Unforeseen obstacles can topple any marketing or advertising campaign. Accordingly, marketers should use predictive analytics to forecast their future through data and identify any critical gaps that need filling.
In 1978, structural engineer William LeMessurier received an alarming message from a college student. LeMessurier had recently completed the new Citicorp tower in Manhattan, a stylish skyscraper with its bottom corners jutting out over space. The effect was riveting, as if the structure floated in air, each corner 114 feet above ground.
An engineering student, asked by her professor to study the design, had called LeMessurier’s office to suggest his Citicorp building might be vulnerable. LeMessurier chuckled at the student’s naiveté, but later re-ran some calculations and in horror realized the student was right. If hurricane-force winds ever hit the building from a certain angle, the pressure might topple the structure. He immediately launched a rebuilding program to strengthen the tower’s frame.
Call it a gap in predictive analytics.
If you’re a marketing executive, you face the same challenge. How do you design a successful advertising campaign when unexpected headwinds may knock over everything? Beyond the elements within your control—pricing, brand, creative, media—you know unseen forces may emerge. PR scandals. A revolutionary competitor. Privacy data breaches. An industry price war.
With so much chaos in the world, you need predictive analytics.
A Simple Game Board for Seeing Your Future
Media planning, buying and analytics agency Mediassociates has guided clients through hundreds of campaigns and seen almost every uncertainty arise. While there is no crystal ball to forecast the future, it has designed a simple Predictive Analytics Game Board strategy to help marketers with data-based forecasting. This is not a technology platform or AI system, but a logic structure to compile what appears to be too much data into clarity on one’s marketing future.
Marketers are often caught between too much microscopic data (usually digital or TV analytics awash in response rates, conversions and attribution modeling puzzles) and too-weighty statistical methods (when the term “regression analysis” comes up, executives’ eyes glaze over). You have an ad server, CRM, DSP, DMP, customer segmentation, industry trends, etc. Because there is so much data, with digital often floating at the top, marketers make the common mistake of looking at what happened yesterday without making any good predictions for the future. Campaign optimization is great, but it’s a bit like driving your car only by looking in the rear-view mirror.
To see far ahead, you need to expand the scope of the data you measure inside and outside your organization, and model it at just the right level to make predictions on likely outcomes. Mediassociates suggests you do this by structuring a “game board” of six measurement boxes. Using this logic structure will help you sort the data you have, and identify gaps in your marketing information systems that prevent you from seeing the future.
The Predictive Analytics Game Board
Let’s go over each box. For each, ask a question: Is your marketing team covering this square?
Forecast
This is most obvious, in which you predict how marketing activities of “A” will drive sales of “B” based on historical performance. Typically, marketers model off past “funnel” benchmarks, breaking down each conversion stage to anticipate how customers flow into the sales system.
For example, if you spend $5 million in advertising this quarter, and anticipate a $100 cost per acquisition, you’ll make 50,000 sales. Great. But you may be forgetting the external “game” square that could help or hinder these results…
Game
No one runs marketing in a vacuum. In the “game” square, marketers must assess external environmental scenarios on what could happen in the future. If you’re a technology startup and Apple launches a similar gadget in November, your performance will be squeezed at Christmas. So is that likely?
Building out external scenarios requires “gaming” potential actions of your competitors, suppliers, market entrants, market substitutes, as well as environmental waves by industry, weather and the economy. Not every hypothetical can be modeled, but you can start by mapping the past few years of external vectors that pushed on your marketing — and then gaming similar scenarios for the near future.
Test
This is what you should do when your marketing is happening right now. Testing is a real-time refinement approach of offer, message and media to feed optimization. Note: Testing does not have to be expensive or risky. New digital services allow rapid prototyping of creative messages to learn consumer receptiveness for as little as a few thousand dollars, and social media advertising can be a forum for speed-testing creative or audience targets.
Monitor
This is how you look outward when your marketing is happening right now. In today’s economy, external environmental factors are shifting more rapidly than ever. Social media sentiment, competitor tracking, news monitoring and evaluation of how you perform in search versus competitors are all urgent requirements to make sure an asteroid-type-event is not striking your marketing campaign when you least expect it. Make sure your reporting systems monitor waves outside your internal marketing pond.
Measure and React
The last two squares are often poorly designed by marketers. A common error is to focus too much on digital data (since it is so easily obtainable). To test how well you can forecast, build a chart of your entire media mix that reaches your audience across their “customer journey” from awareness to acquisition to conversion to loss to win back—and then overlay what percent of each activity you really measure.
If you can capture all data and run it through multi-touch attribution systems: good. If you can’t, consider light statistical modeling to look for correlations between internal marketing actions, outside events and results. And in the “react” square, prepare rapid-response plans for the most likely, or most damaging, scenarios.
Yes, technology platforms can give you better data, analysis, insight and visibility. But before you focus on a single tech solution, map this entire “game board” ecosystem for predictive analytics and look for gaps. If you fill each of these six squares, you’ll have ironclad logic to take to the CEO to explain how you’re going to influence results. Because the only way to influence the future is to try to predict it.
A few years ago, I tracked down the student who called William LeMessurier in 1978 and invited her to a predictive analytics conference. Her name is Diane Hartley, and she was gracious, saying she made the call after running some math that didn’t seem to add up. “It wasn’t that I was so smart,” Hartley said. “I was just curious and wanted to explore all the angles.”
A great lesson for modeling your marketing data in predictive analytics.