Your Meta ads spend is not a branding experiment. If you are putting $3,000, $10,000 or $50,000 a month into the platform, every weak decision shows up somewhere: slower cash flow, stock sitting in the warehouse, a higher customer acquisition cost, or a founder spending Sunday night staring at another vague dashboard.
That is why a proper Meta agency review cannot be based on whether the account manager is pleasant, the reporting deck looks polished, or the agency has a few nice logos on its website. The only useful question is simpler: are they creating profitable, scalable revenue – and can they explain exactly how?
Most agencies fail this test. They hide behind platform volatility, point to impressions and clicks, then ask for more time. Meta can be volatile. That does not excuse a lack of strategy, weak creative, messy campaign structure or no commercial ownership.
Start your Meta agency review with revenue, not ROAS
ROAS matters, but treating it as the whole business is how brands make expensive decisions with false confidence. A 4x ROAS can look brilliant while contribution margin collapses because discounts are too aggressive, repeat purchase behaviour is weak, or the agency is over-attributing sales that would have happened anyway.
Your agency should know the difference between an attractive Ads Manager number and profitable growth. That means asking about blended revenue, new customer revenue, customer acquisition cost, average order value, margin and the amount of spend required to generate the next dollar of revenue.
If they only report platform ROAS, you do not have a performance partner. You have someone reading numbers from a dashboard you can access yourself.
A useful review starts by comparing Meta performance with the wider Shopify picture. Has total revenue increased? Has new customer acquisition become more efficient? Are returning customers carrying the result, or is paid social genuinely doing its job? Context is not a nice extra. It is the job.
Look inside the account, not just at the results
A good result can hide a poor account for a while. Strong demand, a product launch, seasonal momentum or a loyal email list can make almost any ads account look healthier than it is. The account structure tells you whether the agency has built something that can survive the next rough patch.
Campaign architecture should have a reason
There is no universal “best” Meta account structure. Anyone claiming there is has probably copied a template and renamed it strategy. The right approach depends on spend level, catalogue size, creative volume, purchase cycle, market maturity and how much data the account produces.
But the agency should be able to explain why campaigns exist, what each campaign is designed to achieve and when it should be changed. If the account contains a graveyard of duplicate campaigns, overlapping ad sets and old tests that nobody can explain, that is not sophistication. It is neglect.
The same applies to audiences. Founders are often sold elaborate targeting structures because they look clever in a screen recording. In practice, excessive audience fragmentation can choke delivery, create overlap and prevent Meta from learning. Sometimes broad targeting is the right call. Sometimes customer lists, retention audiences or specific prospecting segments earn their place. The point is intentionality, not complexity.
Creative strategy is where most accounts either grow or stall
Meta ads are increasingly a creative and offer problem. Yet plenty of agencies still treat creative as an afterthought: a monthly request for more assets, a few resized product images, then a report saying creative fatigue is the issue.
That is backwards. The agency should have a process for identifying winning angles, building variations, testing hooks, separating message testing from format testing and feeding learnings into the next creative batch. They should know which objections are stopping customers from buying and how the ads are addressing them.
Ask to see the testing rhythm. How often are new concepts launched? What happens when an ad wins? How does the team distinguish a weak creative idea from a weak offer, bad landing page experience or insufficient spend? If the answer is “we test lots of ads”, keep pushing. Volume without a hypothesis is just spending money quickly.
Optimisation needs a commercial lens
Small daily edits do not equal optimisation. Constant tinkering can interrupt delivery and make it impossible to know what caused a change in results. On the other hand, leaving a declining account untouched because “the algorithm needs time” is equally lazy.
The right level of intervention depends on the data. At lower spend, changes need more patience because conversion volume is limited. At higher spend, waste compounds faster and poor decisions need to be caught earlier. Your agency should explain what they are watching, which thresholds trigger action and why a specific adjustment supports the larger growth plan.
