- What testing your SEO really means
- The free tools I run, in order
- Step by step: how I test a site's SEO
- The story: 22 duplicate title tags and a 31% traffic bump
- Testing content and keyword targeting
- Testing whether the traffic does anything
- Why most SEO scores are nonsense
- How often you should retest
- Frequently asked questions
- Where to check the details
The short version: testing your website SEO means crawling it, checking how Google indexes it, measuring speed and mobile usability, auditing your backlinks, and then changing one thing at a time and watching what happens over weeks, not days. Most people skip the crawl and the patience, which is exactly why their "SEO test" tells them nothing. I'll show you the process I run on my own site and on client sites, with the tools, the order, and a real result.
What testing your SEO really means
Testing your SEO is not the same as running your homepage through a free online checker and getting a score out of 100. Those scores feel useful because they give you a tidy number, but I've seen sites score 92 and rank nowhere, and sites score 68 and sit on page one for their main terms. The score measures whether you tick boxes. Google doesn't rank box-ticking. It ranks pages that answer a query better than the competition and can be crawled, understood, and trusted.
A real test has three parts: can search engines find and read your pages, do the right pages rank for the right terms, and does the traffic you get do anything once it lands. Most audits stop at part one. That's the gap I want to close here.
The free tools I run, in order
- Google Search Console - free, and the closest thing to Google telling you the truth about your site.
- Screaming Frog SEO Spider - free up to 500 URLs, crawls your site like a search bot does.
- PageSpeed Insights / Core Web Vitals - Google's own speed and experience data.
- Google's Rich Results Test - checks whether your structured data validates.
- Bing Webmaster Tools - most people forget Bing exists; it still sends real traffic and gives a second data set.
- Ahrefs or SEMrush - paid, but worth it if you want to see backlinks and keyword rankings in one place. Free trials cover a first pass.
If you're already juggling five other business tools, this stack fits neatly alongside the ones I cover in my rundown of productivity tools I've tested over five years running my own business. SEO testing is just another repeatable workflow, not a mystery.
Step by step: how I test a site's SEO
- Crawl the site first. Run Screaming Frog before you touch anything else. It will show you every duplicate title tag, missing meta description, broken internal link, and orphaned page in one pass.
- Check indexing in Search Console. Go to the Pages report and look at "Why pages aren't indexed." If a chunk of your site says "Crawled - currently not indexed," that's Google looking at your pages and deciding they're not worth keeping. That's a content problem, not a technical one.
- Run Core Web Vitals. Anything under 2.5 seconds for Largest Contentful Paint is fine. I've seen sites sit at 6 to 8 seconds because of one auto-play video plugin nobody remembered installing.
- Check mobile usability. Over 60% of most business site traffic is mobile now. If your site fails mobile checks, you're losing more than half your potential audience before they even read a word.
- Search your own keywords manually, logged out, in an incognito window. Tools give you averages. Manually searching shows you exactly what a real person sees, including any featured snippets, ads, or competitor pages sitting above you.
- Check your backlink profile. A thin or spammy link profile can hold a good site back for months. I've written a full breakdown of how many backlinks a website should realistically have if you want the specific numbers by site size.
- Change one thing and log the date. Not five things. One. A title tag, a meta description, an internal link structure. Then wait.
The story: 22 duplicate title tags and a 31% traffic bump
Two years ago, mid rebuild of my own business, I published around 40 blog posts in three months. I was moving fast and not checking my own work as closely as I tell clients to check theirs. Traffic on the new posts was flat for weeks, which didn't make sense given the volume of content.
I ran a Screaming Frog crawl on the whole site out of frustration more than method. It found 22 posts sharing near-identical title tags, all variations of "Lilach Bullock | Blog," because a template default had overridden what I thought I'd typed manually in the CMS. Google was seeing 22 pages that looked interchangeable, so it was picking one and ignoring the rest, or ranking none of them well.
