Every Amazon advertising tool has a features page, and most of them list the same forty things. Bid optimization. Keyword harvesting. Reporting. Dayparting. AI. The lists do not help, because the words mean different things in each product and the features that matter are often the ones described least.
This is the short list: the eight features that decide whether a tool changes results, three that sound important and rarely are, and the question to ask about each. It is written to be used while reading a vendor's page, including ours.
The eight that matter
1. A change log you can read
Every change the tool makes, with the date, the campaign or keyword, the before and after, and the reason. Not a summary. The actual list.
This is the feature that makes every other feature safe. Without it, you are trusting the software on faith, and when ACoS moves you cannot tell whether the tool did it or the market did. Why every account needs a change log covers the case. The question to ask: show me the log from a real account, with the client's details removed.
2. Hour-level scheduling
The ability to change campaign state, budget or bids by hour of day and day of week, not just on and off daily. Dayparting is the largest controllable waste in most accounts, and the console cannot do it. Many tools claim it and offer a daily toggle, which is not the same thing. The question: can I pause from 1 a.m. to 6 a.m. on weekdays only, and raise bids from 7 p.m. to 10 p.m. on weekends? If the answer involves a workaround, it is not the feature.
3. Restore after change
Any change made for an event or a test should carry its own undo: a date or a condition on which the setting returns to what it was. Tools that raise budgets for Prime Day and leave them raised have automated half the job. The post-event budget problem is what this feature prevents. The question: when I set an event rule, where do I set the restore date?
4. Guardrails on every rule
A floor and a ceiling on bids. A cap on daily spend. A minimum data threshold before a rule acts (no bid changes on five clicks). A limit on how far a single rule can move a setting in one pass. Rules without guardrails are how a tool turns a bad day into a bad month. The question: what stops this rule from lowering a bid to zero or raising a budget without limit?
5. Runs on your structure
The tool works with the campaigns you have, named the way you named them. It does not require rebuilding the account its way to function. Some tools do, and for a new account that can be fine. For an account that works, a forced restructure is weeks of disruption and a lock-in: leaving means rebuilding again. The question: can I connect my account and run a rule today without changing anything?
6. Multi-account, same rules
If you run more than one account, or will, the tool should let one rule apply across accounts and show results account by account. Agencies live on this. Brands with two marketplaces need it. The question: can I write a dayparting rule once and apply it to three accounts?
7. Alerts that fire between runs
Spend spikes, campaigns out of budget by noon, ACoS drifting past a threshold. These should reach you the day they happen, not in the weekly report. Alerts are the feature that catches what rules were not written for. The five alerts worth setting covers which ones. The question: what will this tool tell me about, without my asking, and how fast?
8. Transparent pricing that does not scale with spend
Not a feature in the product, but a feature of the relationship. Percentage-of-spend pricing means the tool costs more as you grow and is paid more when you spend more, regardless of whether you spend well. Flat or tiered pricing does not have that problem. The question: what does this cost at twice my current spend?
Three that sound important and rarely are
AI
The label covers everything from a bidding model trained across many accounts to a chat interface over a rule builder. Neither is bad; neither tells you what the tool does. A model that works on high-volume accounts can be erratic on low-volume ones. Ask what the AI decides, what data it uses, and whether you can see its reasoning. If the answers are vague, the feature is a label. Rules versus AI covers the trade-off.
Keyword research
Useful, and available in many places, including the console's own search term report and dedicated research tools. A PPC management tool that includes keyword research is bundling, not differentiating. Judge the tool on management; get research wherever it is best.
Dashboards
Every tool has them, they all look similar, and they rarely change a decision. The console's reporting is adequate for most sellers. A beautiful dashboard on top of a tool that cannot schedule by hour is a beautiful dashboard. Judge the tool on what it changes, not on how it displays what it did not.
Using this list
Take any vendor's features page, including the one for Off Hours, and check the eight. A tool that has all eight and none of the three is unusual and probably good. A tool that leads with the three and buries the eight is telling you where its effort went. The software comparison sorts the market into categories, and the evaluation checklist turns this list into a trial plan.
Frequently asked questions
What is the most important feature in Amazon PPC software?
A change log you can read. Every other feature changes the account; the log is how you know what it did and why. A tool that cannot show you its changes with dates and reasons is a tool you are trusting on faith, and that is the wrong relationship to have with something that spends your money.
Do I need AI features in Amazon advertising software?
Not as a requirement. AI bid models can work well on high-volume accounts and poorly on low-volume ones, and the label tells you nothing about which. Judge the tool on whether you can see and predict what it does. Rules you can read are more useful to most sellers than a model you cannot.
Should Amazon advertising software restructure my campaigns?
Only if you want it to. Some tools require their own campaign structure to function, which is fine for a new account and a real cost for one that already works. A tool that runs on your existing structure, and changes it only when you ask, is easier to adopt and easier to leave.
Off Hours was built around the first seven: readable rules, hour-level scheduling, restore dates, guardrails, your structure, multi-account, and alerts, with a change log on all of it. The eighth is a flat $149 per account. Start a free 14-day trial.