The hierarchy of CRO: the framework behind consistent ecom growth

Jordan Hill, Founder & CEO

Jordan Hill, Founder & CEO

This is the parent article of the That Works Hierarchy of CRO series. The full series: Base (1/3) Offer, Middle (2/3) Messaging, Top (3/3) UX & UI.

What is the That Works Hierarchy of CRO?

The That Works Hierarchy of CRO is a three-layer framework for prioritising conversion rate optimisation efforts in the right order: Offer first, Messaging second, UX and UI last. Developed by That Works Agency and modelled on Maslow's Needs hierarchy, each layer only performs when the one beneath it is solid. Brands that start at the top, obsessing over design and checkout flows, consistently underperform brands that build from the foundation up.


The three layers - in order:


They must be addressed in this order. Most brands start at the Top.

In short:

Fix your offer before your messaging, and your messaging before your UX. The layer below always determines how well the layer above performs.


Why most CRO programmes fail to move the needle

We've watched brands spend six figures with agencies that ran dozens of tests and had nothing to show for it.

The tests themselves weren't the problem. The issue was testing the wrong things - optimising the wrong parts of the business entirely.

The industry has conditioned brands to think CRO is a series of small cosmetic fixes. So that's where they focus: button colours, hero images, checkout flows. And that's exactly why they stay stuck.

A strong offer can compensate for mediocre messaging and average UX. Strong messaging can compensate for imperfect UX. But great UX, built on a weak offer and generic messaging, converts no one. You're building a house on sand and wondering why it keeps sinking.

Base: Your Offer

Why offer is the biggest lever in CRO

We've seen a brand increase their price by 10% and watch their conversion rate hold flat - because the higher price point repositioned them as premium. We've also seen a brand drop their price by 10% to chase volume, only to collapse their margins and destroy the paid acquisition strategy that depended on them.

Adjusting your offer creates a cascading effect across your entire operation. Yet most brands set their pricing once, early, and never revisit it. A 5% price movement in either direction can ripple through your margin, acquisition costs, LTV, and unit economics simultaneously.

What offer optimisation actually looks like

Offer goes well beyond price. It's the full shape of what you're giving your customers:

  • Are you bundling products to increase AOV?

  • Are you offering starter kits that lower the barrier for new subscribers?

  • Are you creating volume discounts that incentivise larger purchases?

  • Is your subscription model structured to maximise LTV, or just bolted on as an afterthought?

Real example: We worked with a supplement brand struggling to scale. Their product was solid - their offer structure was the problem. They offered both one-time purchase and subscription, and while giving customers choice seemed right, it was killing their ability to scale profitably. We removed the one-time purchase option entirely and went subscription-only. Conversion dipped around 8%. But a 100% subscription take rate meant LTV shot up, payback period dropped, and suddenly they could afford to be far more aggressive on acquisition spend. Same product. Same site. Different offer. Different business.

Why brands avoid testing their offer

Because it feels risky. It is risky. The stakes are real in both directions - which is exactly why it's the biggest lever.

Brands with outdated Shopify themes are printing cash because they nailed their offer. Beautifully designed stores with weak offers struggle to break even on ad spend. If you're not regularly testing pricing, bundles, subscription models, and product positioning, you're decorating.

For the full breakdown of offer optimisation, read The Anatomy of The Perfect Offer - Base (1/3) of the That Works Hierarchy of CRO.


Middle: Your Messaging

What good messaging looks like - and why most brands miss it

Once the offer is strong, messaging becomes the next lever. Most brands get this badly wrong.

Take collagen. A 25-year-old buying collagen is thinking about prevention - maintaining what she has. A 55-year-old is thinking about reversing visible signs of ageing. Same product. Completely different emotional drivers.

It goes deeper than demographics. Within that older group alone, you have sub-personas with different primary motivations: hair thickness, joint pain, skin elasticity. If your Meta ads speak to joint pain relief but your landing page only talks about skin benefits, you've just wasted your ad spend. The messaging disconnect kills conversion - no matter how compelling the offer or how polished the site.

What it takes to get messaging right

Most brands create one set of messaging and apply it to everyone. They talk about product features instead of the outcome their customers are actually buying. Generic benefit statements that could apply to anything in the category.

