Data science is not only a discipline for analytics, it is also extremely useful for search engine advertising. Just check out the following example how data science can support your PPC strategy. As we are trying to put a lot of data science knowledge in our latest Google shopping bid automation product Whoop!, we are exploring different features of any kind.
According to recent research studies, weather may have at least some impact on human decisions and behavior.
Too long, didn’t read version:
• The Whoop data science team found interesting weather patterns across customer data.
• Cloud cover correlates with conversion rate, the cloudier the sky, the more the decrease of the conversion rate.
• Device depending patterns showed increasing conversion rates during heat waves only on personal computers.
The variety of weather correlated research studies contains stock market on Wall Street¹ as well as college admission in the United States². Closer to our online shopping environment is weather effect on car purchases³, which reported about correlations between the share of convertibles on car sales and temperature on the day of purchase. While it was expectable that the share of sold convertibles increases with temperature, they also found out that the return rate for this cars increased significant. Thus, long term decisions are influenced by projection bias.
Time to take a look on these interesting weather features for the “Whoop!” data science team! This blog post focuses on cloud cover and very hot days during heat wave periods:
The Top-5 list of real weather data
This is our Top-5 list of real data weather insights this summer.
1. There is a correlation between conversion rate and cloud cover. The cloudier the sky, the more decreases the conversion rate.
2. From all analyzed brands across our customers, “Lego” was sold most on the sunniest days (smallest cloud cover average).
3. The highest conversion rates came from personal computers, followed by tablets and mobile devices. This pattern does not change during heat waves but the conversion rate for computers increased during hot periods.
4. In Germany, a whooping 36% of the analyzed 91 days were hot days over 30°C / 86°F. The next heat wave is coming soon!
5. This is why across our European customers, more than 85 % of cooling fans were sold on hot days with temperatures above 30° Celsius.
We implemented these insights already in Whoop!, stay tuned for more interesting facts from our data science department.
If you’re interested in more details and insights, you should join our free webinar tomorrow (August 26 2015) with Reinhard Einwagner Product Manager of Whoop! & Google Shopping Expert and our friend Frederick Vallaeys (CEO of optmyzr.com).
1) Undergraduate Economic Review Volume 7 | Issue 1 Article 19 / 2011 / Reassessment of the Weather Effect: Stock Prices and Wall Street Weather / Mitra Akhtari / University of California, Berkeley
2) Journal of Behavioral Decision Making / Clouds Make Nerds Look Good: Field Evidence of the Impact of Incidental Factors on Decision Making / Uri Simonsohn / TheWharton School-University of Pennsylvania, Pennsylvania
3) The Quarterly Journal of Economics (2015), 371–414 / The psychological Effect of Weather on Car Purchases / Meghan R. Busse et al