The Mistakes Performance Marketers Make When Hiring an Agency or Building an In-House Team

Key Takeaways

• The mistakes are predictable and self-inflicted: pressure to show results fast makes even experienced marketers abandon the rigor they apply to campaigns when they choose an agency or hire in-house.
• Choosing on unverifiable ROAS claims and polished case studies is the first trap — insist on results you can reconcile against your own booked revenue, not a screenshot from someone else's account.
• Handing over measurement and account ownership is the most expensive mistake: it means you can neither verify what is happening nor leave without losing your pixel, data, and history.
• Match the structure to your stage — a scale agency cannot fix an unvalidated funnel, and an in-house hire you cannot keep fully utilized is an expensive idle asset.
• Watch the incentive: a percentage-of-ad-spend fee rewards the agency for spending more of your money, not for making you more of it — the one behavior that costs you most.
• Judge on the right metric over the right horizon — incremental, reconciled results, not platform-reported ROAS read too early on a funnel that has not matured.

You Know Better — So Why Does This Keep Happening?

Here is the uncomfortable thing about this topic: performance marketers, of all people, should not make these mistakes. You spend your working life refusing to be fooled by a vanity metric, reconciling numbers against reality, and demanding proof before you believe a result. And yet when it comes to the decision of hiring a performance marketing agency or building an in-house team, the same predictable errors show up again and again — the very people trained to see through a good story get sold by one. Why? Because the decision is made under a different kind of pressure. You need results, you need them soon, someone above you is asking when the numbers will move, and that pressure overrides the discipline you apply to everything else. The rigor goes out the window at exactly the moment it matters most.

Ask yourself honestly: when you last evaluated an agency, did you apply the same skepticism to their ROAS case study that you would apply to a campaign report from a channel you distrust? Or did the polish, the confidence, and the urgency carry you past the questions you would normally ask? Most marketers, pressed, admit they did not verify the claims the way they verify their own numbers. That gap — between the rigor you apply to your work and the rigor you apply to choosing who does your work — is where every mistake in this guide lives. The good news is that the mistakes are predictable, which means they are avoidable by simply applying the discipline you already have to the decision itself.

This guide walks through the specific mistakes that cost you, each with a scenario you will recognize and the questions to ask before you make it. It covers both paths — hiring an agency and building in-house — because many of the mistakes are really about the same underlying failure: not applying your own standards to the decision. The point is not that agencies are bad or that in-house is better, or the reverse; it is that whichever you choose, you should choose it with your eyes open, verifying what you are told, owning what matters, and matching the choice to your actual situation. Do that and this decision becomes just another problem you are well-equipped to solve, rather than the blind spot it usually is.

The Six Mistakes, at a Glance

The mistakes cluster into six that account for most of the damage, and they are worth seeing together before going deep, because they compound — a wrong choice on one makes the others harder to catch. The exploded view below lays them out; open each to see what it is, the scenario where it bites, and how to avoid it. Read them not as a list to skim but as a diagnostic of your own last hiring decision, because you have almost certainly made at least one of them.

Notice the through-line: almost every mistake is a failure to apply your own standard of proof and ownership to the decision. You would not run a channel you could not measure; do not hire an agency whose results you cannot measure. You would not optimize toward a metric that does not map to revenue; do not judge an engagement on one. You would not let a vendor hold your data hostage; do not let an agency hold your pixel and accounts. Seen this way, the mistakes are not mysterious — they are the same errors you already know how to avoid, made in a context where you forgot to apply the rule. The table summarizes each mistake, why even experienced marketers make it, and the one question that would have caught it.

The sections that follow take the most consequential of these in depth. As you read, keep asking the same question about your own situation: would I accept this from a channel, a tool, or a report? If the answer is no, you have found a standard you are failing to apply to the decision.

Mistake 1: Choosing on ROAS Claims You Cannot Verify

Picture the pitch. An agency walks you through a deck of case studies: an 9x ROAS for a brand in your space, a halved cost per acquisition for another, a screenshot of an ad account with impressive numbers. It is confident, specific, and exactly what you want to hear, and the urgency of your own situation makes you want to believe it. So you hire them on the strength of the claims — and three months later you are staring at a dashboard that shows a healthy ROAS while your actual revenue has not moved, and you cannot explain the gap. This is the first and most common mistake, and it is one a performance marketer should never make: choosing on results you cannot verify.