Good messaging requires:

  • Knowing your customer at a level most brands never bother to reach

  • Translating features into emotional benefits specific to each buyer segment

  • Deploying that understanding across dedicated landing pages for different personas, dynamic copy on PDPs, and testimonials curated to specific customer journeys

Hard work. Significant rewards.

The compounding effect: A strong offer with weak messaging will underperform. A strong offer amplified by messaging that genuinely connects your customer's problem to your solution - that's where the real gains live. The two compound each other.

For the full breakdown of persona-based messaging, read How to Build Personas From Ecom Data to Win More Conversions - Middle (2/3) of the That Works Hierarchy of CRO.


Top: UX and UI

Why 99% of brands start here - and why that's backwards

Most brands get this wrong - and the reasons are understandable.

Picking holes in a site's look and feel requires no deep customer insight or strategic thinking. Piecing together an offer and messaging strategy that fundamentally changes how a business operates requires both, and most brands haven't built that muscle yet.

There's a psychological pull here too. UX changes are visible. You can show them to your team and say "look, we're doing CRO." Testing your offer or messaging requires more nuance. The changes aren't always obvious on the surface - but the impact on your P&L is.

What UX testing actually delivers

Gains from UX and UI testing are real. A well-executed test can lift conversion by 5%, sometimes 10%. Those wins rarely change the trajectory of your business the way offer and messaging can.

A 5% lift from a UX change is good. A 40% lift from a new offer structure or messaging strategy is business-changing. We've worked with brands doing eight figures a year on Shopify themes that would make a designer wince, because their offer and messaging were doing the heavy lifting.

When UX optimisation is the right move

UX and UI are the polish. Buy area layout. Add-to-cart placement. How product images load. These things matter - only once you've built solid foundations beneath them. They're how you squeeze out incremental gains after the big levers have been pulled.

How agencies make this problem worse

Most Shopify CRO agencies have started offering A/B testing - taking the same surface-level site changes they've always made and split-testing them.

The value in CRO isn't in running tests. It's in the strategic knowledge around what and why something should be tested, and building programmes that compound results over time - starting from offer, moving through messaging, reaching UX and UI last.

Brands end up optimising the wrong things because their agencies never challenge the framing. Running visible, easy tests is good for agency optics. It's rarely good for brand growth.

Frequently asked questions

How do I know if my offer is the problem, or my messaging?

If traffic quality is good but conversion is low across all pages and audiences, the offer is usually the culprit - the product, pricing, or structure isn't compelling enough. If performance varies significantly by ad creative, audience, or landing page, messaging is more likely the gap. The personas and language on your best-performing pages will tell you what's connecting and what isn't.

Can I work on messaging and UX at the same time?

Yes - but test them separately. Running simultaneous changes means you can't attribute results to either. Run your messaging tests first on high-traffic pages, validate the shift, then layer in UX changes once the messaging is confirmed to be working.

How often should I revisit my offer structure?

At minimum, quarterly - and any time you see a meaningful shift in conversion rate, AOV, or LTV that isn't explained by traffic quality. Competitor moves, market conditions, and customer expectations all shift offer perception over time. Brands that set pricing once and never revisit it leave significant revenue on the table.

What's the first test I should run if I've never done CRO properly?

Start by auditing your highest-traffic landing page against your best-performing customer segment. Check whether the headline is written for who's actually buying and whether the primary motivation sits above the fold. Run a headline test that reframes the copy around your lead persona's actual driver. That single test will tell you more about your messaging gaps than a year of UX tweaks.

Does the That Works Hierarchy of CRO apply to all Shopify store sizes?

Yes - the order of priority holds at every scale. The difference is that at smaller scales (under £2M), offer issues are often the dominant constraint. At larger scales (£10M+), messaging sophistication and persona-specific funnels tend to be where the biggest untapped gains sit. UX polish is almost always the last lever to pull, regardless of size.


The bottom line

The That Works Hierarchy of CRO is a framework for building CRO programmes that compound - layer by layer, in the right order.

Offer first. Messaging second. UX last.

The brands that get this right are the ones that compound results, test after test, quarter after quarter. Go deeper into each layer with the full series:

Want to audit where your programme sits in the That Works Hierarchy of CRO? Get in touch →