SEO advice is full of best practices, but not everything has the same impact.
What’s one recommendation that’s technically correct, but you rarely prioritize because the time invested just isn’t worth the result?
@arslnser Building out a perfect llms.txt file. It’s become trendy advice in AI-search circles, and there’s nothing technically wrong with having one — but I keep seeing people spend hours crafting it while their robots.txt is quietly blocking OAI-SearchBot, or their Product schema is JavaScript-injected and never reaches a non-JS crawler in the first place. A beautifully written llms.txt does nothing if the crawler can’t reach or read your actual pages to begin with. It’s a nice-to-have layered on top of the fundamentals, not a substitute for them — I’d always fix crawler access and schema rendering first.
Renaming every existing product image file to include keywords. It’s technically correct, but I’ve never seen enough upside to justify re-uploading a whole catalog, especially when Shopify CDN URLs and theme references can make cleanup messy.
I’d do this instead:
- Use sensible filenames only for new uploads.
- Compress images and serve WebP or AVIF where the theme supports it.
- Add useful alt text to the main 2 to 4 product images, describing what’s actually shown.
- Spend the saved time improving collection copy, internal links, and product titles. Those usually have much more influence on rankings and clicks.
Perfecting every eligible piece of structured data. Once accurate Product, Breadcrumb, and Organization markup is in place, adding more schema types rarely earns enough visibility to justify ongoing maintenance.I would first check whether the pages are indexable, linked from useful collections, and receiving impressions for relevant queries. Then improve the few product or collection pages that already show demand but have weak clicks or unclear shopper copy. If a schema addition supports content shoppers can actually see and Google can use, add it. Otherwise it usually sits behind those basics.
Manually hitting “Request Indexing” in GSC after every deploy, or wiring up IndexNow to ping the second anything changes. Both are technically correct and both feel productive because you get a little submitted confirmation. On a young domain though, indexing was never the actual bottleneck.
I’ve got pages fully indexed, canonical, no robots or schema issues, sitting at low single digit daily impressions. The technical checklist was done weeks ago. What’s missing is authority, more sites linking in, more signal the domain is worth trusting. Pinging Google faster doesn’t touch that. I’d rather spend the hour getting one real backlink than requesting indexing on pages that were going to get crawled within a few days anyway.
@arslnser
One for me is obsessing over achieving a perfect SEO audit score. Its technically good practice to fix every minor warning, but many of them have little to no measurable impact on rankings or conversions.
I’d rather spend that time improving product content strengthening internal linking optimizing site speed or earning quality backlinks. Those activities consistently deliver better results than chasing a 100/100 score in an audit tool.
SEO is about prioritization. Not every recommendation deserves the same amount of effort especially if it wont move the needle for users or organic traffic.
Hi @arslnser,
For me, the biggest waste of time is manually writing a completely unique meta description for every single product variant in a large store.
It is technically a “best practice” to have unique descriptions for every page. However, Google actually rewrites or ignores your meta descriptions most of the time anyway. It changes them dynamically based on exactly what the customer types into the search bar. Spending days writing hundreds of custom sentences for a red shirt versus a blue shirt is exhausting, and it rarely brings in more traffic.
Instead of wasting your time on small tasks that do not move the needle, you should just automate them. I highly recommend using SearchPie: SEO, Speed & Schema. It uses AI to instantly generate all those meta tags and image alt texts in bulk for your entire store. It handles the boring, repetitive SEO work automatically, so you can spend your time on things that actually grow your business, like building backlinks and writing great blog posts.
I would love to hear what SEO task you have completely stopped doing!
Good point. I agree that crawler access and indexability come before things like llms.txt. It’s an interesting addition, but the fundamentals still have the biggest impact.