Although many marketers rely on simplistic or outdated attribution models, improvements are being made.
We forecast that 58.3% of US companies will use multichannel attribution for their digital marketing efforts in 2019, up from 48.6% in 2017.
Multichannel attribution assigns marketing credit to more than one marketing channel or touchpoint. It also provides more complete data on a customer’s path to purchase than does last-click attribution, which credits a purchase or conversion to the last ad clicked by a customer.
While companies can better prove the value of their marketing efforts with multichannel attribution, they have a long way to go in gaining a deeper understanding of their campaign metrics.
For many, bridging traditional media measurement methods—like marketing mix modeling with multichannel attribution—remains a challenge. Even fewer are capable of combining the two and accounting for critical nonmarketing touchpoints, such as call centers, sales teams and the in-store experience.
While there is no panacea for marketers’ attribution challenges, data scientists may be part of the solution. Marketing firms should view their data science hires as a complement to their traditional marketing analytics, not as a substitute, according to Jim Spaeth, partner at media consultancy Sequent Partners.
“Without data science, traditional marketing analysts can’t cope with big data sets,” Spaeth said. “Without traditional marketing analytics, data science overlooks decades of knowledge about consumer behavior and advertising dynamics, and then mistakes data for answers.”
Many marketers agree. In a Salesforce poll of 4,101 marketing leaders worldwide, 42% of respondents said they are using methods driven by data science to measure their marketing success. A similar percentage said they measure their marketing success with attribution modeling, which included marketing mix modeling and multitouch attribution.
"In my experience, successful brands invest a lot of time and resources in making sure that data science and marketing collaborate,” said Nancy Hua, CTO of A/B testing platform Apptimize. “Marketers want to make decisions quickly, while data scientists want to take the time to collect and analyze the results for the most accurate answer possible. At companies that really succeed in digital, there's a healthy tension there.”
“Building attribution solutions internally by hiring data scientists will be possible only for companies that have had success building large-scale data management and science teams, and are able to invest a lot more in the attribution solution than what they would have to pay if they go with an external vendor,” said Brian Baumgart, CEO of marketing analytics company Conversion Logic.
But that’s not to suggest that firms aren’t making room for these high-value roles. Andreas Reiffen, CEO of retail tech firm Crealytics, noted that thanks to advancements in artificial intelligence (AI), some companies are choosing to automate campaign execution. From there, budget is being reallocated to data science.
“You used to need a lot of head count to run your campaigns,” Reiffen said. “With AI, this has started to shift, because you can fully automate lots of these long-running repetitive tasks. That leads us to situations today where we need highly skilled, numbers-driven people—data scientists—who are able to perform these tasks.”
For more on how companies are shifting their internal structures and even their organizational cultures to make way for more data-driven, holistic attribution efforts, stay tuned for eMarketer’s upcoming report series on attribution.
Not sure if your company subscribes? You can find out here.