I rewrote all 22 titles by hand, made each one specific to the actual topic of the post, and did nothing else. Six weeks later, organic traffic to those specific pages was up 31% against the previous six weeks. Nothing else changed on the site in that window. That's the whole point of testing one variable at a time, you know what caused the shift instead of guessing.
Testing content and keyword targeting
Once the technical side is clean, test whether the right page is even trying to rank for the right term. Search Console's Performance report shows you queries a page already gets impressions for but low clicks. That's your gold list. It means Google already thinks the page is relevant, it just isn't compelling enough in the search result to earn the click.
Fix the title and meta description for that specific query, not a generic rewrite. Wait two to four weeks, then compare click-through rate for that query before and after. This is a real test with a real control, your own historical data.
If you're building something with a defined conversion path, like an online course website, test the module and enrolment pages separately from the blog. They compete on completely different keyword intent and mixing the two in one audit hides problems in both.
Testing whether the traffic does anything
Here's where most SEO testing quietly stops, and it's the most important part. You can rank first for a term, get thousands of visits a month, and still make no money if the page doesn't convert. I check average conversion benchmarks against my own numbers regularly, because a small business site converting under 2% usually has a page problem, not a traffic problem. I go into the real figures in my piece on the average conversion rate for small business websites, and it's a useful sanity check before you spend another month chasing rankings for a page that was never going to convert anyone anyway.
Test rankings and conversion together, on the same page, over the same period. A page that moved from position 8 to position 3 but converts at 0.4% needs copy and design work, not another round of keyword tweaking.
Why most SEO scores are nonsense
I'll say the uncomfortable part plainly. Most of the "test your SEO" tools people run give a single score built from generic rules that have nothing to do with your specific competitors or your specific audience. A tool can't know that your top competitor has ten years of domain history and 400 referring domains while you have eighteen months and thirty. It scores you against a template, not against the search results page you're fighting for.
The only test that means anything is comparing your page against the ten pages currently ranking above you for that exact query, on that exact day. Open them. Read them. Time how fast they load on your phone with your data connection, not office wifi. Check if they have something you don't, a comparison table, a video, a calculator, real numbers instead of vague claims. That's the test. It takes longer than running a scanner and it's far more useful.
How often you should retest
Technical crawl: every three months, or after any redesign, plugin update, or CMS migration, because those are exactly when title tags and redirects quietly break.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Rankings and click-through rate for target pages: monthly, using the same date range each time so you're comparing like for like.
Backlink profile: quarterly is enough for most small business sites, since link velocity for a site your size shouldn't be moving fast week to week anyway.
If your last full technical audit was more than six months ago, or you can't remember when it was, that's your answer. Do it this week.
If the technical debt runs deeper than a weekend project, that's exactly where a proper hands-on AI consultant for small business earns their fee, by running the crawl, the vitals check, and the fix list you've been putting off since the last redesign.
The numbers behind this are in my URL Inspection study of 1,995 pages.
For the adjacent case, read <strong>Don't Lose Your Hard-Earned Rankings: SEO Tips for Website Migration</strong>.
When you want a second pair of eyes, here's how I work as a UK SEO consultant.
Frequently asked questions
How do I test if my website SEO is working?
Compare organic traffic and rankings for your target pages over the same 30 day window before and after a single change, using Google Search Console for the numbers. If you change more than one thing at a time, you won't know which one worked.
What is the best free tool to test website SEO?
Google Search Console and Screaming Frog SEO Spider, used together, cover most of what a small business needs: indexing status, crawl errors, duplicate titles, and broken links, all without paying for anything.
How long does it take to see SEO test results?
Give any single change two to six weeks minimum before judging it. Google needs to recrawl and reprocess the page, and search behaviour naturally fluctuates week to week, so a three day check tells you nothing reliable.
Can I test my SEO without any technical skills?
Yes for the basics. Search Console's Pages and Performance reports need no technical background to read, and searching your own keywords manually in an incognito window is something anyone can do in five minutes.
