Why Most "Questions to Ask" Lists Don't Actually Help You Choose
Search for how to pick an Amazon agency and you'll find the same article roughly a dozen times over, wearing different headlines. Seven questions. Twelve questions. Fifty, if the writer got ambitious. They all read fine. Most of them ask about the right things: contract length, reporting, account manager attention, past results. The problem isn't the questions. It's that a list of questions with nowhere to put the answers isn't actually a decision tool, it's just a longer conversation.
Here's what happens in practice. You ask an agency how many accounts their manager handles. They give you a number, or they don't. You ask about reporting. They mention ACoS, maybe TACoS if you're lucky. You ask about contract length and get "flexible, we can talk about it." Now you have a page of notes and no way to compare that page against the notes from the agency you talked to last week, because nothing was scored, weighted, or ranked against anything else. Every agency starts to sound plausible, because a flat checklist treats a strong answer and a mediocre one as interchangeable check marks.
The problem with a flat checklist
A checklist format assumes every agency can technically answer every question, and most can. The gap isn't in whether they respond, it's in the quality and specificity of what they say. "We report on ACoS" and "we report on ACoS and TACoS, broken out by campaign type, updated weekly" are both answers to the same question, but they represent two very different operations. A checklist that just tracks whether the topic got mentioned can't tell you the difference. That's the actual failure mode buyers run into: not a lack of information, but no framework for weighing the information they already collected.
What a scoring framework gives you that a checklist doesn't
A scoring framework forces you to rate the quality of each answer, not just its presence, and then weigh categories against each other based on what matters most for your business. It turns a stack of loose notes into something you can actually compare, side by side, across two or three finalists, using the same criteria each time. That's the entire point of this article: not a longer list of questions, but a way to score the answers you get to the questions you're probably already asking.
Step 1: Confirm You Actually Need an Agency Right Now
Before you score a single proposal, it's worth pausing on a more basic question: do you need an agency at all, right now, at your current size? Plenty of sellers start shopping for agencies out of frustration rather than a clear read on their actual needs, and that frustration can push you toward signing with whoever sounds most confident on the first call, rather than whoever is the right fit.
The revenue and time-availability signal
As a general pattern, sellers doing meaningful monthly revenue with genuinely no time left to manage PPC, listings, and account health are the ones who benefit most from an agency relationship, fast. Sellers earlier in that curve, with more available time relative to revenue, often get more value out of self-management or a freelancer for a while longer. We go deeper into exactly where that threshold tends to sit, and how to reason about your own numbers, in our full breakdown of whether an Amazon agency is worth it at your stage. If you haven't run that math yet, it's worth doing before you take a single sales call, because it changes what you're even looking for.
When self-management or a freelancer is still the better move
If you've got the bandwidth, a smaller catalog, and a business that hasn't yet outgrown what one person managing PPC part time can competently handle, an agency retainer might be solving a problem you don't have yet. That's not a knock on agencies, it's just fit. A freelancer covering a narrower scope, usually PPC bid management, is frequently the right in-between step, and we cover exactly how that comparison plays out in our agency versus freelancer breakdown. Larger brands weighing whether to build an internal team instead of paying an agency retainer at all should read our agency versus in-house comparison before going further down this framework, since the math changes meaningfully at scale.
The 5-Category Evaluation Framework
Once you've confirmed an agency is the right move, the actual evaluation comes down to five categories. Every agency you talk to should get scored, informally or on paper, across all five. None of them matters in isolation, and an agency that's excellent in one category and weak in another is a very different proposition than one that's solidly average across the board.
Category 1: Account Structure and Attention
This is the account manager ratio question: how many other accounts does the person actually running your day-to-day work also handle? It's the single best proxy for how much real attention your account gets after the contract is signed and the honeymoon period of the sales process ends. An account manager juggling a large book of accounts has less time to catch a listing suppression before it costs you a week of sales, or to notice a competitor's bid strategy shift before your ACoS creeps up. This isn't about finding an agency that only handles one client. It's about getting a straight, specific answer instead of a dodge.
Category 2: Reporting Philosophy
Ask what an agency reports on, and you'll usually hear ACoS first. That's not automatically a problem. ACoS is a real, useful metric. But an agency that reports on ACoS and never brings up TACoS is optimizing for and measuring a narrower slice of your business than one that tracks both. ACoS can look great on a single hero campaign while your overall advertising cost against total revenue quietly rises, because ACoS ignores organic sales entirely. An agency fluent in both metrics, and able to explain when each one matters more, is telling you something real about how they think about your account's profit, not just its ad performance.
Category 3: Contract Terms and Lock-In
Contract length by itself isn't automatically a red flag. A 12-month agreement with clear quarterly checkpoints and an easy exit clause is a very different commitment than a 12-month agreement that renews automatically unless you cancel in writing 90 days out, buried in section 14. What you're actually evaluating here is the combination: length, renewal terms, and how easy or hard it is to leave if things don't work out. We break contract red flags down in detail further in this article, and go even deeper on the fine print worth checking in our full guide to Amazon agency contract terms.
