From no presence anywhere to $583K in year one
No website. No marketplace listings. No search footprint. Nobody had heard of the brand and nobody was looking for it, and the product sits around $400, which is not a decision anyone makes on impulse. Twelve months later the account has done $583,218, with $227,059 of that arriving without a paid click.
We spent the first month buying information, not sales
The test launch in September 2025 was a small quantity, and it was never really about revenue. It existed to answer three questions before any money went into inventory: does the listing hold up, what does the keyword picture actually look like, and will the price point stand.
So we ran auto and broad campaigns deliberately loose and treated the early ACoS as a research cost rather than waste. That is an uncomfortable thing to do when the numbers look bad on a weekly report, and it is the reason the rest of this worked.
What it bought us was the single most valuable thing a new brand can have: knowing how people actually search for the product, which was not how the brand described it.
- We rebuilt the listing around the buyer's language
Not ours. The words the brand used internally were not the words converting searches contained, and the listing was rewritten to match the second set.
- We started review velocity on a small order base
So the page was not naked when volume arrived. On a high-ticket product with no brand recognition, an empty review section is the whole objection.
- We protected the brand term early
Once people began searching the brand by name those queries converted at 10% to 29%. At this price point a single click stolen by an established competitor costs roughly $400 in revenue, so defending them was never optional.
How the first year ran
- Sep 2025Test launch
A small quantity, used to validate the listing, the keyword picture and the price point before committing to volume. Advertising ran loose on purpose to gather search data.
- Late Oct 2025Bulk inventory arrived
The rebuilt listing went live against terms the test phase had already proven, with early reviews in place and brand defense running.
- Nov 2025The real launch
First month of proper selling, into Q4 demand, with the keyword and listing work already done rather than being figured out live.
- Dec 2025 onwardScale
The seven months from late December alone account for $378,090 in sales on 1,109 units, and that window does not even include the first weeks of proper selling.
Three decisions that carried the account
Own the modifier, not the head term
Never discount
Treat the replacement consumable as a product line
Why a 2% conversion rate is the right number here
Blended conversion sits around 1.7% across roughly 51,000 sessions, with near-perfect Featured Offer control. On a sub-$40 kitchen gadget that would be alarming. On a roughly $400 appliance that involves installation and a permanent place in the room, 1% to 3% is the normal band. The buyer is weighing the install, the space and a $400 decision, and that takes more than one visit.
The proof that the account converts perfectly well is the consumable, sitting at 15.1%. Same brand, same traffic quality, much easier decision.
Conversion by listing. The consumable is an easy repeat decision; the appliance is a considered one. Both numbers are correct for what they are.
What did not go to plan
- Campaign sprawl
We reached 55 active campaigns, 17 of them with zero sales. Around $9,439 went to zero-order search terms across a 65-day window, roughly 46% of search term spend in that period. That is the honest cost of scaling discovery without pruning on a fixed schedule, and the fix is calendar discipline rather than cleverness.
- Half the backend search term field was never indexed
It was sitting at 488 characters against Amazon's 250-byte limit, so everything past the cutoff did nothing at all. An unglamorous error that had been quietly costing indexation the whole time.
- Growth plateaued rather than compounded
The economics stayed healthy at 4.21 ROAS, 23.8% ACoS and roughly 14% TACoS, but revenue sat flat at about $54K a month with 58% of it ad-attributed. Healthy is not the same as growing, and organic was not carrying its share.
- Reviews became the real bottleneck
A run of negative reviews cost more volume than any bid change we ever made. On a $400 appliance the review profile is the conversion lever, not the ad spend, and that is worth knowing before it bites rather than after.
The first twelve months
| Metric | Value |
|---|---|
| Ordered product sales | $583,218 |
| Units ordered / order items | 1,741 / 1,683 |
| Average sales per order item | $346.53 |
| Ad-attributed sales | $356,159 (61%) |
| Organic and non-attributed | $227,059 (39%) |
| Ad spend | $91,960 |
| ACoS | 25.82% |
| TACoS | 15.8% |
| ROAS | 3.87x |
| Blended conversion rate | approx. 1.7% across ~51,000 sessions |
Seller Central sales snapshot and Amazon Ads console, 28 Sep 2025 to 29 Sep 2026.
This playbook transfers to high-ticket considered purchases where conversion is naturally low, to brands with no existing search footprint to build on, to products with a replacement consumable behind them, and to any launch where the honest answer is that nobody is looking for you yet. It starts by spending a little money to find out how the market actually searches, before spending a lot of money guessing.
We do not guarantee sales or rankings, and we will tell you when the data says no. Send us your last 90 days of Business Report and advertising exports and we will return a written audit with a plan of action, risks included.