A cheap Meta ads retainer can still be your most expensive line item. Spend $3,000 a month on media and $975 on management, and you add 33% overhead before Meta shows a single impression. I have seen owners cut the cheapest expense and leave the expensive leak running. My rule: the cheapest workable setup is the one that covers the work, measures customers and leaves enough money to advertise. Run these five gates before you pay a manager or move budget. Then keep what works, cut what duplicates, or outsource one defined job.

Gate 1: Count the full bill, not just the retainer

  1. Write down your monthly cash limit before reviewing proposals. Include everything you can spend on Meta advertising, not just the amount you intend to send into the auction. A management fee does not come out of a separate, imaginary budget.
  2. Split that limit into four lines: media paid to Meta, management, creative production, and tracking or tools. Keep the lines separate even when a supplier bundles them into one invoice. You need to know what disappears if you cancel that supplier.
  3. Use published management-price ranges as a starting point, not a verdict: agency packages commonly run $500 to $1,400 a month, while freelancers may charge $20 to $45 an hour. Neither number tells you what work you receive.
  4. Divide management by media to calculate your management overhead rate. A $1,000 fee on $2,000 in media adds 50% overhead; a $975 fee on $3,000 adds roughly 33%. This is a cost check, not proof that either arrangement performs badly.
  5. Calculate total cash out separately. With $3,000 in media and a $975 management fee, you spend $3,975 before adding any separately billed creative or tools. Do not confuse management as a percentage of media with management as a share of the total bill.
  6. Ask what the quoted fee excludes. Creative, tracking setup and ongoing edits can sit outside management. A low retainer stops looking low when the work required to keep the account running arrives on a second invoice.
  7. Write down the media left after every required fee. Gate 2 checks whether that amount can support your acquisition target. Do not approve a package because the retainer fits while the advertising budget quietly disappears.

Receipt dividing Meta advertising costs into media, management, creative and tracking

Gate 2: Check whether the result is worth buying

  1. Make a test purchase or submit a test lead. Confirm that it appears in your own records and in Events Manager, with the correct value where applicable. A dashboard full of conversions is not useful if those conversions do not match the business.
  2. Identify the event the campaign optimizes toward. Meta follows the signal you provide; a form submission and a paying customer are different signals. Read the CRM feedback explanation before treating a low cost per form fill as evidence of profitable acquisition.
  3. For lead generation, count qualified leads and booked jobs, not raw form fills. The common mistake is paying for a falling cost per lead while qualification falls faster. Put the qualified outcome beside the platform number before judging the manager.
  4. Check whether lead-stage changes return to Meta through your CRM connection. If you use Conversions API for CRM, confirm the connection carries the identifiers and stage information it needs. Do not assume connecting a CRM automatically means the campaign learns from qualified outcomes.
  5. Set an affordable customer acquisition cost from margin and repeat value. For lead generation, include the close rate when deciding what you can afford to pay for a lead. Write down which outcome each target measures; a lead CPA and a customer CPA are not interchangeable.
  6. Divide weekly media budget by the expected cost of the optimization event. That gives you a rough event count to compare with the usual learning benchmark of about 50 events per ad set after a major edit. Treat it as a planning check, not a guarantee or an automatic reason to stop.
  7. Do not make a keep-or-cut decision without an affordable CPA. If you cannot name the cost the business can carry, you cannot tell whether a cheaper lead, a lower retainer or a new platform improves the economics.

Gate 3: Buy the missing job, not the whole department

  1. Mark which work you cannot cover: creative production, account adjustments or tracking fixes. Be specific about the gap. “Someone should manage this” is a feeling; “we have no new ads ready” is a job you can scope and price.
  2. List the work you already handle. Include comments, basic budget checks and any creative you reliably supply. Do not buy those tasks again by default, but do not count work as covered merely because you intend to get around to it.
  3. Ask how many new images and videos the fee includes each month, and who pays for revisions. Management often excludes creative, and freelancers may expect you to supply the ads. “Creative support” is not a delivery schedule.
  4. Check whether the proposed creative volume is a testing plan or simply a small monthly batch. Two static images a month should not be sold to you as a complete answer to testing and fatigue. Get the actual deliverables into the scope.
  5. Ask who handles account adjustments and who fixes tracking. Keep those responsibilities distinct from reporting. A monthly report can describe a problem without anyone being responsible for correcting it; make the owner of each job explicit.
  6. List every supplier charging for creative, tracking or reporting, then run groas’s double-payment audit. Check the deliverables against one another before adding a retainer. The same setup or monthly report should not become two separate purchases.
  7. Remove confirmed overlap before buying more help. Cancel a tool only if its useful work is genuinely covered elsewhere; challenge a creative retainer that delivers one batch a quarter when your gap is regular production. Cut duplication, not the function you still need.
  8. Outsource the specific missing work. If you buy account management but continue supplying creative late, the missing piece remains missing. A new person checking the account does not manufacture the ads you have not delivered.

