How Dragonfly AI helped Mitsubishi Motors exceed revenue target by 17 percent

The brief 

With the pandemic supercharging digital acceleration across industries, Katie Dulake, General Manager, Marketing & Communications at Mitsubishi Motors UK, understood the brand’s marketing strategy needed, more than ever, to be digital-first. And with above-the-line campaigns put on hold, its digital channels had to be as productive as possible. 

Like so many brands at the time, Mitsubishi Motors UK was in the process of transforming its marketing function, transitioning to an increasingly efficient, results-focused department. Everyone in the auto sector and beyond was looking at how to make better decisions, and keen to explore new ways to measure impact. 

Together with their media company and creative agency – MG OMD and Golley Slater respectively – they planned a new programmatic campaign that would generate sales and help them achieve their targets for the year. Like all the brand’s marketing at that time, it would be a sales and offer based campaign that focused on acquisition. 

Like for many businesses affected by the pandemic, budgets were tight, and everything was now being measured by bottom-line conversions. The goal was to explore ways to do more, better, with the marketing department increasingly critical of where they were focusing their attention. They decided to embark a programmatic campaign which hinged off one consistent offer but, to avoid creative burnout, they wanted to experiment with colours, layouts and messaging to optimise their designs. 

As this was new territory for the team, most of the initial decision-making was based on intuition and gut feeling rather than solid data. It was only once they became stuck on a creative decision that Dulake looked to a solution that could provide data-informed insight. 

The solution 

Deciding that this was a perfect time to embrace experimentation, Mitsubishi chose to exploit cutting-edge neuroscience AI to accurately and instantly predict what grabs audiences attention first when they look at any kind of content. By using Dragonfly AI, the team quickly established the optimal colors and layout to help the messaging stand out and capture the audience’s attention. 

Using the Dragonfly AI Studio eliminated subjectivity from the equation using predictive visual analysis and predictive heat maps. The tool became a solid platform for research, learning, and communication between Mitsubishi and their agency. It gave room for standardiation and synergy, leading to a system where creative could be put through the platform, graded against key metrics and validated, where previously analysis would have been completely subjective.

The tool also enabled Mitsubishi to predict and prevent mistakes that otherwise would not have been noticed until much later, after the creative went live. By using this kind of AI Mitsubishi were able not only to save money but also the time that would’ve been spent on less successful ad testing. The brand utilised the tool across a number of campaigns during this period. 

Impacts of this approach on the wider marketing process included:

1. Increasing efficiency by enabling the team to create visually-ready ads that could be optimized based on performance data post-launch. 

2. Leveraging predictive modelling to determine which routes would obviously fail and which ones would work well in the real world. This allowed for more time to be spent polishing the assets and creating the best content to reach the audience.

3. Challenging and changing the marketing culture – Mitsubishi as a brand became braver and more experimental compared to its biggest competitors. By adopting the fail-fast mentality, they executed more tests and experiments than never before.

4. Accomplishing experimentation on a vast scale that would not otherwise have been possible. 

The results 

1. With the knowledge gained from Dragonfly AI, Mitsubishi increased CTR (Click-Through-Rate) by 10 percent.

2. Combined with increased templating they saved over 87 percent of the creative costs compared to the traditional marketing scenario.

3. In the overall campaign, they accomplished 3x higher ROI than average automotive, exceeded revenue targets by 17 percent and saw a 25 percent uplift in CVR between Q4 and Q1.

Katie Dulake, General Manager, Marketing & Communications, said: “With the knowledge gained from Dragonfly AI, we exceeded our campaign revenue target by 17 percent. Not only did Dragonfly AI contribute to our bottom line, but it was also a catalyst for digital experimentation across our department, enabling us to speed up our test runs and adopt a fail-learn-fast mentality. Utilising it helped us to move our marketing forward.”