The honeymoon period is real, and most brands waste it.
Amazon gives a new listing a window of extra visibility to figure out whether it deserves traffic. Get it right and you compound for a year. Get it wrong and you spend the next six months buying your way out.
The sequence that works, in order
Do not launch until the listing is finished. Not "good enough for now." Every day of the window spent on a weak main image is a day you cannot get back.
Seed reviews before you push traffic. Vine on the new ASIN, 20 to 30 reviews, because conversion with zero reviews is a hole no bid solves.
Launch narrow. One primary keyword, exact match, aggressive bid. You are not trying to be everywhere. You are trying to prove you convert on one term so the algorithm trusts you on the rest.
Price for velocity first, margin second, for the first three to four weeks only. Rank is the asset you are buying.
Then widen. Once the primary term ranks, add the adjacent cluster and let the auto campaign find what you missed.
Two terms, nineteen days
A brand we launched last spring wanted to go live across 40 keywords on day one, because the product was good and they were impatient.
We held them to two terms for 19 days. Both hit page one. Then we opened it up.
Ninety days in, that SKU was doing $71K/month with a 14% TACoS. Their previous launch, done wide and fast, took nine months to reach half that.
Launches are not won by spending broadly. They are won by proving one thing quickly.
What is the single keyword your next launch has to own?