How it works
What happens, in order, from the first email
No dashboard to learn, no onboarding call you have to sit through. The work is a weekly cycle that either sends you a list or makes the changes itself, depending on which tier you are on.
The week, in six steps
The same loop runs every week. On Advisory it stops at step five and sends you the list instead of acting on it.
- 01PullYour account comes in
- 02ScoreEvery product against its own bar
- 03ProposeThe changes worth making
- 04CheckEach one has to clear the checks
- 05ActApplied, or sent to you as a list
- 06ReportWhat changed, and what it did
Then again the following week, with the last report's numbers as the starting point.
Getting started, in order
Two short steps at your end, then the loop above takes over.
-
The free audit
You send me two reports, or grant viewer access and I pull them myself. Back comes a written read of the account: where the spend is going, the terms costing you most with nothing to show, your campaigns ranked by what they actually return, and the three changes I would make in week one. About a business day, no charge, no obligation.
-
Intake, if you want to go further
Three numbers per product: landed cost per unit, selling price, fulfilment cost. That is what gives each product its own bar instead of asking you to guess at one number for the whole account. Twenty minutes for most people.
Then access. On Advisory that is viewer access, which is read only. On Execute, Growth and Continuous it is editor access, which is what lets me make the changes. You grant it through Amazon's own user permissions and you can revoke it yourself at any time without telling me first.
-
The first cycle or two are read only
Your account gets mirrored and watched before anything moves, so I can see how it behaves first. Nothing is written in this phase.
What that actually looks like
This is the deliverable, not a dashboard you have to go and learn. Every change gets a row: what it was, what it became, and one sentence on why.
| Campaign and target | Change | Before | After | Why |
|---|---|---|---|---|
| SP | Massage Ball | Exact | Corekw: massage ball for back | Reduce bid | $1.42 | $1.14 | Returning well under what this product needs to make money. The cut is deliberately partial, so it moves again next week rather than all at once. |
| SP | Posture Corrector | Auto | Discoveryterm: posture corrector for dogs | Add negative | running | negative exact | 37 clicks and $41.20 spent over 60 days with no orders. Checked against the rest of the account first and it is not earning anywhere. |
| SP | Stretch Strap | Exact | Scalekw: physical therapy stretch strap | Raise bid | $0.88 | $1.06 | Clearing its product's bar on a real run of orders. The raise is sized to how much evidence there is behind it, not to how good the ratio looks. |
| SP | Massage Ball | Auto | Discoveryterm: lacrosse ball for trigger points | Promote to exact | auto only | exact at $0.94 | Earned its way out of discovery on its own results, so it gets its own keyword and its own bid rather than sharing one. |
| SP | Massage Ball | Exact | Coreterm: deep tissue massage ball set | Held, kept | $1.27 | no change | Flagged for negation on this week's spend alone. Its longer history shows it earning above this product's bar, so the negation was blocked. |
| SP | Posture Corrector | Broad | Validationkw: back brace posture support | Held, no raise | $0.71 | no change | The ratio called for a raise. Withheld, because at this conversion rate a higher bid buys more clicks that do not convert, which costs more than it returns. |
The two highlighted rows are the ones worth reading. Both are changes the engine worked out and then refused to make, and both are recorded with the reason, because a week where holding still was right should be visible rather than quietly skipped. Layout and reason strings are exactly what the weekly log produces. The campaign names and figures here are illustrative, from my own product catalogue.
Your goals, then the guardrails
Before anything runs we agree what your account is for: what each product has to return, how fast you want to move, and what you would rather I left alone. Every decision is measured against those goals, and these six checks hold the line every cycle.
- A profit floor. No bid goes above what that product can earn.
- A spend circuit breaker. The cycle stops writing if spend leaves its expected range.
- Protection on negatives. A term that is genuinely earning cannot be cut.
- A freshness check. Stale data means nothing gets written that week.
- Cooldowns and step caps. Nothing moves far, or twice, in the same week.
- A daily cap and a kill switch. A hard ceiling, and one switch that stops everything.
The thresholds behind them stay private, and I will walk you through any decision made on your account, any time you ask.
One person, not a rotating account team
The single most common complaint in public reviews of the larger tools and agencies is that the person who sold you the account is gone within a year. I am an Amazon seller who built this system for my own account, and I personally operate every client account on it. The client list stays small because I am the one doing the work. You will always know who is in your account.