The same ads as the control,shown only when the sun was out,converted 76.17% better in the UK.
A global maker of premium garden and patio equipment, selling in 19 markets under its own brand. The products are built to last and carry a multi-year warranty. Most of the selling year falls between March and July. The idea tested here has probably crossed your mind: grills sell on warm days, so show the ads on warm days. Somebody put it against a control group and published the result, including the market where it cost more.
- +76.17%
- UK conversion rate, weather-triggered against always-on
- -39.32%
- UK cost per acquisition against the same control
- +28.37%
- Germany conversion rate, at a 90% confidence level
- 17°C
- the temperature a day had to reach before an ad could run
The forecast decided the media plan, and the forecast kept changing.
A trigger that would not sit still
Ads could switch on only under a clear sky and above a set temperature. In April that threshold was 15°C. As spring warmed up it had to be raised regularly to keep separating a good day from an ordinary one, and by July it sat higher still. The setting needed constant monitoring and quick changes from the team running it.
Short windows that broke the learning
A campaign that goes dark every time the clouds come in never runs long enough for its bidding to settle. Test windows were short, so launches slipped or stopped midway, and the irregular delivery made it harder for the algorithm to learn. Click-through rate came out lower than in the always-on campaigns, which the study puts down, with a "probably", to that stop-start pattern.
Two markets with opposite reasons to test
The UK was picked because its weather is changeable and hard to predict, which makes it a good place to test and a hard place to target. Its test window then brought very few sunny days. Germany is the brand's most important sales market, so it is the market where a bad test would hurt most. The two tests ran on different dates with different setups.
Hold the creative fixed. Let the weather be the only variable.
A live weather feed wired into DV360
The OpenWeather API reports cloud cover and temperature, along with other readings, every 2 to 4 hours. That feed was connected to Display & Video 360 so campaigns launched and paused on their own as conditions changed. Nobody had to decide each morning whether today counted.
Identical creative on both sides of the test
The weather-triggered campaigns used the same creative as the control group, and in the UK the same three best-converting audience groups from the standard campaign. The one thing allowed to differ was the moment an ad could appear. Change the creative as well and any lift could belong to the new ad instead of the sun.
Regions small enough to share a forecast
The UK was divided into four regions: Scotland, England, Northern Ireland and Wales, with targeting narrowed to selected locations. Germany was divided across its 16 states, and the test ran in Bavaria, Baden-Württemberg, North Rhine-Westphalia and Berlin, each with a single best-converting audience group. A national trigger would fire in one state while it rained in the next.
Run head to head against the always-on campaign
Each market ran the weather-triggered version beside a standard campaign that ignored conditions, compared on click-through and conversion rate. The UK test ran from April 22 to May 17, with ads live under clear or lightly clouded skies at 17°C or above. Germany ran from May 13 to May 31 and added no precipitation to the rule. Across both, the minimum temperature kept rising as the weather warmed.
Both markets converted better. Only the UK got cheaper.
Conversion rate of the weather-triggered campaigns against the always-on control in each market, as the study reports it. One bar per country, both measured the same way.
The taller bar is the weaker claim. The German figure carries a stated 90% confidence level, and the UK figure, the one in the source headline, carries none. Cost went in opposite directions, so it stays off this chart. In the UK, cost per acquisition fell 39.32% against the standard campaign. In Germany the standard campaign bought conversions more cheaply than the weather-triggered one, and the study gives no figure for the gap. A fall and an unmeasured rise cannot share one scale, which is why the UK figure sits in the strip above and Germany has no number at all. The study also reports a lower click-through rate beside the higher conversion rate, so fewer people clicked and more of those who did went on to convert. Its summary claims a lower cost per conversion. That holds for the UK and fails for Germany on the study's own figures.
A hunch about your buyers is a test waiting for a control group.
Put a control group behind the hunch before you spend on it. Every seasonal business carries a theory about its buyers, and grills selling on warm days is about as obvious as those theories get. The result here came from running the same creative to the same audiences with and without the trigger, so the weather was the only thing left to explain the gap. Without that control, a sunny fortnight and a good month look identical in your reporting.
Read your results market by market before you roll anything out. The same idea converted 76.17% better in the UK and 28.37% better in Germany, and in Germany it cost more per acquisition than the standard campaign. The study reads this as a sign the approach pays best where the climate is changeable, as in the UK. If your good days are predictable, your always-on campaigns may already be catching them.
These were short tests over a few spring weeks, and the UK run had very few sunny days. Your category and your climate will move the result in either direction, and the 76.17% comes with no confidence level attached. Treat it as a reason to run your own test, and design the control group before the first ad goes live.
This work was delivered by MTA Group, which Zero Fluff Digital is part of, by the same specialists who would work on your account. Every figure here comes from the case study MTA published, republished here with their consent. The reviews behind it are public and verified on Clutch, where clients rate the work rather than the agency describing itself.
Your sales follow the weather. Does your media plan?
30 minutes, no deck. You work out whether a trigger like this could pay on your account, and the control group you would need to prove it.
Book a callNext case studyMarpur: 62.55% revenue growth on a smaller ad budget →