Freelance Technical SEO: What Works and What to Avoid

What freelance technical SEO is

Freelance technical SEO is independent work that makes a client's website easy for search engines to crawl, understand and index. It is the diagnostic half of SEO: instead of writing content or building links, you find the faults that stop pages ranking — broken links, redirect chains, pages Google cannot reach, duplicate content, slow pages, missing structured data — and you fix them or write exact instructions for the client's developer.

Technical SEO suits freelancers who like systems more than sentences. Clients hire you when something is wrong and nobody on their team can say what: traffic fell after a redesign, pages disappeared from Google, or a new site never ranked at all. Your deliverable is a ranked list of faults with fixes, and proof that each fix worked.

Step 1: Crawl the site and compare it with what Google has indexed

Start every engagement the same way. Crawl the site with Screaming Frog SEO Spider — the free version crawls up to 500 URLs, which covers most small business sites, and the paid licence costs $279 a year and removes the cap. Export the list of URLs, status codes, titles, canonicals and directives.

Then open Google Search Console and compare: how many pages does the site have, and how many has Google indexed? The gap between those two numbers is the first finding. Pages that exist but are not indexed, pages indexed that should not be, and pages returning errors all go on the list. Check the sitemap too — it should list the pages that matter and nothing else. A sitemap full of redirecting or error pages teaches Google to ignore it.

Step 2: Fix crawl errors, redirects and duplicates first

Rank the crawl findings by damage, and work in this order:

1. Server errors and broken links. Pages returning 5xx errors or 404s that other pages still link to. Fix the link or restore the page; do not leave internal links pointing at errors.
2. Redirect chains and loops. A redirect that points to another redirect, or back to itself. Replace chains with a single redirect to the final URL.
3. Duplicate pages. The same content at two URLs — with and without a trailing slash, with tracking parameters, or as a print version. Pick one canonical URL and point the rest at it with a canonical tag or a redirect.
4. Orphan pages. Pages with no internal links pointing at them. Google finds pages by following links, so an orphan page is a page Google may never crawl.
5. Blocked resources. Check robots.txt and meta robots tags. A staging rule copied to the live site, or a noindex tag left on a money page, is the most expensive mistake in technical SEO and the easiest to miss.

Do not start on speed or structured data until the crawl is clean. A fast page that Google cannot reach still does not rank.

Step 3: Check speed, mobile rendering and Core Web Vitals

Run the key pages — homepage, top service or product pages, top blog posts — through PageSpeed Insights, which is free. It reports Core Web Vitals: Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift, with field data from real users where available. Open the same pages on a real phone and tap through them. A page that shifts as it loads, or a button that sits under a sticky banner on mobile, is a fault a lab score alone may not show.

Fix the biggest items first: images that are far larger than the space they fill, scripts that block rendering, and fonts that load late. Speed work is never finished — the aim is to bring the money pages into the range Google calls good, not to chase a perfect score on every page.

Step 4: Check structured data and internal linking

Structured data is code that tells Google what a page is: a product with a price, a recipe, an event, a FAQ, a local business. Check the money pages with Google's Rich Results Test approach — the check schema markup guide covers the steps — and fix errors before adding new types. Structured data does not guarantee rich results, but invalid markup guarantees none.

Then check internal linking. Every important page should be reachable in three clicks from the homepage, and the pages that earn money should have the most internal links pointing at them. A blog post that links to a service page passes relevance and authority to it; a service page buried four clicks deep with no links in is a page Google treats as unimportant.

Step 5: Write the report with a fix and an owner for each fault

A technical audit report is a table, not an essay. Each row has the fault, the affected pages, the impact, the exact fix, and who does it — you, or the client's developer. Rank the rows by impact, and put the crawl and indexation faults at the top. Include the numbers: how many broken links, how many pages unindexed, how many redirect chains. The client should be able to hand the report to a developer without a meeting.

After fixes are made, re-crawl and show the result: the broken links are gone, the unindexed pages are indexed, the redirect chains are single redirects. A fix without a re-crawl is a claim, not a result. The SEO migration strategy guide covers the same discipline for site moves, where a missed redirect costs traffic overnight.