Questions that expose a mediocre Meta agency
You do not need a marketing degree to assess whether an agency is in control. You need direct questions and a low tolerance for foggy answers.
Ask how they measure incremental growth rather than simply claiming credit for every tracked sale. Ask what has changed in the account over the past 30 days, why it changed and what impact they expected. Ask which creative themes are working, which customer objections they reveal and what will be tested next.
Then ask the question agencies hate: what would make you change strategy?
A capable operator will answer with specifics. They might point to rising acquisition costs, deteriorating new-customer contribution, a drop in thumb-stop performance, reduced conversion rate, stock constraints or a creative angle that has stopped producing qualified traffic. A weak agency will give you generalities about trends, optimisation and giving it more time.
Also ask who actually works on the account. Many sales calls are run by the sharpest person in the business, while delivery is handed to a junior buyer managing dozens of brands. There is nothing wrong with a team. There is something wrong with not knowing who owns outcomes, who makes strategic calls and how often senior people are involved.
Reporting should make decisions easier
If your monthly report needs a meeting to decode it, it is not reporting. It is theatre.
A useful report should quickly answer three things: what happened, why it happened and what happens next. It should connect spend to revenue outcomes, flag material changes early and name the actions being taken. It should not drown you in reach, CPM, link clicks and engagement unless those metrics explain a meaningful business outcome.
CPM can matter. So can frequency, CTR and conversion rate. They are diagnostic metrics, not trophies. A strong agency uses them to locate the constraint. For example, rising CPM may be acceptable if conversion quality improves and acquisition remains profitable. A high CTR is useless if the traffic does not buy. Numbers only mean something in relation to the commercial target.
This is where many founder-led brands get trapped. They receive plenty of activity but no clarity. Campaigns are launched, ads are refreshed and emails arrive each week, yet nobody can state whether the account is moving towards a profitable revenue target. Activity is not accountability.
Review incentives before you sign or renew
The agency model matters because incentives shape behaviour. A large generalist agency may have resources, but it may also have layers of account management, slower execution and divided attention. A low-cost freelancer may be responsive, but may lack the strategic depth or creative process needed to scale. Neither is automatically wrong. The question is whether the model fits your stage.
For an established Shopify brand, look for a partner with enough capacity to be proactive and enough accountability to care about the result. Be wary of long lock-in contracts with no clear performance milestones. Be wary of retainers that increase automatically as spend rises while service remains unchanged. And be particularly wary of agencies that promise scale without asking about margin, inventory, fulfilment capacity or your actual growth target.
A serious partner will qualify you too. They should care whether the product has demand, whether your site converts, whether inventory can support growth and whether you can supply creative at the required pace. If an agency will take anyone with a credit card, it is probably built for churn.
Underdog Marketing takes the opposite position: a selective engagement, a focus on founder-led Shopify brands already investing in Meta, and a written commitment to grow Meta ads revenue by 30% within 90 days or keep working free until the target is met. That is the kind of standard worth comparing any agency against – not because guarantees remove every commercial risk, but because accountability should be visible before the first invoice.
When changing agencies is the wrong move
Sometimes the agency is not the core problem. If the product-market fit is shaky, the offer is weak, the website conversion rate is poor or fulfilment is creating unhappy customers, a new media buyer will not fix the business. Paid social amplifies what already exists. It cannot turn an unwanted product into a winner through targeting tricks.
Likewise, do not sack an agency after one bad fortnight without checking the evidence. Creative can fatigue. Competitors can become aggressive. Seasonality changes demand. Good operators do not promise a straight line up and to the right. They make volatility understandable, respond quickly and show a credible plan.
But if your agency cannot explain results, cannot demonstrate a testing process, cannot connect spend to profit and cannot state what they own, you already have your answer. Do not keep paying for reassurance when what your business needs is execution.
The right agency review ends with a practical decision: keep the partner and demand clearer targets, fix the business constraints holding ads back, or replace the team before another quarter of spend disappears into polished reporting and excuses.