Technical SEO is invisible right up until it quietly caps every other investment you have made in the site.

We regularly inspect sites where the content is genuinely strong — clear, specific, well-organised — and the rankings simply do not reflect it. The pattern underneath is almost always the same: a technical issue quietly capping visibility that no amount of additional content or backlinks will fix, because the problem sits beneath the content layer entirely. A blocked page cannot rank regardless of how well it is written. A page that takes nine seconds to become interactive on a mid-range phone loses visitors before a single word gets read.

Technical SEO has a branding problem: it sounds abstract and low-priority next to a shiny new content calendar or a paid campaign. In practice it behaves more like plumbing than decoration — invisible when it works, and quietly capping everything built on top of it when it does not.

This checklist is organised the way we actually audit a South African site: crawlability and indexation first, performance second, structure and markup third, then the hosting and infrastructure decisions that generic global checklists tend to skip entirely.

• robots.txt does not accidentally block important sections of the site
• XML sitemap exists, is submitted in Search Console, and lists only canonical, indexable URLs
• No important pages are orphaned — every page you care about is reachable through internal links
• Duplicate content is resolved with canonical tags, not left to compete against itself
• Redirects use 301s for permanent moves; redirect chains are cleaned up, not stacked three deep
• A 404 page exists and behaves correctly — it should not silently return a 200 status
• Search Console shows no unexplained spikes in "Excluded" or "Crawled — currently not indexed" pages

A site that loads quickly on office wifi and a fast laptop can still perform poorly for the visitor who matters most: someone on a mid-range Android phone, on mobile data, possibly on a network under load. Test performance using a throttled mobile connection profile, not just your fastest available device, because that gap is exactly where South African sites tend to quietly lose visitors without anyone noticing in a dashboard built around desktop assumptions.

Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift are not abstract scores to chase for their own sake — they correlate directly with whether a visitor sees your key content quickly, can interact without frustration, and does not experience the page jumping around as it loads. Compress and correctly size images, defer non-critical scripts, and avoid loading a dozen third-party tracking tags that each add their own delay.

Data cost is a genuinely South African consideration layered on top of speed. A heavy, image-bloated homepage is not just slow — it is expensive for a visitor paying for data by the megabyte. Optimising for speed and optimising for data-consciousness usually point in the exact same direction: fewer, smaller, better-compressed assets.

Schema markup — Organization, LocalBusiness, Article, FAQPage, Product where relevant — gives search engines and AI answer systems a machine-readable version of facts that would otherwise need to be inferred from visible text. That reduces the chance of a business being misrepresented in a summary and can enable enhanced result formats in classic search.

Treat schema as a living part of the site, not a one-time technical task ticked off at launch. When a business changes address, adds a service, or updates pricing, the structured data needs updating in the same breath as the visible page copy — conflicting facts between visible content and schema markup undermine trust in both.

Validate markup with Google’s Rich Results Test or Search Console’s enhancement reports periodically. Broken or outdated schema is easy to introduce during a redesign and easy to miss unless someone checks deliberately.

• Every page has a unique, descriptive title tag and meta description
• One clear H1 per page; heading hierarchy is logical, not just styled to look right
• Internal links use descriptive anchor text, not repeated "click here"
• Images have meaningful alt text, both for accessibility and for image search visibility
• HTTPS is enforced sitewide with no mixed-content warnings
• Mobile usability report in Search Console shows no unresolved issues
• Organization, LocalBusiness, Article and FAQPage schema present and validated where relevant

Where a site is hosted affects real-world load times for South African visitors. Hosting on infrastructure physically distant from your primary audience adds latency to every single request, an effect global checklists written for other markets rarely mention because it is often not their bottleneck. Where local hosting or a well-configured CDN is available and cost-effective, it is worth the evaluation.

Uptime resilience deserves specific South African attention. Load-shedding and localised power or network instability can affect origin servers, office-based staging environments, or third-party services a site depends on. A managed hosting or platform provider with genuine redundancy reduces this risk considerably compared with infrastructure that depends on a single office connection staying up.

Set up uptime monitoring with alerting, regardless of hosting choice. A site that goes down for two hours during business hours and nobody notices until a client mentions it is a preventable, embarrassing failure — and one basic monitoring tooling solves cheaply.

1. Crawl the siteRun a full crawl to catch broken links, redirect chains, duplicate titles and orphaned pages.
2. Check Search Console healthReview indexation status, mobile usability, Core Web Vitals report and any manual actions.
3. Test real-world speedRun PageSpeed Insights or a similar tool on a throttled mobile profile for your top five pages.
4. Validate structured dataConfirm schema is present, accurate and matches current visible content.
5. Review uptime and hostingCheck monitoring logs for outages or slow-response periods over the past quarter.

Technical SEO is not a phase you finish at launch. It is a maintenance discipline, and every redesign, plugin update or new integration is a chance to quietly reintroduce a problem you already fixed once.

Nexus technical SEO principle

Few South African SMEs have a dedicated technical SEO specialist on staff, and that is not automatically a problem — it becomes one when nobody is explicitly responsible for it at all. Assign the quarterly audit above to a named person, whether that is an internal marketing lead, a developer, or an agency retainer, and put a date on the calendar rather than leaving it as a vague ongoing intention.

When briefing a developer or agency on a redesign, ask directly how they will preserve current rankings and indexation through the change — a URL structure change, a new content management system, or a platform migration are the three most common moments a previously healthy site quietly loses ground. A specific migration checklist, agreed before work begins, prevents most of this damage entirely.

A services business noticed three previously stable pages slip out of the top ten within the same week, and the first instinct was to blame a Google algorithm update — the explanation that requires no internal accountability. A crawl of the site told a different story: a recent plugin update had quietly added a noindex tag to an entire page template during a routine site change nobody flagged as SEO-relevant.

The fix took twenty minutes once found: remove the tag, request re-indexing, and confirm in Search Console that the pages returned to "indexed" status. The rankings recovered within the following fortnight. The actual lesson was not about the specific plugin — it was that nobody on the team had a standing habit of checking Search Console after routine site changes, so a genuinely simple, fully reversible mistake sat undetected for a week.

This is the value of the quarterly audit cadence above in practice: most technical SEO damage is not caused by a mysterious algorithm shift. It is caused by an ordinary site change nobody thought to check against indexation, and it is almost always fixable in minutes once it is actually found.

Run a full crawl this week and start with the indexation issues — they are often the highest-impact fix, since a page that cannot be indexed cannot benefit from any other SEO work at all.

Test your top five pages on a throttled mobile connection, not your office wifi, and fix the worst offender first rather than trying to perfect everything simultaneously.

Pair this checklist with our local SEO guide and our piece on why website speed costs you customers for the fuller commercial picture technical health feeds into.