The problem is that a ROAS claim from someone else's account tells you almost nothing you can trust. You do not know how it was measured, whether it is platform-reported or reconciled to real revenue, what attribution window inflated it, whether the account was already succeeding before the agency arrived, or whether the case study is a survivor cherry-picked from many failures. A screenshot is not proof; it is a marketing asset the agency produced to sell you, and you know better than anyone how flattering a number can be made to look inside a platform's own dashboard. Ask yourself: if a channel showed you this exact number, would you accept it at face value, or would you demand to reconcile it against booked revenue? You would demand reconciliation — so demand it here too.

Avoiding this mistake is a matter of applying your normal standard of proof. Ask the agency to show results reconciled to actual client revenue, not platform-reported ROAS — and watch how they respond, because the ones who understand measurement will engage with the question and the ones selling a story will deflect. Ask what they measured and how; ask whether the account was already performing before they arrived; ask for a client you can actually speak to about the substance, not just the number. And judge them partly on their answer to the measurement question itself, because an agency that treats measurement seriously is telling you something more important than any case study: that they optimize toward real results rather than dashboard numbers. The claim is not the evidence; the willingness to be verified is.

Mistake 2: Giving Up Ownership of Your Measurement and Accounts

This is the mistake that hurts most and shows up latest, which is exactly why it is so dangerous. In the interest of convenience — or because it was simply framed as the normal way to work — you let the agency set up and own the ad accounts, the pixel, the conversions API, the tracking, and the reporting. It all runs smoothly, the numbers flow, and you never think about it until the day the relationship ends and you discover that your acquisition capability, your data, your account history, and your measurement all belong to the agency, not to you. You are not switching agencies; you are starting over, rebuilding from scratch the assets you thought were yours.

Consider a growth lead who ran paid acquisition through an agency for two years, scaling nicely, everyone happy. When they decided to bring things in-house, they found the ad accounts were in the agency's business manager, the server-side tracking ran on the agency's infrastructure, the audiences and conversion history lived in the agency's tools, and the creative sat on the agency's drives. Reclaiming any of it became a negotiation, and some of it — the accumulated account history and audience signal that made the campaigns work — could not be reclaimed at all. Two years of compounding assets, gone, because ownership was never established at the start. Ask yourself right now: if you parted ways with your agency tomorrow, what would you actually keep? If you cannot answer confidently, you have already made this mistake and just have not felt it yet.

The fix is non-negotiable and should be settled before any work begins, not after: you own the ad accounts (in your business manager, under your billing), you own the pixel and server-side tracking (on infrastructure and credentials you control), you own the conversion and audience data (with a copy in your own systems), and you own the creative. The agency operates these on your behalf with access you can revoke, rather than owning them with access you depend on. This is the single most important thing to get right, because it determines whether you are building a capability you keep or renting one you will lose — and as a performance marketer, you understand better than most that the accumulated data and account history are assets, not conveniences. Never trade ownership of them for the convenience of letting someone else set them up.

Mistake 3: Matching the Wrong Structure to Your Stage

The third mistake is choosing a structure — a particular kind of agency, or an in-house hire — that does not fit the stage your business is actually at, usually because the pitch described a stage you are not yet at. The classic version is hiring a scaling agency, one whose expertise is pouring fuel on a working fire, when your funnel is not yet validated and there is no fire to fuel. They cannot scale what does not work, so they spend your budget trying to force volume through a funnel that has not proven it converts profitably, and both sides grow frustrated as the promised scale never materializes — not because the agency is bad, but because you hired the right agency for the wrong stage. The mirror-image mistake is building in-house too early: hiring an expensive senior media buyer when you do not yet have enough sustained, complex work to keep them fully utilized, so you are paying full salary for partial usefulness.

