A short, repeatable routine for reading your numbers each week and ending with one decision.
Looking at analytics when you are worried and ignoring them when you are busy leads to unsteady decisions. A short weekly routine fixes that. Allow thirty to forty-five minutes at a set time.
Before you start
Choose a question for the week, such as "Is the new opening helping?" Open Analytics and set two equal date ranges: the last full week, and the week before. Menu names can change, so use the official analytics help for the current layout.
The routine
Overview: note the direction of views, watch time and any other headline measure. Write one sentence, not a verdict.
Top and bottom: list your best and weakest videos of the week. Compare each with similar videos of your own.
Reach: for those videos, check impressions and CTR. Did visibility change, or packaging?
Sources: look at the top traffic source and any source that grew or shrank. See Traffic sources explained.
Watch behaviour: open the retention graph for the best and weakest video. Mark the biggest dip and its timestamp.
Comments: read a sample and tag questions, ideas and complaints.
Decide: write one hypothesis and one action. Maybe an opening to rewrite, a title to adjust, or a question to turn into a video.
Log: record the date, video links, hypothesis and action in a simple document.
Rules that keep it honest
Compare like with like: equal periods, similar formats, similar ages.
Do not react to a few views.
Change one thing at a time.
Review good weeks too.
Never predict earnings or growth from one week.
Monthly step
Once a month, read your log from top to bottom. Which hypotheses were supported? Which actions should become habits, and which should be dropped?