Multi-Account Budget Reallocation — How Healthcare Groups Move Spend Mid-Month Without Breaking Learning
Why monthly budgets are a 1990s artefact
Monthly budgets exist because finance teams approve them monthly. They do not exist because Meta's algorithm needs them. Meta's optimiser would happily learn at any cadence — daily, weekly, monthly. The constraint is purely organisational. And the organisational constraint costs real money.
Consider a healthcare group running ₹85L/month across 6 Meta ad accounts: 2 IVF, 2 dermatology, 1 dental, 1 ophthalmology. Each account's audience saturates at different points in the month. The IVF account in Mumbai might saturate by day 18 (Hold Rate dropping, frequency climbing past 4.2). The Bengaluru dental account might still be in efficient learning at day 22, leaving 8 days of demand unfunded. Holding budgets flat means roughly ₹6-9L/month lands on diminishing returns while a different account starves.
The Tuesday Reallocation Ritual
Meta Catalyst IQ runs portfolio reallocation every Tuesday between 10am and 12pm IST. The choice of day is deliberate:
- Tuesday: Monday's data is stable (weekend noise has cleared, Monday's full day has reported)
- Mid-morning: Budget changes have ~12 hours before Wednesday's auction surge in healthcare
- Weekly cadence: 4 cycles per month gives 4 chances to course-correct without resetting learning
The engine ingests 7-day LSQ, CPL, frequency, Hold Rate, consult-show, and treatment-conversion data per ad set across all accounts. It produces a single ranked recommendation list: which ad sets get more, which get less, by exactly how much.
The 4-input scoring model
| Signal | Threshold for "scale" | Threshold for "starve" |
|---|---|---|
| LSQ score (7-day) | >70 | <40 |
| Frequency (7-day) | <3.0 | >4.5 |
| Hold Rate trend | Stable or rising | Dropping >10% week-over-week |
| Consult-show rate | >specialty median + 15% | <specialty median - 15% |
An ad set must hit at least 3 of 4 "scale" thresholds to receive additional budget. Likewise 3 of 4 "starve" thresholds to lose budget. The 2-of-4 zone is held flat — uncertainty does not earn shifts.
The 25%/20% guardrails
Catalyst IQ caps a single day's budget change at +25% up or -20% down per ad set. This is not a compliance rule. It is a Meta learning-phase preservation rule.
Meta re-enters the learning phase when an ad set's budget changes by more than ~30% in 24 hours. Re-learning kills 5-9 days of efficient delivery. So portfolio shifts are stair-stepped: a 50% reallocation happens across two Tuesdays, not one.
An anonymised multi-specialty group — Q1 2026 portfolio shifts
A north India multi-specialty group running 6 Meta ad accounts, ₹62L/month total, adopted the Tuesday Reallocation Ritual in January 2026. Across 13 weekly cycles (Jan-Mar):
| Week | Total shifted | From | To | Net effect |
|---|---|---|---|---|
| W2 | ₹4.8L | NCR IVF (saturating) | BLR Dental (under-budget) | CAC -11% |
| W5 | ₹3.2L | Mumbai Derm (frequency 4.7) | Hyderabad IVF (LSQ 78) | CAC -8% |
| W8 | ₹6.1L | BLR Dental (now saturated) | Chennai Ophthalmology (new launch) | CAC -16% |
| W11 | ₹2.9L | Hyderabad IVF (Hold dropping) | NCR IVF (re-emerged audience) | CAC -6% |
Over the quarter, portfolio CAC dropped 19.4% with no new creative, no audience expansion, and no headline budget increase. Just spend landing where consults were happening.
The 6 behaviours that break this engine
- Capping account-level monthly budget rigidly. Finance teams must agree to a 15-20% inter-account flexibility band each month.
- Reallocating reactively on Friday CPL spikes. Friday spikes are usually weekly cycle artefacts. Wait for Tuesday's stable data.
- Letting clinic owners veto cross-account shifts. Each clinic owner protects their account, which kills portfolio optimisation. Centralise the decision.
- Shifting more than 25% in a day. Triggers Meta re-learning; net negative.
- Ignoring frequency in the scoring. An LSQ-70 ad set at frequency 5.2 is about to collapse — funding it more accelerates the collapse.
- Not stair-stepping back when an ad set recovers. A starved ad set needs slow re-funding, not a sudden 2x bump.
The governance layer founders need
Every Tuesday reallocation produces a 1-page audit trail: which accounts moved, by how much, signal evidence, projected impact. The founder sees it as a Wednesday morning email. No surprises. The CFO sees a quarterly summary showing portfolio CAC trajectory. The agency or in-house ops lead is on the hook for following the engine's recommendation or documenting why they overrode it.
Monthly budget rigidity is a tax on multi-clinic healthcare groups. The clinic chain that moves 8-15% of its portfolio weekly between accounts based on LSQ, frequency, and consult-show data captures the audience-saturation timing variance that monthly-budget peers leave on the table. The discipline costs 2 hours a week and saves 14-22% in CAC.
Want the Tuesday Reallocation Ritual on your group?
We'll plumb the signals, run the first 4 cycles with you, then hand it to your in-house ops lead with a documented playbook.
Book a free audit → WhatsApp ICG