Overlapping invoices for Meta management, creative and tracking

Gate 4: Make a new platform prove its economics

  1. Separate the decision to change managers from the decision to change platforms. A weak management arrangement does not establish that Meta is the wrong place to advertise. Fix the scope and measurement questions before using a platform move to escape them.
  2. Compare customer outcomes in your records, not headline click prices. A cheaper click and a cheaper customer are different purchases. Use the affordable acquisition cost from Gate 2 as the test; a CPC comparison cannot make that decision for you.
  3. Treat reports of higher ChatGPT conversion costs after shifting Meta budget as a warning to test, not proof that one platform always wins. Another advertiser’s lead cost does not tell you what your customers will cost, or whether their definition of a lead matches yours.
  4. Write the test budget as an additional commitment before moving money. Include media, any new management work and tracking costs. A platform switch is not a saving merely because you stop paying one supplier and start paying another.
  5. Decide how much Meta media the test would remove. Return to your expected event count from Gate 2 and check what remains. Do not leave the existing campaign with too little budget to evaluate, then call its weaker results evidence that moving was correct.
  6. Verify tracking on the platform you are testing before interpreting its results. For your Meta setup, check the pixel and Conversions API connection. Do not assume those checks establish that a different platform records conversions correctly, or read an untracked zero as zero demand.
  7. Keep Meta running while testing elsewhere only if the combined budget supports both. If it does not, record that constraint rather than pretending a thin split gives you a useful comparison. Moving the money still carries a cost, even when the new platform looks interesting.
  8. Fix broken measurement before making the move. A new platform does not repair the missing link between leads and customers. If Gate 2 fails, return there before you pay to collect another set of numbers you cannot judge.

Gate 5: Price the exit before you sign

  1. Write down the minimum term, cancellation notice and any setup fee. Published management-cost research includes both contracts with no minimum and commitments lasting 6 to 12 months, with setup fees charged separately. The monthly headline does not show the full commitment.
  2. Calculate what you owe if the arrangement is wrong for you. A cheap retainer with a 12-month lock-in is not the same purchase as a month-to-month service. Price the obligation you sign, not just the first invoice you expect to pay.
  3. Inventory the ad account ID, pixel or dataset ID, Business Manager ID, and the owners of the Facebook page and Instagram assets. Do this before signing or cancelling. An access list is more useful while everyone is still answering your emails.
  4. Check who owns each asset and who merely has permission to use it. Partner access and ownership are not the same thing. Do not rely on a manager’s ability to open the account as evidence that the business controls it.
  5. Get cancellation and handover in writing: which files and assets you receive, in what format, and within how many days. Include access changes and tracking responsibilities, not just the final report. “We will help with the transition” is not a handover plan.
  6. Read refund language literally. A 7-day money-back window is not a performance guarantee; it refunds the fee only under the stated terms. Do not use a short refund window to justify a longer commitment you would otherwise reject.
  7. Follow a documented access handover when switching managers. Use this agency-switching guide to check the sequence, add the new manager yourself and remove the old one yourself. Keep control of the transition rather than asking suppliers to sort it out between them.
  8. Submit another test lead or purchase after the handover. Confirm that the event appears correctly and check deduplication where applicable. A tidy permissions screen is not proof that tracking still works; finish the switch with a working test.

Padlock on a Meta ad account folder beside a cut contract

Make the decision: keep, cut, outsource—or buy nothing

  1. Put four decision inputs on one page: overhead rate, affordable CPA, the missing job and exit cost. Use the same page for every proposal. A persuasive sales call should not change the arithmetic you require the next supplier to beat.
  2. Keep the arrangement when tracking works, qualified leads or purchases tie to your records, and the necessary workload is covered without overlap. Before cutting something useful, check 6 Meta Ads ‘Savings’ That Cost More Than They Save.
  3. Cut the duplicated expense when two suppliers charge for the same work. Also challenge a management fee that excludes creative when nobody has a plan to supply it. Removing the fee alone does not solve the missing-work problem; make both decisions explicitly.
  4. Check combined overhead before keeping separate suppliers. Pay $800 for management and $600 for creative on $2,500 in media, and those two fees add 56% overhead. That does not prove the arrangement is wrong, but it is the bill the results must justify.
  5. Outsource when the missing work is specific, the combined cost leaves a workable media budget, and the scope assigns responsibility. Buy execution against that scope, not another set of recommendations you must find time to implement yourself.
  6. Buy nothing if none of those choices passes. A retainer bought to feel that something is happening still consumes money. Pause the duplicate expense, repair the tracking or define the job before purchasing a package that fails the same gates.
  7. Open Business Manager before you pay or move anything. Screenshot each asset owner, add your overhead rate and affordable CPA, and file the page where the next proposal has to beat it. Make this the check you complete, not the checklist you save for later.
  8. Confirm asset ownership—the item owners skip most. Keep the ad account, pixel and dataset under your business’s control, and grant the manager partner access. Losing control can cost you the history and learning you already paid to build. Check before signing, or your next budget may pay to rebuild them.