How to check your media buyer's work: 12 questions and what the report must include

The owner doesn't need to know how to set up campaigns. You need to be able to read four lines: how much money was spent, how many leads came in, what one lead cost, and how many of them actually paid. Everything else is execution detail that matters to the contractor, not to you. If these four lines don't add up, there's nothing else to dig into.

The fastest way is to open the ad account yourself and check it against the latest report. Spend should match down to the dollar, and the number of results should match the leads in your CRM or Direct. A mismatch in either of these is the entire diagnosis — everything after that is a "why" question, not a "is there a problem" one.

At minimum: administrator access in a Business Manager registered under your name, not the contractor's. That same account should hold your Page, pixel, domain, and catalog. That way, if you switch contractors, you're not just taking a login and password with you — you're taking the entire history: audiences, conversions, campaigns that have already learned, and years of statistics.

If the account belongs to an agency or freelancer, you're renting your own advertising history. Ending the relationship means starting from zero — a new pixel, new data, and the algorithm learning all over again. We covered this in detail in who owns the ad account and data when working with an agency.

A normal reaction from a contractor when you ask for admin access is: "sure, send me your Business Manager ID." A red flag is an explanation of why the owner "doesn't need it," why it will "complicate things," or why it will "break the campaigns." Administrator access breaks nothing: you can see the data and aren't obligated to touch anything.

A proper report is a table for a fixed period broken down by campaign, not a collage of screenshots. It shows where every dollar went and what it brought back. A screenshot with no period, no currency, and no comparison to the previous month isn't a report — it's just an illustration for a conversation.

The required minimum of metrics is below. The formula and logic behind each one are broken down in our guide to 10 marketing metrics with examples: to ask your contractor the right questions, understanding CPL and ROAS is enough — the rest is their job.

These questions don't require any technical background and they don't sound like an interrogation. Someone who's actually running the project answers them on the spot, because the numbers are in front of them every day. Someone who isn't will start steering the conversation toward algorithms, competitors, and "niche specifics." Ask them one after another and listen for where the vague answers start.

• How much money was spent this month, and how much of it went toward testing?
• How many leads came in, and where are they tracked — CRM, a form, Direct?
• What's the CPL for each campaign individually, not on average?
• How has CPL changed compared to last month, and what exactly caused it?
• How many of these leads became clients, and for how much?
• What's the ROAS, and what formula are you using to calculate it?
• Which campaigns did you turn off this month, and by what criteria?
• How many new creatives were launched, and which one gave the lowest cost per lead?
• Which audiences were tested, and which one won?
• Are the pixel and events set up, and how did you verify that conversions are actually being tracked?
• Where is the traffic being sent, and who's responsible for the landing page?
• What specifically will you change next month, and which metric are you aiming to improve?

Open the ad account, set the period to "Last 30 days" and the level to "Campaigns." Then check these columns: spend, results, cost per result, impressions, CTR, frequency. This set is enough to tell whether the project is actually being managed. You don't need to change any settings — just read.

• Compare the total spend in the ad account with the amount in the report — they should match down to the dollar.
• Check the last-edited date on the campaigns: if nothing has changed in a month, no one's been managing it.
• Count the unique ads for the period — how many creatives were actually tested.
• Check the impression frequency: when it keeps creeping up and CPL rises with it, the audience has burned out.
• Compare "results" in the ad account with actual leads in your CRM or DMs.
• Look at how many campaigns are running at once on a small budget — budget scattering shows up immediately.

Bad work almost always looks the same: numbers get replaced with words. Instead of a table — screenshots, instead of CPL — "reach is growing," instead of a plan for the month — "we're continuing to optimize." Every line in the table below is something we've heard from business owners who came to us for an audit after the budget was already spent.

A special note on "the algorithm is still learning" — for the third month running. The learning phase is measured in days, not quarters: in our projects, the first leads come in on day 1–3, and stability arrives by day 14–30. If campaigns stay stuck in the learning phase for months, the problem is in the structure, the conversion event, or the budget — not the platform's patience.

Lead count without quality is the easiest metric to manipulate. You can drive cheap leads in almost any niche — the question is how many of them are even your audience. That's why the report should show the share of qualified leads next to CPL: in the auto case, 45% of 4,263 leads were qualified; in the fitness project, with 784 leads a month, it was 90%.

Who exactly qualifies leads should be agreed on at the start. The minimum working setup: you or your manager mark statuses (qualified / not qualified / purchased), and the contractor checks these statuses once a week and shifts optimization toward the sources that bring in buyers. Without this loop, a media buyer is optimizing blind — and can honestly show you a great-looking but worthless CPL.

You can compare your cost per lead with the market using our roundup CPL by niche — real numbers starting at $1.95. If your CPL is three to four times higher than the niche benchmark and hasn't moved in months, that's a reason for a conversation, not more patience.

Some failures have nothing to do with the setup. A budget below the working minimum gets spread too thin across campaigns and doesn't give the algorithm enough data to learn from — that's exactly why we work with a minimum ad budget of $600/month. A weak offer, an off-market price, or a landing page that loads slowly will ruin any traffic.

The second common cause is follow-up handling. A lead you respond to after six hours costs the same as one you respond to in five minutes, but converts far worse. Before changing contractors, check your first-response time and how many leads even got a call.

An honest contractor brings these things up themselves and shows you where your side of the responsibility lies. One who spends a whole year blaming zero sales purely on "the algorithm" and never once mentions the offer or lead handling simply isn't looking at your business.

If you have access, the numbers in the report match the ad account, and CPL is improving even slowly — this is a working process that needs a monthly plan, not a replacement. If you don't have access, the report is made of screenshots, and none of the 12 questions got a clear answer — you're paying for presence, not results.

A middle-ground step before parting ways is to show the ad account to a second pair of eyes. An independent specialist can review the campaign structure, event setup, change history, and creative quality within an hour. This isn't about "poaching a client" — the conclusion is often "everything's fine, here are three things worth improving."

If you decide to change your setup after the audit, compare your options in agency vs. freelance media buyer — 8 criteria, and check out our results in case studies. Over 4 years and 150+ projects across 40+ niches, we've seen practically every version of what an ad account looks like "before."