The 30-minute store quick-wins audit

It's the first Monday of the month and the owner of a store selling trail shoes has half an hour before the day starts. She could open a long audit report and feel busy, or run the same short check as last month on the same pages and leave with five things to fix, in order. This lesson is the second option: a fixed procedure that ends in a ranked list, not a score, and gets sharper every time you repeat it.

Fix the sample and the sheet first (three minutes)

A repeatable audit needs the same pages every time, so a change between months is a real change and not a different sample. Pick the URLs once: your five highest-revenue collections, your five highest-revenue products, the homepage and one buying guide. The landing page report from Tracking revenue from search and AI gives you the revenue order.

Open a sheet with five columns: URL, check, finding, effort, impact. Every row you add is a finding; the last three minutes rank them. Don't fix anything while auditing, or a short check becomes a lost morning with no list.

Block one: collections (ten minutes)

Collections catch category searches, so they come first. For each of the five:

1. Confirm a 200 status, a title tag that names the category in the words a buyer would search, and a canonical tag pointing to itself. The free On-page meta checker shows all three in one pass. A missing H1 is a finding; more than one isn't, because Google has said multiple H1s are fine.
2. Look for text that says what the collection is and who it's for. A grid of products under a bare heading gives Google nothing to rank, and it's the most common collection finding.
3. Open a sort or filter URL on the collection (anything with a question mark) and check its canonical points back to the clean collection URL. If it points to itself, every filter competes with its parent; Faceted navigation and filters covers the fix.
4. Check where it ranks for its head term with the free Google rank checker and note the position; next month's change in that number tells you whether a fix worked.

Write each failure as a row. Two minutes per collection is enough; if a check takes longer, write "unclear" and move on.

Block two: products (ten minutes)

For each of the five products:

1. The title tag leads with the product name and the attribute buyers search by (size, material, model), not the SKU or a slogan.
2. The description is yours, not the manufacturer's. Copied text earns no penalty (Google's documentation on consolidating duplicate URLs describes canonicalisation, not punishment), but it gives Google no reason to pick your page over the dozen others carrying the same words.
3. Product structured data is present, valid and matches the page: same price, same availability, same name. Paste the URL into the free Schema validator. Product earns a rich result in Google's structured data gallery, so a valid block is worth the time, and markup that disagrees with the visible price is the first thing to fix on that page.
4. The main image has descriptive alt text and isn't a print-sized original.
5. The URL is the canonical one. On Shopify, a product opened from a collection sits at a /collections/.../products/ path and the standard themes point the canonical at the /products/ version; check an app hasn't changed it. Variants follow Variants and duplicate content.

If one of the five is out of stock, note whether the page still returns 200 with honest availability or has quietly become a redirect; Out-of-stock and discontinued products gives the decision rule.

Block three: technical basics (four minutes)

These change rarely, which is exactly why they break unnoticed after a theme update or a new app.

1. Open yourstore.com/robots.txt and confirm nothing blocks /collections/ or /products/.
2. Open your sitemap and confirm a product added this month is listed.
3. Check the homepage and one collection against the Core Web Vitals thresholds on web.dev: Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint within 200 milliseconds and Cumulative Layout Shift under 0.1, at the 75th percentile of real visits. Search Console's Core Web Vitals report shows real-visit results; PageSpeed Insights tests one page. A page that passed last month and doesn't now usually means a new app or script, and Store speed and themes is the fix.
4. Open a discontinued product URL from last month's list and confirm it still redirects to the right place, once, not through a chain.

Turn the rows into a ranked list (three minutes)

A sheet of findings is not an action list. Rank it in two passes.

First, mark impact high for any finding on a page that earns revenue, or that stops a page being indexed: a wrong canonical on a top collection, a blocked folder, a price mismatch in schema. Mark it low for anything cosmetic on a page that earns nothing.

Second, mark effort low for anything you can change in the platform admin today (a title, a canonical setting, a description) and high for anything that needs a developer.

High impact and low effort goes to the top. Take the first five rows, give each a date, and leave the rest. Next month, run the same procedure on the same pages, check whether last month's five are gone, and let the remaining rows move up. The audit never produces a score, only a list that gets shorter.

What to take away

• Audit the same pages every month with the same checks, so a change between runs is a real change.
• Collections fail on thin text and filter canonicals; products fail on copied descriptions and schema that disagrees with the page.
• Technical basics break silently after theme and app changes, so check robots, sitemap, vitals and redirects even when nothing seems wrong.
• The output is five dated actions ranked by impact and effort, not a grade.

Next

Take your ranked list into Days 1 to 30: fix the leaks, where the fixes get scheduled.