Two things worry SEO practitioners most: not knowing whether a redesign worked, and mistaking normal volatility for impact. A single-day position is affected by data refreshes, competitor moves, and even personalization. A three-place jump in one day proves little; two weeks of gains show a page is getting stronger.
Read the curve, then the distribution
The daily average position curve shows the overall direction: whether it moves lower day by day and whether it plateaus. The six-bucket distribution shows quality. An average of 15 could mean 20 keywords spread evenly from 10 to 20, or a few keywords in the top three and the rest beyond 30. The second looks good on average but depends on a single page and carries more risk.
Use landing-page attribution to locate the source of change
When the core keywords of a redesigned page move, the redesign is not necessarily responsible. Use landing-page attribution to see which URL supports those rankings. If the change is concentrated on the redesigned page, the attribution holds. If other pages on the same topic also move, the cause is more likely search volatility or competitors. Competitor insights then show whether a rival rose at the same time, separating your work from environmental change.
Watch new and lost rankings, not just rises and falls
New rankings show content starting to cover new terms, usually from new pages or better internal linking. Lost rankings are a warning that keywords were pushed out. The full evidence chain for an optimization is more new keywords, no increase in lost keywords, and core positions rising day by day. Only when all three appear together is the impact real.