"Weather-based dayparting" sounds clever. It usually means someone wants the ads to flinch every time a rain icon appears. That is a bad clock.
adset_schedule, and it does not auto-spend on sunshine.Why rain-at-3pm is a bad schedule
- Forecasts are regional. Your ad set is often national.
- Showers last 20 minutes. Learning lasts days.
- Fashion demand moves on warm weekends, not on a single wet hour.
- You will never keep the sheet in sync. You will get superstition.
If weather matters, it matters as a card: trigger, eligible SKUs, cover kill, owner. Not as 168 new cells.
When weather and hours might honestly meet
Same-day, local, outdoor, stock in the van. A heat spike by 14:00, you can deliver today, the set is already built. A human turns emphasis up. That is not dayparting. That is ops.
Everyone else: leave the weekly clock stable. Use the weather playbook for sets and creative. UK shoulder-season noise: weather and UK demand.
Fashion example
Warm Saturday forecast in your top ship regions, two days out. Lightweight sets have cover. Do not rewrite Friday's 18:00-22:00 cell. Do swap the prospecting set toward the warm story, and kill it if cover drops. Evening sofa hours stay whatever your Meta breakdown already justified. Fashion dayparting.
What Ralph will not do
No weather-linked hour writer. No unsupervised weather budget. The weather panel is evidence or silence. Daypart suggestions come from hourly ad history, not from Open-Meteo. If those two signals both look loud on a Monday brief, you still approve two different moves. Why Ralph for dayparting.
A side-by-side you can steal
| Weekly daypart | Weather playbook | |
|---|---|---|
| Input | Your hourly CPA / ROAS | 2-day forecast in ship regions + cover |
| Cadence | Change rarely (weeks) | Change when the card fires (days) |
| Action | Ad scheduling / bid adj | Sets, creative, merch modules |
| Kill | Volume collapse, CPA worse | Cover under buffer, forecast flips |
| Ralph | Heatmap + approved Meta clock | Panel / finding, no auto-budget |
If you cannot fill both columns without overlapping actions, you are about to build a rain dance. Stop.
Keep weather and the clock in different drawers
- Write the weekly clock first. From hourly ad data only. Overnight OFF or not. No forecast in this step.
- Write the weather card separately. Trigger, eligible SKUs, cover kill, owner. Copy the weather-triggered playbook.
- Do not add rain to Ad scheduling. If you need a weather move, change sets or creative, not Tuesday's 15:00 cell.
- Same-day exceptions stay human. A local heat spike with stock in the van is a phone call, not an autopilot.
Questions people actually ask
What is weather-based dayparting?
People mean: change the hour schedule when the forecast changes. For most fashion and DTC, that is the wrong tool. Weather moves which SKUs deserve budget over a day or a weekend. It does not mean pause all ads at 15:12 because a shower started.
Should I pause Meta ads when it rains?
Only if rain kills your category in that region and you can prove it. Outdoor furniture maybe. A hoodie brand almost never. Use a weather playbook on sets and creative, not an hourly stop rule.
Is this the same as weather-triggered ads?
No. Weather-triggered ads shift emphasis (sets, creative, merch) when a forecast clears a threshold and stock can take it. Dayparting is the weekly clock. Separate pages, separate jobs.
Does Ralph daypart from the weather API?
No. Weather is demand context and a panel. It does not write Meta adset_schedule. It does not auto-blow budgets.
When would weather and the clock meet?
Rare same-day outdoor or event retail: a heat spike this afternoon, stock in the van, local delivery. That is a human call, not a 168-cell model.
UK weather is noisy. Now what?
Shoulder seasons lie. Use two-day forecast thresholds in top ship regions, plus cover. The UK demand essay is next door.
If the hours are the problem: Ralph reads sixty days of your hourly results, proposes the switch-off with floors, and applies it on Meta after you approve. On Google he recommends the schedule. See Ralph · How Ralph dayparts · Docs