The Best Is Yet to Come — Predicting Customer Behavior With Data
What if you had a crystal ball that would predict customer behavior with uncanny accuracy and gave you the power to create the roadmap necessary to change the future of your business? Would you want one? Turns out that crystal ball is already available; it is called predictive marketing. Predictive marketing empowers digital marketers with the ability to make intelligent, forward-looking decisions through automation, visualization, and user experience innovations coupled with powerful predictive models.
You have always tried to predict the future of your business. Will your latest coupon campaign work? Will this year’s Black Friday be better or worse than last year’s? Looking at year-over-year trends is common for every business, but today, you have access to datasets that could only be dreamed of in the past. This is the new world of data-driven marketing. It is about being proactive not reactive.
**A Staggering Amount of Data Is Available
The amount of data available for predicting customer behavior is unprecedented in the history of commerce. Adult users in the US are spending an incredible 5.6 hours per day with digital media, and over 50 percent of that is on mobile devices. With all that data — and the technology to crunch the numbers — marketers are able to know how their customers are likely to behave, sometimes before they know themselves!
A great example comes from PC manufacturer, Lenovo. Using their analytics software and a predictive model, they are able to predict the likelihood that a visitor to their website will buy something. Then Lenovo’s targeting software delivers customized content that will best fit the customer’s needs. They say it has been accurate nearly 90 percent of the time!
**Predictive Analytics Is About Tomorrow
Using data-driven marketing techniques and predictive analytics is not about finding out whether your customer had a good experience last year. It is about deciding what your customer’s experience SHOULD be like today and tomorrow. It is an ongoing, iterative process that is not about the one-time offer. It is about getting your customers appropriate, well-timed offers and messages at every touchpoint and on every channel, whether on their social media feeds, their mobile browsers, or on the tile floor of their favorite retail stores.
**How Do You Begin Your Data-Driven Marketing Program?
As with anything new, you start small and work your way up. Here are some ideas:
1. Study One Factor — Using basic spreadsheet software, study historic trends in your business to forecast expected revenue tomorrow, next week, or next year, which is useful for setting budgets and goals. Data scientists call this kind of analysis “univariate time series” because you look at only one variable over time, ignoring how other factors might come into play. For example, you might look at the timing of offers you have made and how well they have done.
2. Study Two Factors — Begin using what is called “correlation analysis” to predict customer behavior, and start gaining control over future revenue. Correlation analysis looks at two trends or factors to see how they relate and whether one might be able to predict the other. You can use ordinary spreadsheet software. For example, you might add holidays and the school-year calendar to your analysis in step one. Then, you may notice a correlation between the start of spring break and how successful your offer was. You see the opportunity to make timing decisions regarding your offers that take into account a greater awareness of the customer’s needs.
3. Study Three or More Factors — Known as “multivariate regression,” some of this can be done with spreadsheets, but at this stage, most companies turn to specialized data-driven marketing software. Most spreadsheet software has limitations; if your software lets you have a million rows, it will not be enough if you have 10 million customers. But here, you can start to see the power this analysis can bring to the table. Using our example above, what if you added household income, number of children, and children’s ages to the analysis? You can see how you could more accurately target your ideal customer and properly allocate precious marketing resources.
4. Leverage Real-Time Data — Imagine using multivariate analysis based on data collected in real time, predicting customers’ behaviors instantly, and delivering the appropriate content at the moment they need to see it. This is the most advanced level of analysis, and it only scratches the surface of what is possible.
Here is a great example: An appliance company wants to increase sales of its new high-end dishwasher with an email campaign. Usually, the marketing department would send an email advertising the new dishwasher to all customers, but most customers either would not even open the email because they did not need a new dishwasher, or they would delete it because the appliance was too expensive for them. The results are an expensive email campaign that went nowhere and a warehouse full of unsold dishwashers.
Instead, the marketing department could use predictive analytics to discover that customers are most likely to be interested in new, higher-end dishwashers when their current dishwashers are between seven and nine years old, they have annual household incomes of over $100,000, they own homes of more than 3,000 square feet, they have granite countertops in their kitchens, and they often have teenage kids living at home.
Then, they could design the ad so that potential customers can picture the dishwasher in their own kitchens, perhaps showing a home with granite countertops and a family with teens. The next step is to send emails to the people they identified as most likely to want the new appliance. This kind of campaign results not only in dramatically increased sales, but also customers feeling that the brand understands their needs because they received an offer that was relevant and timely.
When the creative power of your writers, designers, and strategists come together with the mathematical power of data-driven marketing, you can unleash the ability to understand who your key audiences are and create the experiences they want. It is all about making smarter decisions and exceeding your customers’ expectations.
There are signs that data-driven marketing is becoming more widespread, and soon, successful companies will be unable to compete without it. Like anything new, start small and work your way up. In the end, predictive analytics and data-driven marketing techniques are about time-honored business practices — crafting the best customer experience, being of service to those who seek you out, and giving honest value to those you serve.
Want more information about predicting customer behavior? Look no further.