Most SEO problems I see on business websites are not content problems. They are shipping problems. A page that Google cannot crawl, a template that repeats the same title on forty pages, a redirect chain nobody noticed. The marketer gets blamed, but the fix sits in the codebase.
The good news is that technical SEO is mostly a checklist, and developers are the best people to run it. You do not need to write a single blog post. You need to make sure the site you ship can be found, understood and loaded quickly. These twelve checks cover the failures that cost real traffic, in the order I would run them before any launch.
Crawlability and indexing
1. Check that robots.txt is not blocking anything important
It sounds basic, and it still happens. A staging rule like Disallow: / survives the move to production, or a well meant block on a folder takes the product pages with it. Open yoursite.com/robots.txt, read every line, and confirm that the only things blocked are areas you genuinely want hidden, like admin and internal search pages. Then check that the file points to your sitemap.
2. Generate an XML sitemap and keep it honest
A sitemap should list only pages that return a 200 status, are not blocked, and are the canonical version of themselves. Auto generated sitemaps usually handle this, but hand built ones drift. If a page is important enough to rank, it should be in the sitemap and reachable by links from the rest of the site. Orphan pages with no internal links are a smell, sitemap or not.
3. Set canonical tags, and respect them everywhere
Every indexable page should carry a canonical tag, even if it only points to itself. Then check the usual duplication traps: www versus non www, http versus https, trailing slash versus no slash, and print or parameter versions of pages. Pick one version, redirect the rest to it with a 301, and make the canonical agree with the redirect. When the redirect, the canonical and the internal links all say different things, Google picks for you, and it does not always pick the one you wanted.
On page signals your templates control
4. Give every page a unique, descriptive title tag
Title tags are generated by your templates, which means one bad template creates a site wide problem. Check that titles are unique per page, that the page topic sits near the front, and that nothing is truncated into nonsense on narrow screens. If your CMS falls back to the site name when a title field is empty, that fallback will quietly fill with duplicates. Find those pages before Google does.
5. Use one H1 and a sane heading order
One H1 per page that matches what the page is about, then H2 and H3 in order. Heading order is not a ranking magic trick, it is how both crawlers and screen readers understand the structure of the page. Skipping levels or using headings as a styling shortcut makes the document harder to parse for everyone.
6. Keep URLs short, lowercase and readable
A URL like /services/tax-filing tells a visitor and a crawler what the page is before it loads. Session IDs, dates and mixed case in URLs create duplicates and support tickets later. If you are migrating URLs, map every old address to its closest new equivalent and redirect it. Old links from other sites are an asset, and a 301 is how you keep them.
Speed and rendering
7. Compress images and give them dimensions
On most business sites, images are the page weight. Compress before upload, serve modern formats such as WebP where you can, and always set width and height so the browser can reserve space and avoid layout shift. Add descriptive alt text while you are there. It helps image search, it helps accessibility, and it takes seconds per image once it is a habit in the workflow.
8. Protect your LCP and load lazily below the fold
The largest contentful paint element is usually the hero image or the main heading block. Preload the hero image or mark it fetchpriority high, and do the opposite for everything below the fold: lazy load it. Then test on a mid range phone over a real mobile connection, not the office WiFi. A page that feels instant in the office can take ten seconds in the conditions your visitors actually have.
9. Ship HTTPS everywhere, with no mixed content
Every page, every asset, every form. Mixed content warnings scare visitors and can block resources from loading at all. After the certificate is in place, redirect all http versions to https and update internal links so they do not bounce through redirects on every click.
Structure and trust signals
10. Build a 404 page that helps, and monitor real 404s
A good 404 page keeps the visitor on the site: a clear message, a search box, and links to the main sections. Then watch your analytics or Search Console for URLs that return 404 and receive traffic or links. Fix the ones that matter with a redirect to the closest live page.
11. Add structured data where the page genuinely qualifies
If the business has a physical presence or serves a local area, LocalBusiness or Organization markup states the facts machines need: name, address, phone, opening hours, services. Article and FAQ markup help where the content matches. Add it, then validate with a testing tool instead of assuming the JSON is fine. Never mark up content that is not visible on the page.
12. Test like a crawler before you call it done
Run the finished site through Google Search Console URL inspection on a few key templates, check coverage reports after launch, and crawl the site yourself with a free tier tool to catch broken links, redirect chains and duplicate titles at scale. Fifteen minutes of this catches problems that would otherwise surface as a slow traffic decline nobody can explain.
Questions developers usually ask
Does any of this matter if the content is good?
Yes, because crawlers meet the technical layer first. Brilliant content on a page that is blocked, duplicated or painfully slow will underperform average content on a clean site. The two are not rivals. The technical work is what lets the content be seen at all.
How long do fixes take to show results?
Indexing changes can appear in days. Ranking improvements usually take weeks, because Google has to recrawl, reprocess and retest pages against competitors. Ship the fixes, verify them in Search Console, and judge on a month of data rather than a morning.
Where can I learn the business side of this?
If you want the full picture of how these technical pieces fit into a wider strategy for a local market, this complete guide to SEO in Afghanistan walks through keyword research, service pages, Google Business Profile and measurement in the same practical style, written for business owners but useful for the developers who build for them.
The short version
Run the twelve checks, fix what fails, and verify in Search Console after launch. None of this requires a marketing degree. It requires the same care you already put into tests and code review, applied to the layer search engines see first. Ship that, and the content team finally gets a fair chance to rank.
Top comments (0)