The question to ask yourself is diagnostic and uncomfortable: is my funnel actually validated? Do I have evidence that paid acquisition converts profitably at small scale, so that the job is now to scale a working thing — or am I still trying to find out whether it works at all? These are completely different jobs requiring completely different help. If you are validating, you need focused senior expertise to figure out whether and how paid works for you, often a specialist or a lean engagement, not a scaling machine. If you are scaling a proven funnel, you need the capacity and systems to grow it. Hiring for the stage you wish you were at, rather than the stage you are at, wastes months and budget.

For the in-house-versus-agency version of the stage question, the honest test is utilization and breadth. An in-house hire makes sense when you have enough sustained, complex work to keep genuinely skilled people fully occupied, and when the tight integration with your business is worth more than the flexibility of an external team. Below that threshold, you are either underutilizing an expensive hire or asking one person to cover media, creative, and measurement — three specialisms — at a level none of them reaches. Ask yourself: do I have enough work, right now, to keep a senior in-house marketer busy and challenged, across the disciplines I need? If not, an agency or an embedded team gives you the capability at the level your work justifies, and you scale toward in-house as the work grows. Match the structure to the stage, and revisit it as the stage changes.

Mistake 4: Treating It as Media Buying When Creative and Measurement Are the Levers

The fourth mistake is scoping the engagement — or the in-house role — around media buying alone, when on modern platforms the media buying is increasingly automated and the real levers are creative and measurement. You hire someone to 'run the ads', they run the ads competently, and performance stalls anyway, because the thing that would actually move the number is a steady supply of good creative and a clean measurement signal, neither of which was in scope. You optimized the part that automation has commoditized and ignored the parts where the leverage now lives. As a performance marketer you probably know this intellectually, and yet the media-buying framing is so ingrained that the hire or the scope quietly defaults to it.

Consider a team that hired a strong media buyer to fix flat Meta performance. The buyer restructured campaigns, tuned bidding, cleaned up the account — all competent work — and performance barely moved, because the account was being fed the same tired creatives and the same degraded conversion signal it had before. The bottleneck was never the media buying; it was creative fatigue and signal loss, and no amount of account optimization addresses those. The team had solved the visible problem and left the real one untouched. Ask yourself about your own situation: when you imagine the help you need, do you picture someone adjusting campaigns, or someone producing a stream of creative and building a measurement layer you can trust? If it is the former, you may be about to make this mistake.

Avoiding it means scoping around the actual levers. Whether you hire an agency or build in-house, the engagement should explicitly include creative production at the volume the platforms demand and measurement you can trust, not just media management — and if an agency's pitch is all about their media-buying prowess with creative and measurement treated as your problem or an afterthought, that is a signal they are selling the commoditized part. Ask any prospective agency how much creative they produce and how they handle measurement and signal loss; ask any in-house candidate the same. The quality of those answers tells you far more about whether performance will actually improve than any account-management credential, because on today's platforms, media buying is the table stakes and creative plus measurement are the game.

Mistake 5: Accepting Incentives That Reward Spending, Not Earning

The fifth mistake is accepting a fee structure whose incentives point away from your interests, and the most common culprit is the percentage-of-ad-spend model. It sounds fair and it is the industry default, which is precisely why it slips past even careful marketers: the agency takes a percentage of what you spend on media, so as your budget grows, so does their fee. The problem is the incentive this creates — the agency earns more by spending more of your money, not by making more of it, so at the exact moment when the disciplined move is to hold or cut spend on a saturating channel, the pricing model quietly rewards the opposite. You have hired someone whose pay rises when they do the one thing that most often costs you money.

The scenario writes itself. A brand on a percentage-of-spend deal watches spend climb steadily quarter after quarter, always with a plausible reason to increase it, while efficiency slowly erodes and no one on the agency side ever proposes pulling back. Why would they? Every increase raises their fee, and every proposed cut lowers it. The agency is not villainous; it is responding rationally to the incentive you agreed to. But the drag on your economics — the spend that was never questioned because questioning it was against the agency's interest — compounds over a year into a real cost, one that dwarfs any difference in the headline fee. Ask yourself plainly: does the way I pay my agency reward them for spending more of my budget, or for making my budget more efficient? If it is the former, you have built a conflict of interest into the relationship.

