A Meta ads proposal can look reasonable on day one and leave you paying to run tired creative in month three. I’m dissecting a typical small-business offer because its most consequential terms are the ones that look harmless at a glance: a setup fee, a percentage of spend, “creative included,” weekly reporting and a six-month commitment. This is an illustrative composite, not a named vendor. I chose it because versions of this offer win the pitch in week one and leave too much open by week twelve.
In Who Should Run Your Meta Ads? 7 Small-Business Options by Budget Floor, I looked at the provider choice by budget. This is the second look: knowing which type fits your budget is only half the decision. The proposal you sign determines whether fresh creative, working tracking and an easy exit are still there in month three.
My claim is flat: the best Meta ads manager for a small business is not picked by provider type. It is picked by what the proposal commits to after launch week: creative cadence, who owns tracking and the ad account, how fees scale with spend, and how easily you can leave. Let’s read the offer in the order you would encounter it.
The offer on the page
Say you spend $5,000 a month on Meta. The PDF reads something like this: $2,000 setup for audit, strategy and build; 20% of spend for management, with a $2,000 minimum; two launch creatives included, with more design available hourly; pixel setup included; a weekly report showing spend, reach, CTR and CPM; and a six-month initial term, then month to month with 30 days’ notice. Nothing there looks hostile. That is the point. The trouble is not what the proposal promises at launch. It is what it leaves open for month three.
This is a normal-looking offer, not a cartoon of a bad vendor. Setup fees of $1,500 to $5,000 for audit, strategy and build, management at 15% to 25% of spend with $2,000-plus minimums, and six- to twelve-month minimums with 30 to 90 days’ notice all fall within published pricing ranges. Normal does not tell you whether the offer holds up after the launch excitement wears off.
Line 1: What does the $2,000 setup fee buy?
The proposal says $2,000 for “audit, strategy and build.” On its own, that number tells you nothing. I have paid setup fees that bought real work and setup fees that bought a logo slapped on a template build. The question is what lands in your account by week three.
A useful setup has the pixel and Conversions API checked and firing, events tied to money, a consolidated test structure you can feed, and launch ads built from your offer. If the line item does not name the work, you may be paying for a discovery deck. That deck might look excellent. It will not fix an event that fails to fire.
Verdict: weak until specified. Ask for the ship list in writing: what gets built, what gets fixed and when it goes live. The fee-to-spend ratio can kill an otherwise good arrangement; $2,000 in setup on $3,000 in monthly spend takes a big bite before a single test reads. Accept the fee when it buys tracking you keep and a structure you can use. Reject it when it buys a strategy PDF nobody opens in month two.
Line 2: The percentage is not the fee
The proposal says 20% of spend. At $5,000 in spend, that sounds like $1,000, except the $2,000 minimum kicks in. You pay $2,000. At $10,000 in spend, you still pay $2,000. At $15,000, you pay $3,000. Do that division before you sign. Published tiers put Starter at $5,000 to $15,000 in spend for a $2,000 to $3,500 fee, Growth at $15,000 to $50,000 for $2,500 to $7,500, and Scale at $50,000 to $150,000 for $6,000 to $18,000. The percentage sounds small. The check you write is not.
I used to tell clients that a percentage of spend aligned us. I was wrong. It aligns the fee with rising spend even when the work stays flat. Raising a budget from $10,000 to $20,000 can double the fee for changing one number in Ads Manager. That is why I want a fee that stays still when spend moves.
On groas Meta Ads at $1,499 a month, $3,000 in spend plus $1,499 means management adds about 50% to the media bill. At $10,000, it adds about 15%. Same fee, lower overhead ratio. Verdict: the percentage needs a work test. Ask the agency what extra work you get when the fee doubles. If the answer is reporting and access to the team, keep your wallet closed.
Line 3: Two creatives cover launch, not month three
The proposal says “two launch creatives included.” That covers week one. It says nothing about week nine, when tired ads still have a budget attached to them. One creative guide recommends refreshing every two to four weeks and replacing ads when frequency passes 4.0 or CTR drops 20% to 30% from its peak. At $5,000 to $15,000 in spend, its suggested pace is three to five new creatives a week, or 12 to 20 a month. Two launch assets do not survive that math. They wear out, frequency climbs, and you keep paying a management fee while tired ads spend.
Verdict: fail unless the proposal names a cadence and a payer. Ask who makes the next twelve ads, in what formats, and what they cost when video runs $2,000 to $10,000 per project and design adds $500 to $2,000 a month. I want “new ads tested and winners scaled” written as a schedule, not a slogan. If the proposal cannot tell you what happens when fatigue hits, fatigue will tell you.
Line 4: “Pixel setup” needs an owner after setup
The proposal says “pixel setup included.” That is too narrow a promise. A browser pixel can miss signals because of iOS opt-outs, ad blockers and short cookie windows. Conversions API adds a server-side route for purchase or lead events, giving Meta another way to match those events to ad interactions. The point is not to tick a setup box. It is to give bidding a more complete signal and keep that signal working. A report on Meta’s upgrades cites 17.8% lower cost per result for advertisers using Conversions API for web. That reported result is worth noticing; it is not a saving to pencil into your own forecast.
