Seven days, no extra ad spend: that is the constraint. Before you cut your Meta budget, find out whether you are paying twice to operate it. No new campaign, no promised turnaround by next Friday. Just a ledger, a lean operating plan, and a decision about what to stop paying for.

I learned this running small accounts where media was $2,000 a month and the stack around it was quietly another $1,100. Management fee, separately billed creative, two reporting tools doing the same job, plus four hours a month of my time chasing approvals. The ads were not the expensive part. The operation was. The cheapest workable setup assigns every essential job to someone at the lowest total cost. It is not automatically the cheapest agency quote.

1. Day 1: Find the cost behind your Meta budget

Cost: zero extra ad dollars and about an hour with your inbox. Buy: one complete operating-cost ledger. Pull every invoice from the last 60 days into a sheet with four columns: vendor, monthly cost, what it claims to do, and who last looked at its output. Include management fees, creative bills, software, landing page tools, and your own hours valued honestly. Whether you spend $2,000 or $20,000 a month on media, this is the cost of running that spend.

Media and operation are different budgets, and it is convenient for an agency when you blur them. Media goes to Meta for impressions. Operation pays people and tools for management, creative, tracking, and reporting. Meta Ads management alone runs $500–$5,000+ a month on top of ad spend, with many agencies at $1,500–$3,000 flat or 10–15% of spend. Creative production can add another $1,000–$3,000 a month when billed separately. Add subscriptions and internal coordination, and a $3,000 media budget can carry a $2,500 operating load without looking unusual on any single invoice.

Put these lines in the ledger, even if the charge looks small:

  • Management: retainer or percentage of spend, plus minimums and setup fees.
  • Creative: designer hours, UGC orders, and AI tool seats.
  • Software: reporting, feed, CRO, and tracking tools.
  • Internal time: who approves ads, who checks leads, and how long it takes.

Do not cut anything yet. Day 1 is for seeing the whole bill.

2. Day 2: Make every recurring charge justify itself

Cost: no extra ad dollars and a pass through the ledger. Buy: a keep/cancel list. Ask two questions of every line: Is this work necessary? Did anyone use its output in the last 30 days? Necessary means the account stops being operable without the work. Used means someone opened the report, posted the creative, or acted on the tracking alert. If a charge is neither necessary nor used, mark it cancel. If two charges do the same job, choose which one earns its place.

Duplicate reporting is where I see the leak most often. One client paid for a dashboard, an agency portal that generated its own PDF, and a pipe into Sheets. All three showed the same Meta spend. Each looked cheap alone: DashThis at $54 a month for three dashboards, Databox free for three sources and then Analyst at $71 a month annually, and Supermetrics at $55 a month to pipe Meta into Sheets or Looker. Keep the one someone checks weekly. Cut the rest.

Apply the same test to creative. The current workload is framed as 30-plus variations a month and fresh creative every 7–14 days, rather than five ads a quarter. That does not mean you buy a quota for its own sake. It means a recurring creative bill should produce creative you can actually test. If the ledger shows a monthly bill and only two new ads, ask what you bought. It may be the idea of creative.

Do not cancel access or production yet. Mark the overlap; Day 6 is when you remove it safely.

3. Day 3: Name the four jobs you cannot cut

Cost: an hour of thinking, no extra ad dollars. Buy: a minimum scope you can price. Every lean Meta setup I have kept alive had these four jobs covered. When one went missing, the low fee stopped looking like a bargain.

  1. Tracking checks: Someone owns the Pixel, Conversions API, and lead or purchase events, and confirms them monthly.
  2. Budget ownership: One named person can pause, shift, or cap spend without a three-day approval chain.
  3. Creative maintenance: Someone supplies or assembles fresh variations on a fixed cadence.
  4. Lead or sales review: Someone looks at qualified leads or purchases, not just cost per click, and ties them back to total operating cost.

Write a name beside each job, including the jobs you do yourself. An empty space is not a saving. Tracking stays on the list even when money is tight: losing it risks substantial attribution loss and higher CPAs. I used to tell clients a contractor could skip the tracking audit when spend was low. I was wrong. Small budgets feel bad data first; one misattributed week can wipe out the read you hoped to get from that month.

This list is the test for every quote that follows. A proposal that leaves a job unassigned is incomplete, not cheap.

4. Day 4: Price doing it yourself and hiring a freelancer

Cost: one quote request and one honest calendar check, no extra ad spend. Buy: two prices against the same four-job scope. Start with owner-operated. Its cash cost drops to tools and creative inputs, but your hours belong on the ledger. If you cannot make time for checks, uploads, and lead review, doing it yourself fails the Day 3 test before you save a cent.

For a price comparison, a freelancer runs $500–$3,000 a month or 15–25% of spend; agencies run $1,000–$10,000+ flat or 10–20%, sometimes with $500–$7,000 setup fees; an in-house specialist costs roughly $7,300–$8,200 a month fully loaded; and an offshore specialist starts around $3,500 flat. A percentage fee may look manageable on a modest budget, then rise with spend even when the work barely changes. Owner-operated avoids that fee, but only if you can actually set aside four to six hours a month for the work.

