Rankings bounce after a change because Google re-crawls the page, re-evaluates it against the query, and effectively tests it at different positions while it gathers fresh signals, and the position only settles once that re-assessment has run its course. The bounce is not a verdict, it is the re-evaluation happening in real time, and reading it as a final result is the most common mistake people make in the days after an edit.
When you change a page, you reset part of what Google knows about it. The crawler has to revisit, the content has to be reprocessed, and the page’s relevance and quality have to be weighed again against the competing results. During that window, Google often shows the page at varying positions to see how it performs and how the new version compares, rather than locking it into one spot immediately. That experimentation is what you observe as a page jumping from position four to nine to six within a short stretch.
The instability is temporary because it reflects an incomplete picture. Early on, Google is working with partial signals about how searchers respond to the updated page and how it stacks up against neighbors, so its placement keeps adjusting as more data arrives. As those signals accumulate and stabilize, the testing narrows and the page converges on a position that reflects the new version’s actual standing. The same pattern shows up after on-page edits, content rewrites, and link changes, and it is distinct from the broader early volatility a brand-new page goes through before it has any established baseline at all.
The practical takeaway is to wait for the settle before you judge the change. Resist the urge to revert an edit or pile on more changes the moment the page dips, because reacting mid-bounce restarts the whole re-evaluation and prevents the page from ever stabilizing. Give it a couple of weeks of relative calm, then compare the settled position to where the page started. That settled number, not the noisy days in between, is the real measure of whether the change worked.