By better understanding our customer (personas, typical website journey, critical purchase information etc) we were able to design and A/B test better-converting versions of the Blunt website. This was underpinned by immaculate tracking – telling us not only what customers cared about on their way to purchase, but which exact ad (or number of ads) maximised advertising budgets.
We began to treat different customer segments differently (older/younger, richer/poorer, city/urban etc, technical/emotional buyers) to personalise the buying experience and maximise conversion.
Finally, our weather based ads automation triggered ads to serve during bad weather (rain and wind) drastically increasing the likelihood of someone visiting the website to actually make a purchase.
We built a predictive model and an interactive dashboard that gives recommendations on which countries, cities and postcodes provide the highest likelihood for successful entry. This meant less wasted advertising, a cheaper entry into lucrative new markets and the ability to automate acquisition marketing in the future.
Our team used 18 major variables and their interactions to predict the likelihood of someone being “willing and able” to purchase a Blunt Umbrella. This machine taught model uses indicators such as a real-time “rain-wind index” and “average resident walking distance” to recommend the best spread of advertising budgets across each country.