Ranking is not the finish line. Ten results answer the same question and one of them gets the visit. This lesson is about that moment, and about measuring it honestly afterwards.
If a word here is unfamiliar it is defined in the SEO glossary. Two public penalties and one problem we found on our own network are pulled apart in the case studies, and the longer pieces live in the library.
The description does not move your position. It decides whether a person picks you out of a list where everyone is technically correct. The strongest ones name the specific thing the reader gets and admit a limit, because a line that promises everything reads like every other line on the page.
Featured snippets, people also ask, image packs, local packs and shopping units sit above or between the ordinary results and take attention from them. Being first in a list that starts halfway down the screen is worth less than it used to be, and that is a positioning problem rather than a ranking one.
A result whose title matches the exact question can outperform the one above it, because people scan for their own words rather than reading in order. Position matters, and match matters too, and only one of the two is under your direct control this week.
Position tracking measures the search engine. Clicks and impressions measure people. Sessions measure your analytics setup. They disagree, and the disagreement is normal rather than a fault. Pick the one that maps to the decision you are making and stop quoting whichever number is highest that week.
Run the free check first, then tell us what it found. We come back with what we would fix first and why, at no cost.