You open your rank tracker and it says position 4. You search Google yourself and find the page halfway down page two. Your client searches from their office and sees a competitor where you should be. All three can be “right”, because each of you is looking at a different results page.
This is the most common complaint in SEO reporting, and it is rarely a bug. Google does not have one results page for a keyword. It builds a page for each searcher, and the differences add up fast. Here is where they come from, and how to get positions you can actually trust.
Seven reasons the numbers differ
1. Location: city, not just country
Google tailors results to where the searcher is, and for many keywords that means the city, not the country. “Car rental” in Riyadh and “car rental” in Jeddah are two different pages. So are “dentist” in Dubai Marina and in Deira. Many tools check from one location per country, often a data centre far from your customers, and report that as “Saudi Arabia” or “UAE”.
If your customers are local, a country-level position is an average of nobody’s experience.
2. Personalisation
When you search while signed in to Google, your history, past clicks and the sites you visit often shift what you see. If you check your own site every day, Google learns you like it, and you will see it higher than a new visitor does. That is why your own search is the least reliable rank check you can do.
3. Device
Mobile and desktop results are ranked and laid out differently. On mobile, the map pack, People also ask and videos push organic results further down, and some pages rank on one device but not the other. A tracker that only checks desktop can miss what most of your traffic sees.
4. Freshness and estimates
Some tools refresh rankings weekly or monthly for most keywords. Others fill gaps with estimates built from older data or clickstream panels. Rankings move daily, especially after updates, so a number that is a week old can be confidently wrong.
5. How positions are counted
Is the first result under an AI Overview position 1, or position 3 because two SERP features sit above it? Tools count differently. Some count only organic links, some count every block on the page. Neither is wrong, but comparing two tools that count differently will never match.
6. Which page ranks
Most tools report the best position for your domain. If your blog post ranks 6th for a keyword you want your product page to own, the report still says “6”. The keyword looks healthy while the page you care about is nowhere.
7. Reused and averaged data
Large SEO databases track millions of keywords, so they share and reuse data wherever they can: one check standing in for several locations, or positions carried over between refreshes. We call the result “SERP poisoning”. The data looks precise, but it describes a results page nobody actually saw.
How to get positions you can trust
You can’t remove the variables, but you can control them. A trustworthy rank check:
- Runs a live Google search at the time of the check, not a database lookup.
- Uses a clean, logged-out browser with no history, cookies or sign-in.
- Searches from the exact city your customers are in, using Google’s own location names.
- Checks the device your customers use, or both.
- Tracks the specific page you want to rank, not just the domain.
- Keeps a screenshot of the results page, so anyone can see what was found.
That last point matters more than it sounds. When a client says “I searched and you weren’t there”, a screenshot of the check, with the date, city and device, ends the argument in seconds.
Track the page, the place and the moment, and the numbers start to match what customers see.
How SERP Tracker does it
SERP Tracker was built to do exactly this. Every check is a live Google search in a clean, logged-out Chrome browser, in the country, language, city and device you set. It reads the top 30 results, records which of your pages ranks, flags when Google picks a different page than the one you assigned, and saves a screenshot of the page.
Checks run on our servers, or in the free desktop app, which runs the same logged-out Chrome session over your own internet connection. See all the details on the features page.
What to do next
Pick the five keywords that matter most to your business. For each one, set the target page and the city your customers are in, and run a live check. Then compare it with what your current tool reports. If the numbers differ, the screenshot will show you why.
