Most agency-comparison content is written by agencies trying to sound neutral while steering you toward hiring them. This is a genuine checklist, use it against us too, and if any answer we'd give doesn't hold up against it, that's worth knowing before you sign anything.
Start with what you're actually solving for
Agencies broadly split into a few types: pure media buyers (they run your ads, nothing else), full-funnel growth partners (media, creative, retention, analytics), and specialists (Meta-only, Google-only, or channel-specific). A brand with strong in-house creative and just needs media execution has different needs than one that needs the whole system built from scratch. Getting clear on which category you actually need before evaluating specific agencies saves a lot of wasted conversations with agencies whose core strength doesn't match your actual gap.
Questions that actually reveal fit
"What number do you get judged on internally?" If the honest answer is platform ROAS, that's a flag. Agencies optimizing for a number the platform reports itself have less incentive to surface uncomfortable truths, like when an account structure or offer problem is capping performance, not a media problem, since flagging that might mean less credit for the platform-reported wins they're measured on internally.
"Show me a time performance looked good on the dashboard but you flagged a problem anyway." This tests whether the agency will tell you things you don't want to hear. If they can't produce an example, they probably haven't, or won't, and that's a meaningful signal about how the relationship will play out once something inevitably goes wrong.
"How do you handle returns and RTO in your reporting?" For Indian D2C specifically, this is a real test. An agency that reports gross ROAS without adjusting for returns or RTO is reporting a number that doesn't reconcile with your actual P&L, and that gap compounds the longer the relationship runs without anyone catching it.
"What happens when a campaign underperforms? Walk me through the actual process." Vague answers about "optimizing" are a flag. A specific diagnostic process (checking creative fatigue, checking account structure, checking offer strength, checking measurement) signals a team that actually investigates rather than just adjusting bids and hoping.
"Can you walk me through a client relationship that ended, and why?" Every agency has had client relationships end. How they talk about it, whether with blame, defensiveness, or a genuinely reflective account of what didn't work, tells you something about how they'll handle friction with you eventually.
"Who specifically will be working on my account day to day?" Sales conversations are often led by senior people who won't touch the account after signing. Getting clarity on the actual account team, and ideally meeting them before committing, avoids a common bait-and-switch dynamic in agency relationships.
Red flags worth taking seriously
Guaranteed ROAS numbers before they've seen your account. No legitimate agency can honestly guarantee a specific performance number without first understanding your product, margins, existing account history, and market, so a guarantee offered upfront is either naive or a sales tactic.
Reporting that only shows platform-native metrics with no attempt to reconcile against Shopify or bank data. This means either the agency hasn't built that capability, or hasn't prioritized it, both of which matter for a business trying to make decisions based on actual profit rather than dashboard numbers.
Reluctance to share a client reference who's been with them over a year, since that's usually when the real relationship, not the onboarding honeymoon, shows up. New clients are often still in a positive, high-attention phase; long-tenured clients have seen the agency through actual ups and downs.
Contract terms that lock you in without a reasonable exit clause. A long minimum commitment with no meaningful off-ramp signals either confidence bordering on arrogance, or an awareness that clients tend to leave once the initial performance period ends.
What good looks like in the first 30 days
A proper audit of your existing account and business context before any recommendations, not a generic optimization plan handed over on day one before the agency has actually looked closely at your specific situation. A specific, written diagnosis of what's actually constraining growth, not just a list of tactics to try. Clear agreement on what success is measured against, ideally contribution margin or profit-adjusted metrics, not just ROAS, established explicitly before work begins rather than left ambiguous.
Evaluating the pitch versus evaluating the actual work
There's a meaningful gap between how well an agency pitches and how well they execute, and it's worth being deliberate about testing for the second, not just being impressed by the first. Ask for a small, scoped starting engagement or a paid audit before committing to a full retainer, if the agency resists this, it's worth understanding why. A confident, competent agency should be comfortable proving value on a smaller scale before asking for a larger commitment.
Pricing models and what they incentivize
Flat retainer pricing is neutral with respect to spend level, the agency's fee doesn't change based on how much you spend, which removes one potential conflict of interest. Percentage-of-spend pricing means the agency's revenue grows as your spend grows, which can subtly incentivize recommending more spend even when it isn't the most efficient use of budget. Performance-based pricing tied to specific outcomes sounds appealing but is genuinely rare in practice and often comes with structural caveats (minimum spend commitments, specific attribution definitions favorable to the agency) that are worth reading closely rather than taking at face value.
FAQ
Should I choose an agency based on case studies alone? Case studies show what's possible, not what's likely for your specific situation. Ask how similar your case-study brand actually is in category, spend level, and starting problem, a case study from a brand in a completely different situation tells you less than it might appear to.
Is a bigger agency always better than a smaller one? Not necessarily. Larger agencies often mean more account layers between you and the person actually making decisions. Smaller, senior-led teams can mean more direct access, but check their capacity to actually service your account well, a small team overextended across too many clients has its own risks.
How long should I give an agency before judging results? Most meaningful structural changes (account rebuilds, measurement fixes, creative testing cycles) take 60 to 90 days to show a clear read. Judging inside the first 30 days usually measures the wrong thing, often just how quickly they got set up rather than whether the underlying strategy is actually working.
What's a reasonable minimum commitment length for a new agency relationship? Three months is common and reasonable, enough time to see meaningful results from structural changes without locking in for an extended period before either side has real evidence the relationship is working.
Should I expect to speak to the same person throughout the engagement? A consistent primary point of contact who has real visibility into your account is reasonable to expect. Some rotation on execution-level team members is normal at larger agencies, but the strategic relationship shouldn't feel like starting over with a new person every few months.
Want a straight, outcome-focused read on your current agency setup, whether that's us or not? Talk to us at growth@adtitudemedia.com.