Who the client is
The client is a global manufacturer of garden and patio equipment, operating in 19 markets, with over 1,000 products in the premium segment and a seasonal business peaking between March and July (the same client as the Home & Garden market-entry case).
The starting point
- Campaigns dependent on weather conditions. Activations could only take place in clear skies above a temperature threshold; initially, in April, 15°C, raised regularly as the weather warmed up, higher by July.
- Short testing periods and campaign instability. Limited testing windows often delayed or interrupted the campaign, making it harder for the algorithm to learn and optimize.
What was done
The idea came from the project team: since the product belongs to the home & garden category, sales should rise during warm, sunny weather. The client tested it in two key markets: the United Kingdom, where changeable weather creates ideal test conditions, and Germany, the brand's most important sales market.
- The campaign ran in DV360 on weather data collected in real time from the OpenWeather API (updates every 2–4 hours: temperature, cloud cover, probability of precipitation), which automatically launched or paused the campaign.
- Creatives were identical to those of the control group, so only the weather varied. CTR and conversion rate were the comparison metrics.
- United Kingdom: four regions (Scotland, England, Northern Ireland, Wales); ads when the sky was clear or lightly clouded and the temperature at least 17°C; April 22 to May 17; the three best-converting audience groups from the standard campaign. Very few sunny days during the test had a negative impact on results.
- Germany: campaigns across 16 regions (states), including Bavaria, Baden-Württemberg, North Rhine-Westphalia, and Berlin; clear sky, minimum 17°C, no precipitation; one best-converting audience group per region; May 13 to May 31.
The result
A/B tests compared the weather-based campaigns with standard always-on campaigns.
| Market | Conversion rate | Cost per acquisition (CPA) |
|---|---|---|
| United Kingdom | +76.17% vs the standard campaign | −39.32% vs the standard campaign (control £20.51, variant £12.45 per the source chart) |
| Germany | +28.37% vs the standard campaign, at a 90% confidence level | Slightly higher than the standard campaign |
The weather-based campaign had a lower CTR, probably due to its intermittent nature, but a significantly higher conversion rate than standard campaigns. The source concludes that weather-based activity can particularly improve effectiveness in countries with a variable and unpredictable climate, such as the United Kingdom.
What happened next
Despite the unfavorable weather in the United Kingdom, the tests delivered a lower cost per conversion and showed the client that the strategy was practical. Testing and refinement were planned for the next season.
"We have a weekly call to talk about what was accomplished the past seven days and what we're going to work on in the next seven. They're very responsive; communication is really great compared to the agency we worked with before. If there's a need for an ad hoc call, we use a chat channel or email."
