Where they were
Halvard was doing about $40,000 a month on their own site with good margins and a mailing list that actually opened. Every agency they had spoken to wanted to list all nineteen products on day one and spend $8,000 a month finding out which ones worked.
That plan has a specific failure mode. Nineteen new listings with no reviews all compete for the same launch budget, none of them get enough velocity to rank, and four months later you have nineteen ASINs sitting at page six and a storage bill.
What we did
-
01
Launch four products, not nineteen
We looked at search volume, review depth on the incumbents, and Halvard’s own bestseller data, then picked four SKUs where the top listings had under 400 reviews. Everything else stayed off Amazon until those four had traction. That decision is most of why the launch budget was $6,200 and not $30,000.
-
02
Get the foundations right before spending anything
Brand Registry, an A+ layout that works on a phone, and image stacks built from the questions Halvard’s own support inbox already answered. We spent three weeks on this with zero ad spend running. Launching into a page that does not convert just buys you expensive data.
-
03
Reviews first, rank second
Vine on all four, plus the Request a Review automation on every order from day one. We did not touch a manual campaign until each ASIN had fifteen reviews, because bidding for a keyword you cannot convert on is a donation.
-
04
Spend hard, then stop
Weeks four to nine ran deliberately unprofitable to buy velocity and rank on eight terms. From week ten we cut bids on anything already ranking organically in the top five, because paying for a click you were going to get free is the most common leak we see in launches.
-
05
Add the rest of the catalogue slowly
Three more SKUs went up in month four, riding the review syndication from the parent listings and the brand searches the first four had generated. Nine of the nineteen are still not on Amazon and probably never should be.
Where they are now
Month five closed at just under $34,000 on Amazon with a TACoS of 14%. The first profitable month was week eleven, which is faster than we would normally promise and had a lot to do with the fact that Halvard already knew their customer.
The direct site did not shrink. It grew 9% over the same period. Some of that is Amazon shoppers searching the brand and buying direct, which is the version of cannibalisation nobody complains about.
Everyone else pitched us a big launch. Sentinel talked us down to four products and a smaller budget, which is not a great way to win a bigger retainer. It is a very good way to get a second year.
Figures come from the client’s Seller Central and advertising accounts over the period described. Past results are specific to that brand, that catalogue, and that category.
