What accessibility evidence should a Shopify brand capture from every theme update?
A theme update can silently erase months of accessibility work. Capture a dated evidence set before and after every update: automated scan results, manual keyboard checks on key templates, and a change log that ties each finding to a fix. This is the paper trail that turns "we care about accessibility" into something defensible.
Why theme updates are the danger zone
Theme updates replace templates wholesale. Custom accessibility fixes baked into the old theme, focus styles, ARIA labels, heading order corrections, can vanish when the new version overwrites the files. App embeds and custom sections may also behave differently under the new markup. The store can go from clean to broken in a single deploy, with nobody noticing until a customer complains or a demand letter arrives.
Capture the before state
Before applying the update, freeze the current accessibility state. Run an automated scan across the key templates and save the dated report. Note the theme version, the date, and any known issues already tracked. This baseline is what the after state gets compared against, and it proves the store was in a known condition before the change.
Capture the after state
After the update, on a staging or preview theme first, repeat the same scan and diff the results. Every new finding is a regression until proven otherwise. Then do the manual checks automation cannot: keyboard through the homepage, a product page, the cart, and checkout. Test the mobile menu, the search, and any custom sections the update touched.
Log each finding with the template, the WCAG criterion, and a screenshot or description. Assign each one a fix and a retest date. The log does not need to be fancy. It needs to be dated, specific, and complete.
Keep the change log with the theme
Store the evidence with the theme version it describes. When the next update comes, the log shows what was fixed last time and what to watch for this time. Over a year, the accumulated logs tell a story of continuous attention that no single audit can match.
If a demand letter ever references the period after an update, the evidence set answers the only question that matters: did the brand check, find, and fix? A dated yes, with the scans to prove it, is the strongest position a store can be in.
When the update breaks something major
Sometimes the diff is not a handful of findings but a wholesale regression: the new theme handles navigation, forms, or product options in a fundamentally less accessible way. When that happens, the evidence set becomes a decision document. The brand needs to weigh the update benefits against the accessibility cost, and the log makes that tradeoff explicit instead of invisible.
The options are to fix the new theme before launch, to delay the update until the fixes are in, or in rare cases to stay on the old theme version. None of these are free, which is exactly why the evidence matters. A brand that can show it measured the regression, priced the fix, and chose deliberately is in a far stronger position than one that shipped blind and hoped.
Whatever the decision, record it with the evidence. Future updates, future team members, and if it ever comes to it, future legal scrutiny, all benefit from a record that shows the brand was paying attention.