Average position is moving. Should you change anything?

A report showing average position 8.4 does not mean your page sits in one fixed place. It combines recorded appearances across the selected report. Treat a change as a prompt to look at the underlying traffic, especially when the website is starting to appear for more searches.

Begin with a smaller comparison

Choose one page, one search type and the same country and device where possible. Compare a complete period with another period of equal length. Then inspect a few important queries rather than making a decision from the whole-site average.

Google explains the metric and its aggregation rules in the search performance documentation. Keep the reporting scope beside any position figure you share with someone else.

A falling average can hide a useful gain

Illustrative example: a guide used to appear mainly for one narrow search. This month it also appears for a larger set of related questions at lower positions. Those additional appearances can make the average position look worse. The guide may still receive more clicks on its original query and more total relevant visits.

The next action is to compare the established query group separately from newly appearing queries. If the original group is steady, the new exposure may be a content opportunity rather than a reason to rewrite a working page.

The reverse can happen too. If a page loses many low-position impressions, its average can improve while clicks fall. A greener position number is not automatically a healthier page.

Look for a meaningful loss

When clicks and positions fall for the same important queries across a sustained period, open the page and check for changes. Did the offer disappear? Is the page accessible? Did a template update hide the main answer? Have relevant alternatives become more useful?

A query that asks for a comparison may deserve a comparison table. A query that asks how to fix a problem may need a tested sequence. Position alone cannot tell you which content change is needed.

Keep the action tied to the page

Write down the affected query group and the answer the page is supposed to provide. Inspect that answer before selecting an edit. If you cannot find a specific weakness, gather more evidence rather than inventing a task to make the dashboard look productive.

AbstractMetrics shows average position alongside impressions and clicks. Its opportunity screens use position ranges to find candidates for review, not to certify a stable ranking. Use the shared reports or a flexible MCP query to investigate a narrower slice when the average is ambiguous.

Report the decision clearly

A useful note says, “The established queries held their position, while new queries added impressions. We will improve the section answering the new question.” Another useful result is, “No change yet. The observed movement is too small and the traffic mix changed.” Both are more actionable than a position number without context.