Why your Google rank looks different
The position you see is personalised, local and device-specific. Here is what changes it, what a rank tracker measures instead, and what you cannot control.
The position you see when you search for your own keyword is not the position your customer sees. Google personalises results by search history, location, device and language, and the page itself moves as SERP features push organic results down. Two people in the same city can see different orders on the same query, at the same minute.
That is not a bug and it is not something you can switch off. It means a position is only meaningful when you say whose Google you are talking about.
| What differs | How much it moves a position | Can you control it |
|---|---|---|
| Your own search history | A few places, on queries you visit often | No — only avoid measuring on it |
| Country and language | Whole different result sets | Yes: measure per market |
| City or neighbourhood | Large for local intent, nil for the rest | Partly: a tracker fixes a market, not a street |
| Desktop vs mobile | Different order, sometimes different pages | Yes: track both |
| SERP features above the results | Pushes #1 down the screen without changing it | No |
| Time of day, data centre | A place or two, transiently | No |
Your own searches are the least reliable measurement
Google uses your recent activity to re-rank results. If you have clicked your own site on a query a dozen times, you are the worst possible judge of where it sits for a stranger.
Incognito helps less than people think. It drops the signed-in history, but the IP address still says which country and roughly which city you are in, and the device still says what kind of screen is asking. It is a different personalised result, not a neutral one.
There is no “true” rank you can reach by clearing cookies. The closest honest thing is: a position measured for a stated country, language and device, with no account attached. That is what a rank tracker is for — ours runs the query without a search history and reports what it sees, which is also why the number can differ from your own browser by several places.
Location changes the result set, not just the order
For anything with local intent — a service, a shop, anything Google thinks is near-me shaped — the results are assembled for a place. The same query from two cities is not the same query.
For non-local queries the country still matters: google.com from Vietnam and
from the United States return different sets, because the country and interface
language change which pages qualify. This is why a tracker asks for a market
rather than a single “Google”.
If your customers are in one city and you are measuring from another, the number is not wrong so much as about somebody else.
Mobile and desktop are two different rankings
Google indexes mobile-first, but it still returns a different order per device, and the mobile page shows fewer organic results before the fold. A keyword can sit at #6 on desktop and #11 on mobile.
Track whichever device your traffic actually arrives on. If you do not know, Search Console’s Performance report splits by device — measure that before deciding which one to pay for.
“Position 3” can still be below the fold
The position number counts organic results. It does not count what sits above them: ads, an AI overview, People Also Ask, a local pack, images, video.
So two things follow. A rank that has not moved can still lose clicks, because a feature appeared above it. And a tracker that counts every block on the page would report a different number from one that counts only organic results — ours counts organic, because that is the number you can act on.
What you can act on is the snippet itself, which is the part of the result you control. Our SERP snippet preview shows where Google truncates a title and description in pixels.
Before any of this, the page has to be indexed
A page that is not in the index has no position at all, and no amount of rank tracking will tell you that — it will just report “not in the top 100” every day. If a keyword never shows a position, check the page is indexed first with the free index checker, and if it is not, why is my page not indexed covers the six mechanical causes in order.
How to measure it so the number means something
- Fix the market. One country and one interface language per project. Do not compare a Vietnamese result with a US one and call it movement.
- Fix the device. Track desktop and mobile separately, or pick the one your traffic uses.
- Measure without your own history. Your browser is not a measuring instrument. Use a tracker, or at minimum a different network and no account.
- Measure on a schedule, not on a hunch. Positions move daily on their own. One check tells you almost nothing; the same check every day for two weeks tells you the direction.
- Record the ranking URL, not just the number. When the page Google ranks changes, that is usually the real story — two of your pages are competing for one query.
You can see what a neutral measurement returns for your own site with the free keyword rank checker: a domain, up to three keywords, a market and a device, with no account.
What you cannot control, and should stop trying to
- Personalisation for individual searchers. Someone who visits you weekly will see you higher. That is not a rank you can report.
- Which features Google puts above the results. You can compete for them; you cannot remove them.
- Small daily movement. A place or two either way is noise. Treat a trend over weeks as signal, and a single day’s change as nothing.
The useful question is not “what is my rank”, but “what is my rank for this market, on this device, over the last month, and which page is ranking”.
How to check your Google ranking compares the ways of taking that measurement, and why did my Google rankings drop is the one to read when a position you were happy with has gone.