Worked example: a hypothetical 300-page ecommerce audit

An online store with 300 product and category pages hires you for a one-off technical audit. The whole site fits under Screaming Frog's free 500-URL cap, so the crawl costs nothing.

The crawl finds 18 broken internal links, 9 redirect chains, 24 product pages with duplicate titles, and 11 pages that exist but are not indexed. Search Console shows 260 pages indexed out of 300. PageSpeed Insights shows the category pages loading slowly on mobile, with oversized product images as the main cause.

Your ranked fix list starts with three items. First, fix the 11 unindexed pages — 6 are blocked by a stray noindex tag left from a sale, 5 are orphans with no internal links — because those pages cannot earn anything while Google cannot see them. Second, collapse the 9 redirect chains to single redirects and fix the 18 broken links, because every crawl wastes effort on them. Third, compress the category page images, because that one change improves every category page at once.

You deliver the report, the client's developer makes the fixes in a week, and your re-crawl shows zero broken links, zero chains and 296 pages indexed. Two months later Search Console shows clicks up from 4,100 to 4,900 a month. The audit fee paid for itself in the unindexed pages alone.

Freelance technical SEO audit checklist

• ☐ Full crawl completed, with status codes, titles, canonicals and directives exported.
• ☐ Indexed pages in Search Console compared with pages on the site.
• ☐ Sitemap checked: only live, indexable pages listed.
• ☐ Broken links, redirect chains and loops listed and fixed.
• ☐ Duplicate titles and duplicate pages identified, canonical chosen for each.
• ☐ robots.txt and meta robots checked for accidental blocks.
• ☐ Core Web Vitals checked on money pages with PageSpeed Insights.
• ☐ Mobile rendering checked on a real phone.
• ☐ Structured data validated on money pages.
• ☐ Internal linking checked: money pages within three clicks of the homepage.
• ☐ Report written with a fix and an owner per fault, ranked by impact.
• ☐ Re-crawl run after fixes, with before-and-after numbers recorded.

Common mistakes in freelance technical SEO

• Auditing without access to Search Console. A crawl shows what the site has; Search Console shows what Google sees. Without both, half the faults stay hidden.
• Optimising speed before fixing indexation. A fast page that is not indexed ranks nowhere. Crawl and indexation faults come first.
• Writing a report a developer cannot use. "Fix the redirects" is not a fix. "Change /old-page to redirect once to /new-page" is.
• Forgetting the re-crawl. Fixes that are never verified are the most common reason a client pays twice for the same audit.
• Adding structured data to broken pages. Markup on a page Google cannot index or that returns an error is wasted work. Fix the page, then mark it up.
• Chasing a perfect speed score. A money page in the range Google calls good is enough. Hours spent moving a score from 92 to 98 are hours not spent on indexation.

Freelance technical SEO tools and what each one costs

The free tools cover most audits. Screaming Frog is the first paid tool a freelance technical SEO buys, and the 500-URL free cap is the line that decides when: a site under 500 pages can be audited for free, and a larger one needs the $279-a-year licence. See is ChatGPT good for SEO before using AI drafts in a client report.

Five technical SEO terms worth knowing

• Crawl budget: the number of pages Googlebot will crawl on a site in a given period. Wasted crawls on errors and duplicates leave less budget for real pages.
• Canonical tag: a tag that tells Google which URL is the main version when the same content exists at more than one address.
• Orphan page: a page with no internal links pointing to it, so Google may never find it by crawling.
• Redirect chain: a redirect that points to another redirect instead of straight to the final URL. Chains waste crawl budget and slow users down.
• Core Web Vitals: Google's three user-experience measures — Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift.

For the speed side of this work, the Core Web Vitals optimisation guide covers the fixes in depth. Adziq's SEO services page explains how a technical engagement is scoped when a client wants the work done for them.

Common questions

Sources

• Screaming Frog SEO Spider pricing
• Google Search Central: SEO starter guide
• Google Search Console help
• PageSpeed Insights
• Google Search Central: Core Web Vitals

Adziq sells marketing services and does not accept payment for placement in its articles. See our editorial policy.