Shopify redesign or CRO: which should you do first?

Should you redesign your Shopify store or optimise it?

The honest answer depends on what is wrong. The two jobs fix different things, and picking the wrong one is the usual way to spend months and see no change in orders.

A redesign replaces the theme or large parts of it: layout, templates, styling, often the code underneath. Conversion rate optimisation (CRO) leaves the structure alone and changes specific parts of the buying path, one at a time, measuring each against a control. A redesign is a big bet with one result at the end. CRO is a series of small bets that each report back.

So the question is not which is better. It is whether the problem you have is in the foundation or in the details.

When does a redesign make sense?

A redesign earns its cost when the foundation is the problem. These are the signs:

• The store is slow for reasons the theme causes. If the theme ships a large amount of JavaScript, loads fonts and sliders nobody uses, or builds every page from heavy sections, trimming apps will only go so far. In one build of ours, the rebuilt theme brought Largest Contentful Paint from 4.1s to 1.2s and cut the JavaScript shipped from 2.4MB to 310KB. Google rates an LCP of 2.5 seconds or less as good, per web.dev, so the first number was well outside the line and the second well inside it. That kind of gap is hard to close with settings.
• The mobile experience is broken in ways that need structural change. Tap targets, a buy button that sits out of reach, a gallery that does not behave on a phone.
• The theme is hard to change. Heavy customisation on an old theme means every edit risks breaking something, and upgrades are not possible. Our guide to theme customisation versus apps covers how that happens.
• The brand and the offer have moved. If you reposition the range, change the audience or add a major product line, the old structure may not fit what you now sell.

If two or more of these are true, no sequence of small tests will fix them. Fix the foundation first.

When is CRO the better first move?

CRO is the better move when the theme is sound and you can see where buyers stop. The signs:

• Core page speed is acceptable on mobile and the theme is maintainable.
• The funnel shows one or two clear weak steps, such as add to cart or checkout completion. Our guide to reading a low conversion rate shows how to find them.
• You have enough traffic to test. How much you need depends on your baseline rate and the smallest lift you care about, as set out in how to A/B test a Shopify store.
• You want to know why something worked, so you can apply it elsewhere.

A single change can matter a lot. In one test we ran, one variant showed the shipping cost on the product page while the control showed it at checkout. Each group had 24,105 sessions. The control produced 265 orders and the variant 578, a +118% difference in orders per session. A sticky buy box test finished at +31%. Each test changed one thing on the page. Those results are specific to where they ran, and yours will differ, but they show what a narrow, measured change can do when it removes a real point of friction.

What does a redesign put at risk?

A redesign is a change to a live store that already earns. Three risks are worth naming before you sign anything.

Attribution. If you launch a new theme and orders rise, you do not know which change did it. If they fall, you do not know either. You have swapped a working system for an unmeasured one. The way to limit this is to record a baseline before the switch (conversion by device and by source for at least four weeks) and, where traffic allows, to run the old and new themes against each other rather than replacing one with the other.

Search visibility. If the new build changes URLs, titles, structured data or internal links, search engines have to relearn the site. Keep URLs the same wherever you can, map any that must change, and follow a written plan. Our migration SEO checklist covers freeze, map, redirect and verify.

Regression. A new theme can quietly drop things the old one did: tracking events, schema markup, tested elements that were working. List what the current store does well, including the parts nobody designed on purpose, and check each is still present before launch.

CRO carries a different risk. A badly run test can mislead you, for example by stopping early when one variant looks ahead. The CRO roadmap shows how to set the stopping rule before launch.

What is the order when you need both?

Most stores that ask this question need a bit of both. The order that wastes the least:

1. Measure first. Funnel by step, split by device and source, speed on mobile. This takes days and settles the argument.
2. Fix the foundation if it is broken. If speed or mobile structure is the problem, rebuild or repair that part with a speed budget, and keep the URLs stable.
3. Let the baseline settle. Give the new build a couple of weeks of traffic so you have a clean starting point.
4. Test, one change at a time. Start with the weakest step in the funnel.

This also changes what a redesign should be asked to do. It should be built for testing: sections that can be changed without a developer, clean events for analytics, and room to try variants without breaking the rest.

How do you decide for your own store?

Answer three questions in writing.

1. Is speed the problem? Check the product page on mobile against the LCP threshold above. If the theme is the cause, that points to a rebuild. If apps are the cause, see the third-party scripts guide first.
2. Where does the funnel leak? If one step is clearly weak, test there before changing the whole store.
3. What would a redesign put at risk? If the answer is a lot, such as strong organic search or a store that converts well in some segments, the case for a full redesign has to be stronger.

If you cannot answer the first two from your own data, you are not ready to choose, and the first piece of work is to measure.

What should you fix first, and how long does it take?

Measuring takes a week. That alone often shows which path you need. If the foundation needs work, scope the repair against a speed budget and agree the list of things the new build must keep. If it does not, pick the weakest funnel step and queue the first test.

Our Store Teardown does this on your own store: a written audit of the conversion path, with the fixes ranked by likely revenue impact, so you can see whether you need a rebuild or a test queue before you commit to either. The services page lists the build and optimisation work that follows, and the Shopify CRO agency page covers the ongoing testing side.