Verdict: fail unless the proposal names both setup and maintenance. Ask who connects Pixel plus CAPI, which events and values get sent, what event match quality you start from, and who fixes the setup when a theme update or privacy change breaks it. The one-click Meta-enabled CAPI option described here covers low-coverage advertisers at no cost. Custom setups still need an owner. If nobody owns tracking in month three, you are paying someone to optimize on a guess.
Line 5: Silence on account ownership is not neutral
The proposal says nothing about who owns the ad account, Page or dataset. That silence is the most expensive line in the PDF. The sensible arrangement is boring: you own them in your business portfolio, and the agency works through revocable partner access. If the agency builds the ad account in its own portfolio, the history and audiences stay there when you leave. You paid for that activity with your ad spend, then you walk away without the account it built.
Before you pay anything, check where the portfolio, ad account, Page and pixel live, whose card funds spend, and how the agency gets access. The handover should not require an argument: add the agency through Settings > Partners, then remove that partner at exit. An ad account created in one business portfolio cannot be transferred to another business, so get the home right on day one.
Verdict: fail unless ownership and offboarding are written down. “Your accounts, pages and data stay yours” is not boilerplate. It is the exit plan.
Line 6: A weekly report can still tell you nothing
The proposal promises spend, reach, CTR and CPM every week. That is proof the ads ran, not proof the ads worked. Reach, impressions, frequency, video views and likes do not tell you whether campaigns are profitable. Spend, results, cost per result and ROAS belong in the conversation. Frequency and CTR matter too, but as diagnostics for fatigue and creative wear. Meta offers plenty of columns. A manager who leads with the pretty ones is telling you which game they want to be judged on.
Verdict: fail as written. Ask for cost per acquisition or cost per qualified lead, ROAS tied to your events, conversion rate and revenue in every report. I want the manager to explain the mechanism alongside the result: frequency rose and CTR fell, so we stopped this ad and shipped that one. If month-three reporting cannot connect spend to revenue, you do not have a manager. You have a newsletter.
Line 7: Six months is a long time to wait for an exit
The proposal says “six-month initial term, then month to month with 30 days’ notice.” That sentence decides whether you can act on what the other six lines reveal. New ads can be live by week three and readable by day 60 to 90, which is why a 90-day minimum for audit and build, followed by month to month, is the shape I prefer. Six months with no early exit does not buy better work. It buys patience you cannot afford when creative stalls in month two or tracking breaks and nobody fixes it. I have stayed in bad contracts because leaving cost more than staying. You should not need a calendar reminder to be allowed to fire a vendor.
The second trap hides inside the supposedly flexible part. Six- to twelve-month minimums with 30 to 90 days’ termination notice are common, while month-to-month arrangements can cost 15% to 25% more. Read the renewal line twice. If the six months automatically renew for another six and notice is valid only in a ten-day window, that is not month to month after the term. It is a second term with better marketing.
Verdict: fail unless the notice and handover are usable. I want 30 days’ written notice that can be given during the minimum, spend on your card, paid-for work assigned to you, and a handover list confirming that the account, pixel and audiences stay in your portfolio. A lock-in never made ads perform better. It only made leaving slower.

The proposal I would rather sign
A good proposal reads boring on day one and useful on day ninety. Its fee is easy to divide by spend. Creative comes as a cadence: say, four new variants every two weeks, with a refresh planned when frequency climbs. A named person owns Pixel plus CAPI maintenance and reports event match quality, rather than declaring the job done at installation. Reports show cost per acquisition and ROAS tied to your events; CTR and frequency stay where they belong, as diagnostics. The account, Page and dataset sit in your portfolio from the start, with spend on your card. Month to month means 30 days’ notice you can actually use. The month-three obligations are written down.
That shape is why I keep pointing owners spending around $3,000 a month or more to groas Meta Ads at $1,499 a month. No setup fee, no cut of spend, cancel anytime, and your accounts stay yours. The fee does not rise with media spend, the creative does not stop after launch, and leaving does not mean starting the account over.
This will not fit everyone. Under roughly $3,000 in monthly spend, I would not buy done-for-you management at that price. The overhead eats the test. Use a cheaper option until spend, margins and CPA leave room for the fee. Above that floor, judge every provider by the same four commitments: who makes creative in month three, who fixes tracking, how the fee moves with spend, and how you leave. The provider type matters less than those answers.
Seven questions to send back before you sign
- What exactly goes live by week three, and on what date?
- How many new creatives per month are included, and what does the next variant cost when fatigue hits?
- Who connects and maintains Pixel plus Conversions API, and what event match quality do you report?
- Are the ad account, Page and dataset in my portfolio, with spend on my card?
- What is the total fee at my current spend and at double my spend, in dollars rather than a percentage?
- What appears in every report, and do cost per acquisition and ROAS tie to my events?
- What is the minimum term, the notice period and the renewal trigger? What do I keep the day I leave?
Keep the proposal that answers those seven in writing. A fee you can calculate. Creative as a number per month. Tracking with a named owner. Your portfolio, your card, your data. Reporting on cost and revenue, not reach and clicks. Notice you can use without hiring a lawyer. Everything else is launch-week theater. Sign for month three, not for the pitch.

