The science of audiences

Man riding backwards on a horse...teach the world to sing...the Pepsi challenge...Blendtec...Apple lemmings superbowl commercial...Excedrin headaches...there is no denying that when you catch lightning in a bottle with great creative, glorious things happen.

However, I want to use math and science to make great things in marketing recurring events, where repeatable success will come from understand the math of audiences and marketing, using the right targeting and tactic choices based on evidence turned into principles.

This is the part of marketing that is still evolving and meeting resistance from those who preach that the old ways of mass marketing still rule the roost. This despite my white papers, in collaboration with others that prove that certain targetable consumer segments will PREDICTABLY return 5-16 TIMES the ROAS of other consumers. #Persuadables #Outcomesbasedmarketing2.0.

In an era of addressable marketing, where even TV advertising is shifting from pre-formed linear audiences towards addressable marketing, it's time to augment MASS thinking with MATH thinking.

In addition to my normal newsletter that I e-mail, I am starting this new newsletter on LinkedIn to share what I have learned and what we are yet to learn about the science of repeatable success in marketing. Marketers started this journey around 1990 with marketing mix modeling...how to systematically shift funds from low to high performing channels and sub-channels. But now, we must go further...we must create a science that is foundationally centered on consumers. Who to target, what are the need states that make consumers more receptive, what is the right creative for performance vs. brand-building...and more.

In the past 1-2 years, using the math of of logit functions, monte carlo simulations, beta distributions, and calculus, the #MMA (who I have consulted with for 6 years) and #Neustar created and proved that a particular targeting strategy could out-perform reach based media planning by 50%. Working with #Truthset, using Bayesian math, we created a system for identifying IDs that were most likely to possess asserted attributes, improving the quality of 3rd party audiences. Working with #DISQO, we uncovered which patterns of digital behaviors that had the highest conditonal probabilities of leading to an add-to-cart event.

All of this work had these things in common...

--rooted in math

--generalizable truths

--offered approaches for improving marketing effectiveness that were not parochial to a campaign but should work each and every time.

My hope is that this monthly newsletter becomes a forum for exchanging the best in scientific thinking about how to make marketing work better.

Thank you for joining this community!

Regards,

Joel