Why Google rankings are fluid rather than fixed
[]Updated in 2026: Personalisation and rank tracking have changed since the original post, but its main point remains: there is no single permanent position for a page.
Google rankings move because results depend on the query, location, device, available content and changing ranking systems. Treat a position as one observation; investigate sustained changes across useful query groups.
”The ranking” is a sample
Two people can receive different local results because distance matters. Search features can vary by device and query. New pages enter the index, old pages change, and Google regularly tests and improves its ranking systems.
That does not make rank tracking useless. It makes the measurement conditional. Record the market, location, device and query set, then compare like with like.
Intent can change beneath the keyword
The same words can represent different tasks. A query may favour a current news result after an event, a product page when buying intent dominates, or an explanatory page when people need to learn. Google documents freshness systems for queries where recent information is expected.
If the result types change, your page may lose visibility without becoming worse. It may no longer match the task the result page is serving.
Investigate patterns, not blips
When visibility falls, compare:
- several weeks rather than one daily reading;
- groups of related queries rather than a trophy keyword;
- impressions, clicks and qualified outcomes;
- affected countries, devices, pages and result types;
- site releases, indexing changes and announced search updates.
Do not diagnose an algorithm experiment from one movement. First rule out measurement differences and changes you control. Then inspect whether competing results answer the task better.
A useful ranking report explains the distribution and the consequence. “Position moved from four to seven” is trivia until you know whether relevant visibility and business outcomes changed with it.