Category 4: Transparency and Proof of Work
Will the agency show you an anonymized sample report before you sign? Will they walk you through a real change log, the actual list of what they changed on a client's account and when? Will they hand you two or three reference clients you can call directly, not just a polished case study on their site? These are all forms of the same underlying question: can this agency show its work, or does it only talk about its work in the abstract. Reference clients specifically are worth dwelling on, because research into how buyers actually shop for agencies shows most never ask for them at all, which makes it one of the more useful, underused filters available to you. Our own case studies page is one example of what visible proof of results can look like; use it as a reference point, not a sales pitch, when judging what other agencies show you.
Category 5: Scope Fit
A PPC-only agency and a full-service agency should not be evaluated with the exact same weighting. If you need bid management and nothing else, an agency's SEO capability is irrelevant to you, and you shouldn't penalize a specialist for not offering it. If you need coordinated PPC, listing optimization, creative, and account health monitoring under one roof, a narrow PPC specialist is the wrong fit regardless of how good their bid management is. Look at a dedicated PPC agency's scope against an SEO-focused offering and a full-service model to get a feel for how differently these are structured before you decide which one you're even shopping for.
The Scorecard: How to Score an Agency During (or Right After) a Sales Call
Here's where the framework becomes something you can actually use live, not just read. For each of the five categories, rate the answer you got as green, yellow, or red, right after the call while it's fresh. Don't average them into one vague overall impression; keep the five scores visible side by side so a lopsided profile, strong in two categories and weak in three, doesn't get smoothed over into "seemed fine."
| Category | Strong answer looks like | Weak answer looks like | Weight if budget-tight | Weight if growth-focused |
|---|---|---|---|---|
| Attention (manager ratio) | A specific number, offered without hesitation | "It depends on the account" | Medium | High |
| Reporting philosophy | Mentions ACoS and TACoS unprompted | Only ever says ACoS | High | High |
| Contract terms | Short initial term, plain-language renewal | Long lock-in, vague on renewal until asked twice | High | Medium |
| Transparency / proof | Offers a sample report and references unprompted | "We can't share client data" as a blanket answer | Medium | High |
| Scope fit | Clearly matches what you actually need, no more, no less | Pushes a bigger package than your account needs | High | Medium |
Green flag, yellow flag, red flag answers for each category
A green flag is a specific, confident, checkable answer. A yellow flag is vague but not evasive, worth a follow-up question before you decide it's a real problem. A red flag is a dodge, a non-answer, or an answer that reveals a structural issue (like an account manager who genuinely doesn't know how many accounts they run, which suggests nobody at the agency is tracking that number at all). One yellow flag on an otherwise strong scorecard usually isn't a dealbreaker. Two or more red flags across different categories usually is, regardless of how good the pitch deck looked.
Weighting the categories based on what you actually need most
If you're budget-constrained, weight contract terms and scope fit more heavily, since a bad exit clause or an oversized package does the most damage to a tight budget. If you're focused on aggressive growth, weight attention and transparency more heavily, since a distracted account manager and an agency that won't show its work are the two things most likely to slow down a growth push you're paying a premium for.
Questions to Ask, and What a Good vs Bad Answer Sounds Like
The scorecard above tells you what to weigh. This section gives you the actual phrasing to listen for, category by category, when you're on the call.
| Question | Green flag | Yellow flag | Red flag |
|---|---|---|---|
| How many accounts does my manager personally handle? | A specific number, with context on account complexity | "A handful, it varies" | "It depends on the account" with no follow-up detail |
| Can I see an anonymized sample report? | Sends one within a day, unprompted mention of TACoS | Sends a generic template, only after asking twice | "We build custom reports after onboarding," no example offered |
| What's the contract length and how do I exit? | Explains term and notice period clearly, unprompted | Gives the term but is vague on notice period until pushed | "Standard terms, it's all in the contract" with no walkthrough |
| Can I speak with two or three current clients? | Offers names within a few days | Offers one, or a written testimonial instead | "Our clients prefer confidentiality," offered as a blanket rule |
On account manager workload
Listen for specificity. An agency that genuinely staffs accounts thoughtfully knows this number and isn't shy about it, because it's a point of pride, not a liability. Vagueness here usually means either nobody's tracking it, or the honest number is one they'd rather not say out loud.
On reporting and metrics
Beyond ACoS versus TACoS, ask how often reports update and whether you'll see a change log alongside the numbers, a record of exactly what was adjusted and when. A report with numbers but no narrative of what actually changed leaves you unable to connect cause and effect.
On contract terms
Get the renewal mechanism in plain language before you sign anything. "Auto-renews unless you cancel with 60 days' written notice" is a fact you want spoken out loud on the call, not discovered nine months in.
On proof of past work
A confident agency treats this as a normal, expected request. One that treats it as an imposition is telling you it's not something they're used to being asked, which is itself informative given how rarely buyers apparently ask for it.
Red Flags That Should End the Conversation
Some issues are worth a follow-up question. Others are worth walking away over. The difference usually comes down to whether it's a single weak answer or a pattern across categories.
