, and follow-up happens in the CRM rather than the inbox. Zone distribution (Liège dominant at 48 %) frames the next paid campaigns without guesswork.

Read the full mission Discover all missions

Amecam-built professional vehicle outfitting

06 · Let’s talk

We work directly. No sales intermediary.

Tell us what you’re looking for. We come back to you for a first chat. No pitch, no slides: we check if it’s a fit.

You write to the squad, not to a sales rep.

FAQ

Frequently asked questions.

How many A/B tests can you run in parallel?

Depends on traffic. For statistically significant results without diluting conversions, minimum 5,000 visitors per month per variant on a CTA test, more for a full landing page test. We prioritize via PIE score (Potential · Importance · Ease): one prioritized test at a time rather than five tests cancelling each other out.

What's the difference between UX audit and A/B testing?

UX audit answers « where it breaks and why », based on session replays and heatmaps. A/B testing answers « does this fix actually win ». The audit poses hypotheses, the test validates or discards them. The two don't replace each other, they build on each other.

How long until we see conversion results?

First complete audit in 2 to 3 weeks. First A/B test launched after 4 to 6 weeks (time to verify tracking and prioritize hypotheses). First statistically valid result after 4 to 8 weeks of testing depending on traffic. A serious CRO program runs over 6 to 12 months, not over a sprint.

What tools do you install?

For audit: Hotjar or Microsoft Clarity for session recordings and heatmaps. For testing: VWO, Optimizely or Google Optimize depending on your stack. For analytics: we rely on your GA4 if clean, otherwise we move upstream to our server-side tracking service.

What type of sites do you work with?

E-commerce (Shopify, WooCommerce, Prestashop), B2B with quote-request forms, SaaS with sign-up funnels. The method is the same: audit, prioritized hypotheses, rigorous tests. The difference is in the minimum traffic volume required for reliable tests.

How do you measure CRO mission ROI?

On two axes. One, conversion rate gained (before/after, per funnel step). Two, additional revenue at constant traffic (typically +10 to +30 % over 6 to 12 months for a well-run CRO program). We compare against your acquisition cost to calculate net ROI: one conversion point gained is often more profitable than ten points of paid acquisition budget added.

Who this service is for

4 Belgian contexts where CRO pays off.

E-commerce with a catalogue to scale

You sell online, your traffic exceeds 5,000 visitors/month per funnel step. You can test rigorously, measure step-by-step gains, scale what pays. Without that volume, we wait or target the test on the busiest zone.

eg: Wallonia pure-player 8-50K visitors/month, basket 50-300 €

B2B with a quote-request form

You collect B2B leads through a form. Too short and you get noise, too long and leads drop off. We audit the funnel step by step, test form versions, plug qualification into your CRM (HubSpot, Salesforce, Odoo).

eg: consulting firm 100-300 leads/month, lead-to-deal 10-25 %

SaaS with a sign-up funnel

You sell a self-service product. Sign-up → activation is your main drop point. Funnel audit, tests on the first step (email vs OAuth, price vs trial), pricing page rewrite.

eg: B2B SaaS 10-50K visitors/month, sign-up conversion 3-8 %

Catalogue site with multiple product pages

You have a catalogue of dozens of product pages. UX audit by category, optimisation of high-traffic pages, redesign of dedicated landing pages for paid campaigns.

eg: Wallonia retailer 50+ product pages, average basket 100-500 €

so? shall we settle in?

We're ready to turn your existing traffic into business numbers. We get back to you for a first chat, no pitch.

Get in touch