Digital marketing success requires a unique mix of data scientists and creative marketers. Here’s how to help those teams do their best work.
The advent of digital tools has upended age-old processes in marketing and advertising. Digital marketing technology is now a requirement for identifying, attracting, and retaining customers in an omnichannel world.
A new e-book from the MIT Initiative on the Digital Economy highlights learnings from the 2022 MIT Chief Marketing Officer Summit held this spring. The topline message to marketing executives: Add data, analytics, and algorithms to better reach socially-linked modern consumers.
Here are MIT Sloan researchers’ top digital marketing trends for 2022:
ocial consumers in broad digital and social media networks
Today’s consumers make brand decisions based on a very broad set of digitally connected networks, from Facebook to WhatsApp, and the mix is constantly in flux.
Since social consumers are influenced by what social network peers think about different products and services (a trend called “social proof”), marketers must employ granular analysis to really understand the role of social media in marketing, according to IDE directorSinan Aral.
Aral examined 71 different products in 25 categories purchased by 30 million people on WeChat and found significantly positive effects from inserting social proof into an ad, although the effectiveness varied. For example, Heineken had a 271% increase in the click-through rate, while Disney’s interactions rose by 21%. There were no brands for which social proof reduced the effectiveness of the ads, Aral said.
Video analytics on TikTok, YouTube, and other social media
TikTok influencers loom large, especially with Gen Z. The problem is whether or not those viral influencer videos actually translate beyond attention into sales.
Research shows that engagement and product appearance isn’t the crucial factor — it’s more about whether the product is complementary or well-synched to the video ad. And the effect is more pronounced for “product purchases that tend to be more impulsive, hedonic, and lower-priced,” according to research conducted by Harvard Business School assistant professor Jeremy Yang while he was a PhD student at MIT.
Measuring consumer engagement with machine learning
Call it the “chip and dip” challenge: Marketers have long grappled with how to bundle goods, finding the right consumer products to combine for co-purchase from a huge assortment. With billions of options, this research is exacting and massive in scale, and data analysis can be daunting.
Researcher Madhav Kumar, a PhD candidate at MIT Sloan, developed a machine learning-based framework that churns through thousands of field scenarios to identify successful and less successful product pairs.
“The optimized bundling policy is expected to increase revenue by 35%,” he said.
Using machine learning to forecast outcomes
Most marketers are concerned about retention and revenue, but without good forecasts, decisions about effective marketing interventions can be arbitrary, saidDean Eckles, social and digital experimentation research group lead at IDE. Instead, update customer targeting through use of AI and machine learning to forecast outcomes more quickly and accurately.
In collaboration with the Boston Globe, IDE researchers took a statistical machine learning approach to analyze the results of a discount offer on customer behavior after the first 90 days. The short-term surrogate prediction was just as accurate as a prediction made after 18 months.
“There’s a lot of value to applying statistical machine learning to predict long-term and hard-to-measure outcomes,” Eckles said.
Adding “good friction” to reduce AI bias
Digital marketers talk frequently about reducing customer “friction” points by using AI and automation to ease the customer experience. But many marketers don’t understand bias is a very real factor with AI, said Renée Richardson Gosline, lead for the Human/AI Interface Research Group at IDE. Instead of getting swept up in “frictionless fever,” marketers must think about when and where friction can actually play a positive role.
“Use friction to interrupt the automatic and potentially uncritical use of algorithms,” Gosline said. “Using AI in a way that’s human-centered as opposed to exploitative will be a true strategic advantage” for marketing.
If you are interested in original article by Tracy Mayor you can find it here