Digital Analytics and Testing
You read one footwear retailer's month of testing honestly. A product-page change showed +18.6% on day 4 and +3.6% on day 28, and you work out the sample that effect would have needed. Then you see, by simulation, what repeated looks at a running test do to a 5% false-positive rate. And you meet a conversion rate that rose 9.2% while every device segment fell.
Published by Merchandising Academy · First lesson free to read
Course value
What will you be able to do?
Work outcome
You can lay out a space or a page against a commercial objective, run a test whose result you can actually trust, and read the analytics without mistaking traffic for demand.
Who it is for
Brand and buying-office teams.
What you will produce
You read one test three ways: a 95% interval worked out from the counts, the sample the believed effect needed, and the effect the business could act on. You build a peeking table measured at 1, 4, 10, and 28 looks. You run a twelve-segment scan and find its 46.0% chance of a false winner. And you put a conversion rate on one fixed traffic mix, so it says what the site actually did.
Learning format
3 lessons · 0 templates · workplace calculations and decisions.