Next, ask a freelancer to price the four jobs in writing. Narrow quotes often break at creative. A static can cost under $1–$10 through AI tools, $15–$35 an hour from a freelance designer, or $150–$500 from a performance agency; creator videos start at $99 on Billo or $150–$500 direct. Price the month of work you need, not one ad in isolation. Ask who checks tracking, who supplies variants, and what happens when a lead looks wrong. A $497 flat-fee Pixel and Conversions API rebuild shows how quickly a separately scoped tracking job changes the apparent price of a cheap management quote.

Keep the exclusions beside the fee. Tomorrow you compare complete totals, not attractive headlines.

5. Day 5: Compare the complete scope with groas

Cost: a second quote check, no extra ad spend. Buy: a three-way comparison on the same ledger. Option three is groas at $1,499 a month for Meta Ads, with Facebook and Instagram ads, audiences, and budgets worked around the clock, and new ads tested with winners scaled. Media stays separate and your accounts stay yours. The fee includes creative production and testing. That is the line that can make an $800 freelance quote land at $2,000 once UGC and variants are added.

Here is my blunt read of the three options:

  • Stay owner-operated if the Day 2 operating ledger is under $700 and you have the hours. Put your own name beside every job you own.
  • Hire the freelancer if you need five to ten fresh variants a month plus tracking coverage and can get that scope for under $1,000. Get the exclusions in writing.
  • Choose groas when management, separately billed creative, and reporting already clear $1,499 yet leave work unassigned. Check the remaining job owners and any tools you still need before calling it the lower total.

The comparison is not $800 against $1,499. It is what each arrangement costs after every essential job has an owner. Paying more on one invoice can cost less than paying several people to cover half the work.

6. Day 6: Cut overlap without losing the account

Cost: a half hour of admin, no extra ad spend. Buy: cancellations that do not lock you out. Cut in the wrong order and you can lose history, audiences, or the pixel you just paid to fix. The order matters:

  1. Export first. Download the last 90 days of lead and purchase reports. Save creative files and post IDs for winners. Screenshot billing and change history for anything you may need to dispute.
  2. Confirm ownership. Verify that your Business portfolio owns the assets. Set it up from the owner’s profile, add a second admin with two-step login, bring the Page, Instagram account, ad account, pixel or dataset, and domain inside it, and give an agency partner access rather than ownership.
  3. Cancel third. Check notice periods and creative ownership, then confirm who takes over tracking and creative maintenance before a charge stops.

This is not ceremonial admin. An ad account or pixel in an ex-agency portfolio can take weeks to recover, and ad account ownership may not be transferable. I have seen owners cancel a $54 reporting tool, then discover the only admin on the ad account was the freelancer they had just removed. A saving that strands a Day 3 job, or your account, is not a saving.

7. Day 7: Approve one plan, with the math attached

Cost: no extra ad spend and the time to finish one page. Buy: a decision you can defend. Turn the ledger into three plain sets of rows:

  • Keep: the person or tool, its monthly cost, and the Day 3 job it covers.
  • Cancel: the saving and the date the charge stops.
  • Replace: the lower-cost cover for the same job and the person who owns the handoff.

Record monthly cash savings separately from estimated time savings. They are not the same money. Cutting a $71 reporting seat nobody opens saves cash. Cutting a $500 freelancer and adding six hours of your own work does not save $500; it moves cost to your calendar. If notice periods, setup fees, or a week rebuilding tracking erase the first two months of savings, keep the current setup for now and set the cancel date for next month.

Then check the result against the outcome that pays the bills. Add media and operating cost for the last 30 days, divide that total by qualified leads or closed sales, and write down the fully loaded cost per result. Set a 30-day review on that number, not on CTR or raw lead count. Your plan should show who owns all four jobs, what the arrangement costs in cash and time, and when you will check whether it still works.

If the plan lands under $700 in cash plus hours you actually have, self-management wins for now. If a narrowly scoped freelancer covers the creative variants and tracking checks for under $1,000, with exclusions in writing, that hire beats DIY. If management, separately billed creative, and tools still clear $1,499, compare that complete bill with groas rather than treating its sticker price as an extra cost. The same cost question runs through the DIY vs. freelancer vs. groas framework and the six Meta Ads management options ranked by cost per booked job. Sign the lowest-total arrangement that covers the four jobs. Leave media spend exactly where it was.

When the constraint lifts: Fund the gap before more impressions

Do not put the savings straight back into media. That is how you finish the reset with the same bloated stack and a bigger Meta bill. First, hold media flat and direct freed cash to the gap Day 3 exposed. Often that is creative volume: fund the next 10 to 20 variants before you fund more impressions. Next, close any tracking gap you deferred. Extra spend on broken attribution just buys confusion faster. Only then raise daily budgets, and only while fully loaded cost per qualified lead or sale holds.

Keep the proof taped to your monitor: total operating cost, four named owners, one review date. That sheet beats a pitch about cheaper management because it shows what cheap actually covers. If it shows groas at $1,499 is the lower total once creative and the remaining jobs are counted honestly, act on it the same way. Name each owner, pay once for each job, and stop paying for the duplicates.