The operator's daily & weekly workflow
The finale. You now understand the whole machine - auction, structure, levers, creative, scale, measurement. Today turns knowledge into a repeatable operating rhythm: exactly what to do daily, weekly, and monthly. Your operating manual.
The hardest discipline (Day 5, Day 17): most of the time, the right action is NO action. The system needs stability to learn. A great operator checks often but TOUCHES rarely - and when they touch, it's one deliberate, diagnosed change. Activity is not progress.
1The daily routine (10-15 min)
A quick health check, NOT a tinkering session:
- Spend pacing - delivering as expected? Sudden under-delivery = a flag (cap too tight, learning limited, disapproval).
- Major anomalies only - CPA/ROAS wildly outside normal? Note it, but apply Day 17 (noise/learning/attribution?) before acting.
- Ad disapprovals / errors - anything rejected or not delivering.
- Signal sanity - events still firing? (Day 10)
Default daily action: observe and leave it alone. Daily edits reset learning.
Open your saved "Daily health" preset and scan top-down. Most rows are quiet - that's the point. You're hunting for the one or two things that genuinely need a hand today:
Action today: fix the rejection (request re-review or replace the asset). Everything else: observe - the "learning limited" flag is a weekly diagnosis, not a same-day edit. To build the preset: Columns ▾ → Customize columns → pick spend, results, delivery status → Save as preset ("Daily health"). Repeat for a "Weekly trends" preset (CPM, CTR, conversion rate, frequency, CPA).
2Policy & account safety
Disapprovals are the one daily item that can escalate into something expensive, so it's worth knowing the ground rules before you ever hit one. Meta's rules live in the public Advertising Standards (the Advertising Policies in the Business Help Center) - bookmark them; that page is the source of truth, not forum lore.
For DTC and lead-gen advertisers, the rejections you'll see most often cluster around a few categories: personal-attribute targeting in the copy ("Are you over 40 and struggling with…"), unsupported or sensational health/finance claims, before/after imagery, restricted-goods language, and landing pages that don't match the ad or are missing a clear privacy policy. Most rejections are about wording and the destination page, not the product itself.
When an ad is rejected, you can request another review from the ad's status - Meta re-checks it, often within a day. If you genuinely think it complies, request the review; if you can see why it tripped, fix the copy or asset and relaunch rather than appealing the same ad repeatedly.
Why this matters beyond one ad: repeated violations compound at the account level. A pattern of disapprovals can lead to ad-account restrictions that pause all your delivery, which is far costlier than any single rejected ad. Treat clean compliance as protecting the asset, not as red tape.
Spamming "Request review" on the same unchanged ad, or relaunching a rejected asset under a new name to dodge the system. Both read as evasion and accelerate account-level risk. Read the stated reason, fix the actual cause, then resubmit once.
3The weekly routine (the real work, 60-90 min)
Where decisions belong, on enough data to be meaningful:
- Performance review - 7-day trends on CPM, CTR, conv rate, frequency, CPA/ROAS (Days 13, 17). Shapes, not single days.
- Creative review - which won/lost/fatigued? Cut losers, note winners, queue iterations (Days 12-14).
- Scaling decisions - bump winning budgets gradually (~10-20%, Day 16); open horizontal fronts.
- Launch new tests - feed the testing campaign fresh concepts; graduate winners (Day 12).
- Signal & reconciliation - check event match quality; reconcile vs backend / blended (Day 18).
This weekly cadence IS the iteration loop (Day 14) operationalized.
4The monthly / quarterly routine
- Incrementality & new-customer view - run/refresh a holdout or geo test; check cost per NEW customer and blended ROAS (Day 19).
- Account structure audit - has it sprawled? Consolidate (Day 3).
- Strategic creative direction - which ANGLES win at the theme level? Brief the next wave.
- Budget allocation across campaigns/clients and goal-setting.
Day 1's three levers: the machine took two (targeting, bidding) - the third, creative, is yours, plus the signal and judgment that feed it.
- The auction rewards relevance → great creative wins cheaper.
- Structure + learning demand consolidation and patience.
- Broad targeting + automation free you to focus on the inputs that matter.
- Signal feeds the algorithm; creative is the lever and the ceiling of scale.
- Measurement honesty + incrementality keep you optimizing for real business impact.
You don't row (the algorithm rows). You set the destination (strategy), keep the instruments accurate (signal), choose the cargo and sails (creative), check heading daily but adjust course only with reason, and periodically verify you're sailing toward real land (incrementality), not a mirage of reported ROAS. A frantic captain spinning the wheel every minute capsizes the ship. A calm one with a steady rhythm and good instruments crosses oceans.
The modern stack clients pay for - strategy, creative, media buying - maps exactly to where value now sits: Meta automated the buying, so the winning operator leads with strategy + a relentless creative system + signal/measurement rigor + incrementality-grade honesty. You now have the operator-level playbook to build and lead that, not just the theory. The button-clicking is commoditized; everything this course concentrated on is the moat.
You've now mastered the demand side — the buying system this whole course built. Its companion, the Creatives course, builds the supply side that feeds it: 20 days from the Great Inversion (Day 1) through the creative genome, persona-led creative and AI production, to a compounding creative engine (Day 20). Great creative feeding a well-run buying system is the whole machine — start the companion course next.
Confusing checking with touching. New operators turn the 10-minute daily check into daily edits - pausing yesterday's "loser", nudging budgets, swapping creative - and keep the account in permanent learning reset. The second trap: skipping the weekly block because "nothing looks broken", then making panic changes off one bad Tuesday. The routine only works if the daily default is hands-off and the weekly slot is sacred.
Finale recap — your operating manual
- Governing principle: check often, touch rarely - deliberate single changes only.
- Daily (10-15m): pacing, anomalies, disapprovals, signal - default to observe.
- Weekly (60-90m): trends, creative cut/queue, gradual scaling, new tests, reconciliation.
- Monthly/quarterly: incrementality + new-customer growth, structure audit, creative direction, allocation.
- The whole game: own creative + signal + judgment; let the machine buy; optimize for real incremental growth.
That's the course. You've gone from zero to the full operator playbook - the hands-on media-buying layer that sits beneath every good strategy; fluency comes from running it. Lock it in with the final exam, then go build.
Three quick questions to lock in this module. Tap an answer to see if it lands.