Zohar Bronfman is the CEO and co-founder of Pecan.ai , a predictive analytics platform designed to solve business problems .
Have you prioritized the future of your business? Maybe it's time to look to the past for solutions.
History buffs know that the past provides valuable information for solving the problems of the present.
In today's marketing world, we are witnessing a revival of a once remarkable innovation, a practical solution.
That solution is Marketing Mix Modeling, or MMM. MMM is a powerful tool for understanding marketing performance and planning marketing strategies.
And given the big changes in marketing and data that have become such a hot topic recently, MMM is surprisingly old. Originally developed as models used to predict how political campaigns would affect voting patterns, the earliest forms of MMM date back to the late 1960s. This version of MMM runs on computers programmed with punch cards.
Do you know another characteristic of history, how it repeats itself? This is also true here. Today, MMM is experiencing a renaissance in the context of artificial intelligence and machine learning. Looking back on this story, there is much to learn about successful marketing today.
Marketing with the electronic brain
Harvard University historian Jill Lepore wrote about simultics in If/Then. How the Simulmatics Corporation invented the future . Simulmatics was founded in 1959 and originally sought to influence American politics through data.
The company is what The New York Times called "an electronic brain designed to evaluate voters' responses to campaign questions to make strategic recommendations." The idea was for computers to simulate when and where candidates should be declared to win an election.
But over time, simulmatics realized that this work was not limited to political campaigns. Advertising agencies and marketers also struggled to decide where to focus their resources to achieve their business goals.
But, as Lepore describes it, simulationists realized that "...you can't build a model without data, and these media companies have surprisingly little information about who is reading the books and magazines they print or listening to the news." : Articles and films. or produced (television... was an exception).
Lack of appropriate distribution channel information. Struggling to understand and plan for consumer behavior?
Are sellers feeling this all-too-familiar vibe?
The first attempts to develop and improve MMM arose from the situation we are in today. While we have now replaced some of these channels with mobile, influencer marketing and outdoor advertising, the challenge remains the same. There is not always enough information to determine the most effective use of marketing resources.
Calculations and information are separated
Finally, simulmatics, marketed as "media-mix," simulate an advertising campaign to predict its effectiveness. Advertising agencies that collect consumer data can use modeling techniques to more accurately predict campaign performance.
But in the 1960s, gathering data to build models was more difficult. The Simulmatics team even tasked them with generating data from children's TV shows.
The slow and laborious process of data collection was one of the challenges faced by the MMM pioneers. Another problem was processing capacity. Since computers still relied on magnetic tape, punch cards, and vacuum tubes, nothing was faster than building and maintaining MMM models.
Even the biggest early adopters of simulmatics, such as Nestlé, Colgate-Palmolive, and General Foods, probably rarely refined their models. So while customer behavior certainly changes in the months between new models, they still have to base their plans on old data.
Automated MMM generation based on machine learning
Fortunately, marketers no longer need magnetic tape to count and hire their families to help them gather information.
Today, automated data feeds make it easy to collect data from multiple marketing channels and data sources. These include CRM systems, customer data platforms, advertising platforms and web analytics. Data can also be aggregated from relevant external sources, such as industry data or weather data.
MMM can then assess the potential impact on revenue or other business results of each of these channels, as well as channels for which information is not available.
In addition, powerful cloud computing has made the modeling process more flexible, faster and easier to update. Change can be made in minutes, not weeks. Instead of waiting months for models to build, marketers can update models every week to make decisions based on the latest data.
Machine learning and more advanced computing have made these models more accurate. By identifying gaps between spend and targeted results, these reliable models can guide the optimization of marketing budgets across all channels, even when data is limited or scarce.
Another new feature of today's fastest and most reliable MMMs is the ability to quickly model different budget scenarios and their potential impact on business results. Modeling tools provide effective suggestions on how to allocate costs given a variety of objectives and constraints.
Finally, as in the days of Simulmatics, some consumers are concerned about the privacy and security of their data and its use. When using MMM, only transition layer data is required, not individual layer data. Thus, MMM is a marketing measurement tool that respects privacy principles and considerations while helping marketers make accurate data-driven decisions using the most up-to-date data and models.
With the addition of machine learning capabilities, MMM is not only back, but at the forefront of the marketing revolution.
Move into the future with MMM
Another advantage of modern MMM is that it no longer requires advanced statistical skills or specialized knowledge. Marketing organizations can implement MMM in a number of ways, including small-code predictive analytics platforms, open-source packages for use by experienced data scientists or external consultants.
Being prepared for a truly data-driven approach to marketing strategy and planning is just as important as technical preparation. Our Pecan research found that more than half of marketers still believe that many decisions are based on “guesswork” despite the availability of data and technology.
But if your organization is ready to move away from intuition to a more confident and forward-looking marketing strategy, modeling new and improved marketing mixes can be helpful for your team. no punch cards required.
The Forbes Technology Council is an invitation-only community for CIOs, CTOs and CEOs of world-class technology companies. What do I qualify for?
