Technical SEO in an afternoon

5 min read

For a site under a few hundred pages, technical SEO is a short list of pass-or-fail checks: can Google reach every page, index it, and serve it fast enough on a phone. Pass them once, re-check them once a quarter, and spend the rest of your time on pages. The hundreds of warnings your audit tool shows you are mostly warnings.

The plumbing

I think of technical SEO as 'the plumbing'. Nobody praises it, everybody notices when it is blocked, and once it works you leave it alone. The industry that makes it feel complicated works on sites with a million pages, where crawl budgets and canonical conflicts genuinely cost money. Your espresso shop has sixty pages. Google can crawl sixty pages before lunch.

The rule: fix anything that stops Google reaching, indexing or showing a page, and ignore everything that merely could be tidier. The consequence is that this is an afternoon, not a project, and the afternoon repeats once a quarter. Everything on the checklist is free to check and most of it is free to fix.

Check How, free Pass
Indexed Search Console, Pages report. Also search site:yourshop.com Every page on your map is listed as indexed. None of them sits under "crawled, currently not indexed"
Crawlable Open /robots.txt No Disallow line touches your content. Nothing blocks CSS or images
Sitemap Open /sitemap.xml. Submit it in Search Console It exists, lists your real pages, and has no deleted URLs in it
HTTPS Type the http:// address It redirects to https:// and the padlock shows
Mobile Open three pages on your phone Text readable without zooming, buttons tappable, nothing cut off
Speed PageSpeed Insights, mobile tab Passes Core Web Vitals, or fails only on images you can compress today
One address per page Try www and non-www, with and without the trailing slash Everything redirects to one version. Each page has a canonical tag pointing at itself
Redirects Open five old URLs from before your last change Each one lands on the right page in one hop
Text in the HTML View the page source and search for a sentence from the body It is there. Some crawlers, AI ones included, do not run JavaScript
Structured data Rich Results Test Article or Organization markup validates. Optional, and last
1 CRAWL 2 INDEX 3 SERVE robots.txt allows itsitemap lists ita link points at itthe server answers indexed, not "crawled, not indexed"one canonical addressthe text is in the HTMLnot a duplicate of your own page httpsworks on a phonefast enough, images compressedredirects land in one hop can Google reach it?did Google keep it?can Google show it? Each gate is pass or fail. A page that fails gate one never reaches gate three, whatever the content.
Google's three stages as three gates, with the checks that belong to each. The afternoon checklist above is these twelve lines, in order.

Triaging an audit report. The tool returns a few hundred items. Sort them by one question: does this stop a page on my map from being crawled, indexed or shown? "Page blocked by robots.txt" on a product page: fix today. "Redirect chain" on an old URL that still gets links: fix today. "Missing alt text" on a decorative icon: ignore. "Title too long" by four characters: ignore. "Low word count" on the contact page: ignore, it is the contact page. "Not indexed" on the descaling guide: that is the only item on the list that matters, and it was on page nine of the report.

Worked example: sorting hundreds of warnings into two piles

The afternoon

  1. Open Search Console's Pages report and compare it with your map. Every page on the map should be indexed. The ones that are not go to the top of the list.
  2. Open /robots.txt and /sitemap.xml in a browser. A stray Disallow and a sitemap full of deleted URLs are the two most common blocks on small sites.
  3. Type your address with http, with www and without, with and without a trailing slash. Everything should land on one version in one hop.
  4. Open three pages on your phone and run them through PageSpeed Insights. Compress the images it names. Most small sites fail speed on images alone.
  5. View the source of your most important page and search for a sentence from its body. If it is not there, your text is loaded by JavaScript, and that is the one item on this page worth calling a developer for.
  6. Check five old URLs. Renamed sections, old blog posts, last year's product pages. Each should redirect once, to the right place.
  7. Put the date in the calendar for three months from now. Then close the audit tool and go back to the map.

When you need this, and when you don't

You need the afternoon once now, once a quarter after that, and once more after any redesign, platform change or domain change. Those three events break more plumbing than anything else.

You don't need a technical SEO retainer, a monthly audit subscription, or a fix for every warning. Call someone for three things only: content that only appears with JavaScript, a site migration, and a hacked site. Everything else on this page you can do yourself with a browser.

Next

  • Step 6 is done. Next: Run it, the weekly routine for one person, and the version for clients.
  • SitePerfector's site health check runs these checks on your site and flags the ones that block indexing: Site health.

Ready to grow your online presence?

Full access for 14 days — your topic map built, your gaps found, your content plan ready. Then Basic at $39/mo, locked for life for the first 50 members.

Try everything for 14 days — $7

$7 today · then $39/mo · cancel anytime