Four: run the work by hand inside Seller Central, hire a freelancer per function, hire a team that brings its own tooling, or build a small reporting sheet of your own. Automation executes decisions you have already made. If your targets are undefined, it repeats the guess faster.
The short version
- Decide the policy before you buy the engine. A rules engine with no policy is a random number generator with a dashboard.
- Score the options on decision quality, not feature count. Every tool in this category shows you a similar chart.
- Ask any candidate, human or software, for two advertising targets: one for a product at launch, one for the same product at maturity.
- Automation is strongest where volume is high and stakes are low. Negative keywords, harvesting, budget caps. It is weakest at judgment.
- Your own hours are a real line item. Put a price on them before you compare anything.
Score the four alternatives yourself
Do not start from a feature comparison. Start from a weighted scorecard, because the differences between these options sit in things no feature list mentions. Score each option from 0 to 5 on each row, multiply by the weight, and add it up.
| Criterion | Weight | What a 5 looks like |
|---|---|---|
| Decision quality by product stage | 30 | Targets differ by product, and somebody can explain why each one is set where it is |
| Coverage outside advertising | 25 | Listings, creative, inventory, and account health all have an owner |
| Speed from signal to action | 15 | A bad search term is negated this week, not next month |
| True cost including your hours | 20 | The invoice plus your time is lower than the alternatives |
| Willingness to stop | 10 | Something in the arrangement can conclude that a product should be cut |
The weights are mine and you should argue with them. What matters is that you weight decision quality above features, because that is the axis on which these four options actually differ.
Running it by hand
Seller Central and the ads console do the job for one or two products. You get complete visibility and no subscription. The cost is your calendar and your consistency, and consistency is where hand-run accounts fail. Scores well on cost, badly on speed once you pass three or four products.
Freelancers per function
You buy hands and, if you choose well, judgment in one area. It scores well on decision quality inside that function and poorly on coverage, because nobody owns the problems in between. You also inherit the project management.
A team that brings its own tooling
You pay a fee and the tooling is inside it. Scores well on coverage and speed, and the honest weakness is that you are now dependent on somebody else's attention. That is why the staffing questions matter so much before you sign.
A reporting sheet you build yourself
Underrated. A weekly export into a sheet that shows search term spend, conversion rate, and rating movement gives you most of the visibility of a paid suite. It scores badly on speed and well on cost, and it teaches you what the numbers mean, which is worth more than the sheet.
The question that separates policy from software
Ask for the advertising target on a product in its launch window, and the target on the same product once it is established. Those two numbers should be far apart, and anyone who says so without prompting understands the mechanism.
At launch you are buying rank and review velocity, so an aggressive ACoS is the correct answer and an efficient one means you are quietly conceding position to a competitor who is willing to pay. At maturity the product defends its place and efficiency is the whole job. A single target applied across a catalog will overspend on the mature half and underspend on the new half at the same time, which is a hard trick to pull off and an expensive one.
Software will happily hold whichever target you type in. That is exactly the problem. The tool is not wrong, it is obedient.
What most agencies will not tell you
We use the same category of software you can buy. So does almost everyone in this industry. The tooling is not the differentiator, and any pitch built around a proprietary dashboard is a pitch about the wrong layer. What you are actually paying a team for is who reads the output, how often, and what they are allowed to change without asking.
The second thing rarely said: for a single product with modest spend, a subscription plus two focused hours a week from you will beat a retainer. A monthly fee that swallows the margin on one product helps nobody, and we say so on audits and lose the deal.
The third: automation applied to an account with a conversion problem will optimize the account into a smaller version of itself. It cuts spend on terms that convert badly, which looks like efficiency and is actually retreat. Fix the page, then automate.
Related answers
- All in one Amazon software vs specialist freelancers
- Alternatives to hiring a full-time Amazon manager
- Dayparting and PPC automation for Amazon ads
- Global Amazon PPC tools compared
- Build vs buy for your Amazon channel: the complete guide
If you want the two target numbers written against your own catalog before you renew a subscription, ask for the free audit at Flapen.
