Rocket Fuel: Measuring the Effectiveness of Online Advertising Case Summary This case study digs deeply into the businesses TaskaBella and Rocket Fuel. Rocket Fuel’s goal was to use artificial intelligence and analytics to revolutionize the digital advertising industry. Rocket Fuel generated about half a billion dollars in sales a year between 2008 and 2016. To conduct a trial campaign for a recently released handbag model, TaskaBella reached out to Rocket Fuel. Following extensive discussions and calculations, a proposal was agreed upon by both businesses. Rocket Fuels job was to calculate the size of the control group automatically based on the campaign perimeters, and TaskaBella’s job was to understand the factors that determine the resulting size, as they also tried to avoid having a large control group. This trial ran for four months. The ads first were enabled to display anywhere within milliseconds, then the publisher connects with vendors to put the impression opportunity out to auction. In order for its bid to quality, Rocket Fuel needs to respond within 100 milliseconds after receiving the bid. Once this is done, it applied ML models to predict the probability on what the user will act on. Once the trial was over, the two companies were able to calculate that out of the 59,000 users identifies, around 15,000 bought the handbag. Rocket Fuel then was able to use these results for analytics to calculate if the trial was effective or not. After the two companies met in person to get the analytics of the trial and measure if the total cost in advertising was profitable compared to the value of the conversions, they discussed the effectiveness of the trial. They concluded that the advertising was effective because it not only increased profits, but also conversion rates. Overall, this was a very successful trial advertising campaign for both Rocket Fuel and Taskabella as well as giving both companies lots of data and research for future studies about conversion rates and profitability.
