Signal Automation
Weather- and time-based ads automation for Google & Meta
Signal Automation is HeyMetrix's rule engine that pauses, activates, or re-budgets Google Ads and Meta Ads campaigns based on weather forecasts and time of day. Every rule can be backtested against historical weather data and rolls back automatically when the trigger condition ends.
Demand for many products moves with the weather: ice cream and cold drinks with heat, heating service with cold snaps, delivery with rain, travel with sunshine. Signal Automation lets your campaigns react to those signals automatically, without Google Ads scripts or a developer.
- Google Ads + Meta Ads
- Forecasts up to 7 days ahead
- Backtesting on historical weather
- Automatic rollback
- Full audit trail
Last updated:
How does weather-based ads automation work in HeyMetrix?
A Signal Automation scenario combines a condition, an action, and a rollback strategy. The condition watches real-world signals (weather forecasts or time of day) for a region you pick on a map. The action changes something in your ad account when the condition fires. The rollback strategy defines how everything returns to normal afterwards.
An example scenario: “If the temperature in Berlin exceeds 30 °C tomorrow, raise the ice-cream campaign budget to €50/day, and roll it back when the heat wave ends.” Setting this up takes four steps:
- 1Pick the region on a map, or let HeyMetrix auto-detect it from your campaign's geo targeting.
- 2Define the condition: weather metrics (for now, or 1–7 days ahead) and/or time of day, combinable with AND/OR logic.
- 3Choose the action: pause or activate campaigns, ad groups, ad sets, or ads, or set a campaign's daily budget to a specific amount.
- 4Set the rollback: manual, time-based, or automatic reversal when the condition flips back.
Which weather signals can trigger a rule?
Weather conditions come from Open-Meteo forecast data and can look at the current weather or up to 7 days ahead. A rule can therefore act on tomorrow's heat wave today, before the demand spike rather than after it. Time-of-day conditions can be combined with any weather condition using AND/OR logic.
Available conditions
| Signal | What it covers |
|---|---|
| Temperature | Threshold above/below, now or forecast 1–7 days ahead |
| Precipitation probability | Chance of rain/snow for the selected day |
| Wind | Wind speed thresholds |
| Humidity | Relative humidity thresholds |
| Cloud cover | Share of cloud cover |
| Weather condition | Categorical states like clear, rain, or snow |
| Time of day | Time windows, combinable with all weather signals |
What can a scenario actually do in the ad account?
Actions execute directly on Google Ads and Meta Ads through the official APIs, the same connections HeyMetrix uses for monitoring. No scripts, no spreadsheets, no zaps.
- Pause or activate campaigns, ad groups (Google) or ad sets (Meta), and individual ads.
- Set a campaign's daily budget to an absolute amount: on Google at campaign level, on Meta for CBO campaigns at campaign level and for ABO campaigns at ad-set level.
- Every platform mutation is recorded with its pre- and post-state in the execution audit trail, so you can always trace exactly what a rule changed and when.
What happens when the weather changes back?
Automation without a way back is a liability. A rule that raises a budget for a heat wave must not leave it raised in the rain. Signal Automation ships three rollback strategies, chosen per scenario:
Rollback strategies
| Strategy | How it reverts |
|---|---|
| Condition reversal | Automatically reverts when the trigger condition flips back: budget raised for a heat wave returns to normal when the heat wave ends. |
| Time-based | Automatically reverts after a set number of minutes. |
| Manual | Stays until you revert it yourself, for changes you want to review first. |
Can you test a rule before it touches live campaigns?
Yes, twice over. Test mode lets you “test fire” a scenario as a dry run: HeyMetrix evaluates the condition and shows what would happen, without changing anything in the ad account.
Backtesting goes further: HeyMetrix replays your rule against historical weather data for the selected region and shows how often it would have fired in the past. A rule that would have triggered 40 times last month probably needs a stricter threshold, and you find that out before it spends a cent. The AI agent explains backtest results in plain language.
How does this compare to Google Ads weather scripts?
The traditional way to run weather-triggered ads is a Google Ads script wired to a weather API. That is flexible, but it needs someone who writes and maintains JavaScript, covers only Google, and offers no built-in safety net. The honest comparison:
| Google Ads weather script | HeyMetrix Signal Automation | |
|---|---|---|
| Setup | JavaScript + weather API key, maintained by a developer | Visual editor, no code |
| Platforms | Google Ads only | Google Ads and Meta Ads |
| Weather data | Depends on the API you wire up | Open-Meteo forecasts, now to 7 days ahead |
| Rollback | You script it yourself | Manual, time-based, or condition reversal |
| Testing | Preview runs, no historical validation | Dry-run test mode + backtesting on historical weather |
| Audit trail | Script logs, if you build them | Pre/post state per platform mutation, scenario history |
What are typical use cases for weather-triggered ads?
Any product whose demand correlates with weather or time of day is a candidate for weather targeting. Patterns we see agencies build:
- Food & beverage: push ice cream, cold drinks, or beer-garden campaigns when heat is forecast; delivery campaigns when rain is forecast.
- HVAC & trades: activate heating-repair campaigns on cold snaps, air-conditioning campaigns on heat waves.
- Retail & garden: garden centers on sunny weekends, umbrella and rainwear promotions ahead of rain fronts.
- Travel & leisure: outdoor activities, pools, and excursion targets on clear days; indoor alternatives (cinemas, museums) on rainy days.
- Time-of-day patterns: lunch offers only before noon, nightlife campaigns only in the evening, combined with weather where useful.
How does the AI agent help with automation?
You don't have to invent scenarios from scratch. The Monitoring Agent knows your project's campaigns and their geo targeting, suggests complete scenarios (“this campaign advertises beverages in Munich, want a heat rule?”), and explains backtest results. You review and approve; the agent never activates a scenario on its own.
Honest scope
Signal Automation actions currently cover pausing/activating entities and setting daily budgets. There are deliberately no automated bid-strategy or creative changes. Weather conditions are forecast-based (Open-Meteo) with a maximum look-ahead of 7 days. HeyMetrix is in early access; final pricing is not set yet.
Frequently asked questions about weather-based advertising
Which ad platforms support weather-based automation in HeyMetrix?
Google Ads and Meta Ads (Facebook/Instagram). Rules can pause or activate campaigns, ad groups/ad sets, and ads, or set campaign daily budgets, on Meta at campaign level for CBO and at ad-set level for ABO campaigns.
Do I need Google Ads scripts or a developer for weather-triggered ads?
No. HeyMetrix connects via OAuth and executes actions through the official platform APIs. Scenarios are configured in a visual editor: no JavaScript, no API keys, no spreadsheet plumbing.
What weather data does HeyMetrix use?
Forecast data from Open-Meteo: temperature, precipitation probability, wind, humidity, cloud cover, and categorical weather conditions, for the current moment or 1–7 days ahead, per region you select on a map.
What happens if the weather changes back?
That's what rollback strategies are for. With condition reversal, the rule automatically reverts its own change as soon as the trigger condition no longer holds. Alternatively, revert after a fixed time or manually.
Can I see what a rule would have done before activating it?
Yes. Test mode dry-runs the scenario without touching the ad account, and backtesting replays the rule against historical weather data for your region, showing how often it would have fired.
Does weather-based advertising actually improve performance?
It works where demand genuinely correlates with weather: seasonal food and drink, HVAC, travel, garden, delivery. The backtest tells you how often your specific rule would have fired, so you can judge the impact for your case with data instead of guessing.
Put your campaigns on autopilot for the weather
Get early access and let the agent suggest your first automation scenario. Testing is free.
Get Early Access