MIAMI, FL: Brands could benefit from including factors like customer lifetime value in their marketing-mix models, rather than focusing "maniacally" on sales, according to a leading executive from AT&T.

Greg Pharo, the company's Director/Market Research & Analysis, discussed this topic at the Association of National Advertisers' (ANA) 2015 Masters of Measurement Conference.

Most marketing-mix models, he reported, have a "maniacal" focus on transactions – an area that is clearly vital to consider, but not to the extent that other factors are then neglected.

"While transactions are important," said Pharo, "there's also a lot of value you can unlock by optimising on relationships, rather than just isolated transactions." (For more, including further tips for improving marketing-mix models, read Warc's exclusive report: How AT&T enhances marketing-mix modelling.)

In better understanding the "why" alongside the "what" of each purchase, Pharo proposed that researchers double down on customer lifetime value.

The important matters to address here, he suggested, are "customer relationships, profitability over time, [insights] into the future and, very importantly, the impact on – or the profitability for – various segments of your customer base and potential customer base."

Successfully analysing and incorporating such issues into marketing-mix models is complex, as they cover a wide range of revenue and cost drivers.

"That's a lot of work," he said, "but it pays benefits, because you will get more accurate optimisation – [and] better, more insightful results – if you're able to optimise based on these customer relationships, rather than just on transactions."

Such recommendations for updating marketing-mix models came as part of Pharo's wider assertion that these vital tools remain "underdeveloped" in many cases.

"They haven't reached their full potential. As such, they can produce inferior results – especially if a mixed model is built solely from the blinders of media only," he said.

"It's important to expand mixed models [beyond] the basic media and advertising elements that you probably include in [your current] mixed models to incorporate the rich details that link up the rest of the business analytics."

Data sourced from Warc