The fix is to choose a structure whose incentives align with your profit. A flat retainer decouples the fee from your budget, so the agency has no reason to push spend and every reason to make it efficient. A hybrid of a base fee plus a component tied to a genuine business outcome aligns the bulk of their upside with your results. Whichever you choose, the test is simple: will they proactively recommend cutting spend on a channel that has stopped performing, or does their pay depend on you never doing that? Ask the question directly in the negotiation and listen for whether they acknowledge the incentive honestly or wave it away. As a performance marketer, you understand incentives drive behavior better than anyone — apply that understanding to how you pay the people spending your money.

Mistake 6: Judging on the Wrong Metric, or Too Early

The final mistake is evaluating the engagement on a metric that does not map to your business, or on any metric before the funnel has had time to mature — and it cuts both ways, causing you to keep bad engagements and kill good ones. The wrong-metric version is judging on platform-reported ROAS, which the agency's own tools produce and which is subject to attribution inflation, rather than on incremental, reconciled results that reflect what actually happened to your revenue. A dashboard showing a strong ROAS can coexist with flat or declining real revenue, because the platform is claiming credit for conversions it did not cause. You, of all people, know that platform-attributed ROAS and true incremental impact are different numbers — so do not judge your agency on the one you would not trust from a channel.

The too-early version is the opposite error: killing an engagement or a channel before it has had time to work, because the numbers had not moved in the first weeks. Paid acquisition, especially into a considered purchase, takes time to ramp — the algorithms need to learn, the creative needs to be tested, the funnel needs cycles to reveal its true performance. Judge it in the first three weeks and you are judging noise, and you may cut something that was about to work or, worse, conclude that a fundamentally sound approach failed. The scenario is common: a team pulls the plug at week four, having given a considered-purchase funnel no chance to mature, and concludes the agency or the channel does not work, when the truth is they never let it run long enough to find out. Ask yourself: am I judging this on a number that maps to my actual revenue, and have I given it a fair horizon for my funnel's real cycle?

Avoiding both errors means deciding, before you start, what success looks like and over what horizon — and defining it in terms of incremental, reconciled results on a timeline that matches your funnel, not platform-reported numbers on a timeline that matches your impatience. Agree with the agency (or set for your in-house team) what the meaningful metric is, how it will be measured against reality, and how long the fair evaluation window is given your sales cycle, and then hold to it rather than reacting to the daily dashboard. This is simply the measurement rigor you already have, applied to the evaluation of the engagement itself — and it protects you from both the false confidence of an inflated ROAS and the false despair of a funnel judged before it matured. Get the metric and the horizon right, and you will make sound decisions about who is actually working; get them wrong, and you will reward the wrong people and punish the right ones.

Agency or In-House — and the Questions to Ask Before You Decide

Underneath all six mistakes is the agency-versus-in-house choice itself, and the honest answer is that neither is universally right — the correct choice depends on your stage, your scale, and how central paid acquisition is to your growth, and it changes as those change. The table below lays out when each tends to fit, but treat it as a starting frame, not a rule, because your specific situation overrides the general pattern. The real skill is not picking a side once and forever; it is matching the structure to where you actually are, and being willing to change it as you grow.

Whichever way you lean, run the decision through the questions that catch the six mistakes, because asking them is the whole discipline. Before you sign or hire, ask: Can I verify your results against real revenue, not a platform dashboard? What do I own — accounts, pixel, data, creative — and what do I keep if we part ways? Is this structure right for my actual stage, or the stage the pitch describes? Does the scope include creative and measurement, or just media buying? Does your pay reward spending my budget or making it efficient? And how will we define success, on what metric, over what horizon? Every one of these is a question you would instinctively ask about a campaign; the mistake is failing to ask them about the decision.

The meta-lesson is the one we started with: you already have the discipline to make this decision well — you apply it to your work every day. The mistakes happen when the pressure to show results fast makes you suspend that discipline at the moment of choosing who does the work. So slow down enough to apply your own standards to the decision, verify what you are told, own what matters, match the choice to your reality, and judge it on numbers you trust over a fair horizon. Do that and you will avoid the errors that trip up marketers who should know better — and if you want a second, independent read on an agency you are evaluating or an in-house plan you are weighing, that pressure-testing is exactly the kind of thing our team is glad to help with.