Machine learning may be a buzzword, but Norwegian Air’s marketing head and the CEO of ad targeting platform AdTheorent point to tangible, if initial, results.
As the competition among long-haul, low-cost carriers intensifies, Norwegian Air and its long-time ad tech partner wanted to start 2019 with the intent to accomplish three primary goals. The first was to drive bookings to Norwegian’s branded website.
The second campaign priority was to get traction for Norwegian’s brand identity as “the world’s best long-haul, low cost airline.”
And the third charge: harness data that can be used to build follow-up campaigns and improve the airline’s marketing overall.
Among the initial results the two companies were willing to share delivering including noting that “cost per booking” (aka “cost per action in general online ad parlance) that was 170 percent lower than the CPA goal (no dollar figures were provided).
As for the future insights, the machine learning-based marketing effort found a difference between travelers who went to Norwegian’s booking site over wi-fi at home and at work. The prospective fliers who accessed Norwegian’s site at home tended to book flights at a rate 4X greater than users at work
“Airlines must use smart advertising tactics to engage customers who are the most likely to book flights,” said Marina Suberlyak, Head of Marketing North America, Norwegian. “AdTheorent’s machine learning and predictive targeting capabilities enabled us to spend our limited advertising dollars more efficiently by rapidly identifying opportunities to optimize where and how our brand was positioned to the most qualified buyers.”
Kambr Media caught up with Jim Lawson, CEO of AdTheorent, and Marina Suberlyak, Head of Marketing North America, Norwegian, to get a few more details about the campaign.
Kambr Media: How did Norwegian come to choose AdTheorent for this campaign?
Marina Suberlyak: We have worked with AdTheorent for several years and they consistently move the needle for our brand. AdTheorent leverages learnings from past campaigns to continue to drive down the CPA for each campaign.
In terms of the two main goals of the collaboration with Norwegian and Vizeum, can you describe AdTheorent’s role and strategy in attempting to drive bookings and raising Norwegian’s profile as the “world’s best long-haul, low cost airline?”
Jim Lawson: Norwegian engaged AdTheorent for this campaign because we have a proven track record of using targeted digital advertising to drive actual business goals – in this case, flight bookings. Our data science team developed custom machine learning models to target users within target DMAs with high predictive scores, identifying those most likely to engage with the ad and complete a booking.
How did it differ from past campaigns? Or was this simply an effort to amplify past efforts in terms of driving bookings and solidifying Norwegian’s profile as the “world’s best long-haul, low cost airline?”
MS: In addition to past learnings, this campaign utilized insights from current efforts, using real-time feedback, to optimize targeting and placement and therefore improve our ability to drive bookings at lower cost.
What were the main challenges in both areas that you sought to address?
JL: When using advertising to drive a very specific business action, targeting must be precise. For the Norwegian campaign, the main challenge was identifying efficiently the users most likely to take the desired action of booking a flight.
How does Norwegian view the competitive challenges at the moment this campaign kicked off? Was it largely combating seasonal trends associated with slower winter airline bookings?
MS: As most in our industry, we do see similar seasonality effects; however, that doesn’t equally translate into advertising results. In fact, since AdTheorent’s technique used machine learning, more data would naturally drive better results over time, so we actually saw CPA improvements as we moved through the year and into Winter.
With this campaign, were you focused on particular areas/regions or demographics?
MS: This campaign was focused on key markets including Austin, Boston, Chicago, Denver, Florida, Los Angeles, New York and San Francisco/Oakland. AdTheorent developed custom machine learning models to target users within the target DMAs who were deemed most likely to engage with the ad and then complete a booking.
Is there anything you can say about the initial performance of the effort?
JL: Campaign performance exceeded Norwegian’s goals, delivering a cost per flight booking that was 170 percent lower than the “cost per action” (CPA) goal. AdTheorent has partnered with Norwegian for several years and we are able to leverage data and learnings from past campaigns to deliver bookings with increasing efficiency.
How are machine learning-based marketing tools impacting ad strategies for considered purchases like airline bookings and brand affinity?
JL: Machine learning and “Artificial Intelligence” buzzwords are everywhere these days . . . At AdTheorent, we use the industry’s best machine learning platform and we provide “data science as a service” as part of every campaign, but we realize that this is all noise unless we drive actual business results for our clients.
Measurable performance and solutions-oriented executions drive value and customer retention. Most digital ad solutions focus on clicks or engagement metrics which don’t always translate into direct business outcomes. Norwegian wanted to move beyond basic measurement tactics to demonstrate real-world value – and that is our specialty.
How do you expect this campaign to influence the next marketing and ad targeting efforts? Is the focus likely to continue to drill down on these two goals of bookings and raising awareness of Norwegian’s long-haul, low-cost carrier profile?
MS: Our commercial strategy remains largely the same – to drive bookings and increase awareness of our proposition. We are obviously very pleased with the results AdTheorent delivered through its advanced approach, and we are looking forward to continuing the partnership and extracting further value.