Think you can’t predict what will happen between now and the year’s end? Media+ data sage Mike Ruff thinks otherwise, and shared his perspective with MediaPost’s Marketing Insider. Read the complete article here or below.
By Mike Ruff
This holiday season is a key time for you and your competitors during a year when everything seems more expensive than it was last year. With shifts in consumer spending, it’s more important than ever to optimize your ad spend to maximize your return. One way to do that is to partner with your analytics team about how you can leverage data to edge out the competition through predictive analytics.
What Is Predictive Analytics?
Predictive analytics uses historical data, combined with statistical analyses and machine learning, to predict future outcomes. It sounds complex, but you likely use predictive analytics in your everyday life.
Forecasting the weather combines historical weather data with current observations, radar, and other instruments. The output tells us the likelihood of temperature, precipitation, wind, and more. It has become so common that we don’t even consider the complexity behind the cartoon weather map, which relies heavily on predictive analytics.
For example, there may be a 70% chance of rain tomorrow, so we’d forecast for rain. Also, we may be able to combine all of the data to determine that the high temperature tomorrow will be 70 F +/- 4 F.
How Do We Interpret Predictive Analytics?
The 70% chance of rain tells us that it’s more likely to rain than not. However, there is still a 30% chance that it won’t. Predictive analytics is not a crystal ball, but in terms of the weather, it can help us be more comfortable, decide whether to buy tickets to a baseball game, and can even save lives.
When we interpret the output of predictive analyses, we combine these forecasts with other analyses to help us make better holistic decisions based on the totality of circumstances, including predicted outcomes.
What Predictive Analyses Should We Look at For The Holidays?
While there are endless opportunities for predictive analytics, here my top three holiday analyses:
Demand forecasting – Holiday marketing budgets can be all over the map, from way too high to way too low. By combining your historical first-party data with industry forecasts, you can set much better sales expectations and ensure your marketing budget is a good fit.
Product mix – Knowing what products sell with what other products is a powerful upselling opportunity that can be automated. This is especially effective when selling typically high-margin attachments. Your first-party product sales data will be the key data source. Once you uncover attachment forecasts for this holiday season, make it next year’s goal to test new attachments to improve the model.
Dynamic pricing – If you sell a deep inventory product for $100 at 50% margin, wouldn’t you want to know if someone would buy it for $99? Sure, you lose the $1, but your margin still may be great and you might not want to turn them away over $1. Dynamic pricing allows you to use your audience data to target and retarget prospects with changing pricing, thus increasing your holiday sales.
How Can I Maximize Holiday Success?
Since the holiday season is short and profitable, it’s important to get started early. Once insights are uncovered, use October and early November to test your ideas at a lower-risk time. Most importantly, you’ll have to fail fast during the holidays. Once you see that an audience or channel isn’t working, redirect that budget to more profitable and effective ones. Making a change mid-campaign is a sign that your team is being a good steward of your resources.
[Read the full list of insights at MediaPost here.]