A 12-month contract alone isn't disqualifying. Neither is ACoS-only reporting alone, or an account manager who won't give a specific ratio. Two or three of these together, on the same call, is a genuinely different signal than any one in isolation. Treat the pattern as the red flag, not the individual data point.
Contract red flags
A 12-month or longer lock-in combined with automatic renewal and no clear, written notice period is the single most common complaint pattern that shows up once sellers try to leave an underperforming agency. Ask directly what happens if you want out at month four.
Reporting red flags
ACoS-only framing, with no willingness to also discuss TACoS or a full profitability view, and no visible change log tying specific account actions to specific outcomes. If an agency can't show you what actually changed on your account and when, you're paying for activity you can't verify.
Attention red flags
A manager who won't say how many accounts they run, or an agency that dodges the question by talking about "team-based support" instead of a direct answer, usually means the honest number is higher than you'd be comfortable with.
Pricing Models You'll Encounter and What Each One Incentivizes
Pricing isn't part of the five-category framework directly, but it shapes incentives enough that it deserves a quick mention here. You'll typically run into three structures.
| Model | What it incentivizes | One-line takeaway |
|---|---|---|
| Percentage of ad spend | Growing your ad budget, not necessarily your profit | Commonly cited in the 10-30% range; watch for budget-increase requests without profit justification |
| Flat monthly retainer | Predictable cost, no built-in reason to inflate spend | Cost stays flat as your ad spend scales, unlike percentage models |
| Hybrid | A blend, only as fair as its actual split | Evaluate the specific base-fee-to-percentage ratio, not the label |
This is a deliberately short summary. Pricing structure deserves its own full treatment, including how to model which model actually costs less at your specific spend level, which we cover in our complete Amazon agency pricing guide.
Comparing Your Shortlist Side by Side
Once you've run two or three finalists through the same scorecard, lay the results next to each other. This is where the framework earns its keep, because it surfaces tradeoffs a gut-feel comparison hides.
Say Agency A quotes a lower monthly fee but gives you a one-paragraph scope of work with no month-one deliverables listed. Agency B costs more but hands you a detailed breakdown of exactly what happens in week one, week two, and week four. On price alone, A wins. On the scorecard, A scores yellow or red on transparency and possibly scope fit, since a vague scope of work often means vague expectations later, while B scores green on the same categories. The price difference might be entirely justified once you see it isn't really a price difference, it's a difference in how much specificity you're paying for.
Or consider a brand doing around $120,000 a month in revenue that receives a proposal quoting 15% of ad spend on a 12-month contract, where the sample report shows ACoS and nothing else. Run that through the framework: contract terms score red (long term, no mention of exit terms until asked), reporting scores red (no TACoS, no profitability lens), and two red flags out of five categories is enough to seriously reconsider before signing, regardless of how the sales call felt in the moment.
What Good Onboarding Looks Like Once You've Chosen
Choosing well doesn't end the evaluation, it just moves it into a new phase. A strong onboarding process usually includes an account audit in the first one to two weeks, a documented month-one plan with specific deliverables (not just "we'll optimize your campaigns"), and a first reporting cycle that arrives on schedule and actually reflects the reporting philosophy they described on the sales call. If the agency that promised weekly TACoS-inclusive reporting shows up with a monthly ACoS-only summary in month one, that's worth raising immediately, not waiting out. Running an account audit yourself before onboarding even starts, through a service like a dedicated account audit, also gives you a baseline to measure the new agency's early impact against, rather than taking their word for what's improving.
Frequently Asked Questions
How do I know if an Amazon agency is legitimate?
Ask for their scope of work in writing, a sample of an anonymized client report, and two or three reference clients you can actually contact. A legitimate agency answers all three without hesitation. One that gets vague about any of them, especially references, is telling you something worth listening to.
What's a normal contract length for an Amazon agency?
Three to six months is commonly cited as a reasonable initial term, long enough for PPC and listing changes to show real results, short enough that you're not trapped if the fit is wrong. Month-to-month after an initial term is a strong sign of confidence. Anything at 12 months or longer with automatic renewal deserves extra scrutiny.
Should I ask an Amazon agency for references before signing?
Yes, and most buyers skip this step, which is exactly why it's a useful filter. A real agency can produce two or three clients willing to talk about their actual experience, not just a case study slide. If they can't, ask why.
What's the biggest red flag when hiring an Amazon PPC agency?
The combination of a long contract, automatic renewal without clear notice terms, and reporting that only ever mentions ACoS. Any one of those alone might be explainable. Together, they usually mean you're locked into a relationship with no visible way to check whether it's actually working.
Can I switch agencies later if this one doesn't work out?
Usually yes, though the contract terms determine how painful it is. This is exactly why contract length and notice periods belong in your evaluation before you sign, not after you've decided you want out. Ask directly what the offboarding process looks like.
Do I need different criteria for a PPC-only agency versus a full-service one?
Some criteria stay constant (account manager attention, contract terms, transparency), but scope fit changes which questions matter most. A PPC-only agency should be judged almost entirely on bid strategy and reporting depth. A full-service agency also needs to prove it can coordinate SEO, creative, and account health without any one function getting neglected.
