How Do You Audit a 400-Article Shopify Blog Without Breaking It?
If your Shopify blog has been running for the better part of a decade, it is almost certainly carrying damage you cannot see just by clicking through it. Old campaign tracking links go dead when a third-party tool is retired. Product photos get replaced during a catalogue refresh, leaving old articles pointing at colourways or SKUs that no longer exist. Articles written years ago sometimes get republished or backdated incorrectly, so a 2015 award announcement shows up looking like it happened last month.
None of this shows up in a normal editorial read-through. The page still looks fine. The problem only surfaces when a customer clicks a "shop this look" link and lands on a 404, or when Google recrawls the page and quietly drops it a few ranks for pointing at dead inventory. We recently ran exactly this kind of audit and repair pass across a 400+ article legacy blog, and the process is worth documenting because most of the failure modes are invisible until you build a tool that looks for them specifically.
Why manual review doesn't catch these bugs
The two most damaging bugs we found were both HTML defects that render perfectly well in a browser while being structurally broken. The first was a nested anchor tag: an image wrapped in a link, sitting inside a product card that was itself wrapped in a link to a collection page. Browsers silently auto-close the outer tag the moment they hit the inner one, so the layout still looks intact, but the "shop the collection" link around the whole card is dead. A tag-balance checker won't catch this, because both tags still have a matching close; they just nest illegally.
The second was colour and SKU drift. An inline product image and its link both pointed at what looked like a live product page, but the catalogue had moved on: that exact colourway had been discontinued, and the image URL happened to prefix-match a different live variant's filename closely enough that nobody caught it on a visual skim.
Catching either of these takes two things a human editor doesn't have on hand while reading a blog post: a full HTML structure parse, and a live snapshot of the current product and collection catalogue to check every link against.
Building a QA toolchain instead of reviewing by eye
Rather than manually re-checking hundreds of articles, we built a Dockerised QA tool that runs against a full CSV export of the blog, pulled via the Matrixify app. It checks every article for three classes of problem.
Structural HTML integrity comes first: unclosed tags, nested anchors, the malformed float and clear divs that accumulate in any template a store has outgrown twice. Then import packaging, which is the unglamorous stuff that silently breaks a bulk upload. UTF-8 BOM presence, exact column naming, correct blog-handle values, command flags. A single mis-cased field here can import six articles under the wrong blog and quietly break every cross-link pointing at them, which is not the sort of thing you notice until someone complains months later.
The third check is the one that took the longest to build and has paid off the most: live-catalogue validation. Every product and collection link in every article body gets cross-referenced against a current export of the store's actual live products and collections, so dead SKUs and retired colourways get flagged before they ship rather than after a customer finds them.
The tool also renders a side-by-side review page, showing every article's rendered body next to its flagged issues, so a human still makes every final call but reviews one clean page instead of opening each article individually. Once the nested-anchor check went in, it started catching that bug class automatically on every future batch, not just the one that surfaced it.
Reviving old content safely
Not every fix in a legacy blog audit is about breakage. Some of the highest value work is reviving content that was quietly buried. In this project, that meant scanning every article, published and hidden, for publish-date mismatches against the actual event described (an award announcement dated years after the award, for example), and identifying a batch of genuinely good, on-brand heritage content that had never been published at all.
The rule we apply here: never touch dates or visibility on old content without checking it against source-of-truth data first. A misdated article fixed carelessly can do more SEO damage than leaving it alone. Every backdate and republish in this audit was cross-checked against the original event date before anything was changed.
Treat bulk content edits like code deploys
Once a blog crosses a few hundred articles, "review it by eye before publishing" stops being a QA strategy and becomes a hope. Software teams solved this problem decades ago: when you find a bug, you write a test for it, and the test runs forever after. Content deserves the same treatment. Every defect we caught by hand in this audit became a permanent check in the tool, which means the next 400 articles get the benefit of this bug hunt without anyone repeating it.
If you're sitting on a large legacy blog and suspect it has quietly accumulated dead links, wrong images, or buried content, talk to us about a content integrity audit.
References
- Matrixify. (n.d.). Import and export data for Shopify. matrixify.app
- Google Search Central. (2025). How Google discovers, crawls, and indexes web pages.
- Shopify Help Center. (n.d.). Managing blog posts and articles.
