by Eric Bradlow
As a marketing professional, it’s in your best interest to take a strategic approach towards identifying, prioritizing and over-serving your highest value customers, as well as those who will likely have high future value, and new prospects. By using predictive analytics, you can gain meaningful, undiscovered insights into your target customers. And you can design a product experience that deeply resonates with your customers at just the right moment.
Indeed, predictive analytics can be applied across the entire customer journey – from awareness, to consideration, to purchase – and on through post-product satisfaction and ultimately referral. The most effective brands today truly fixate on customer experience and take an outside-in approach to understanding their customers’ underlying motivations and needs. And they do it by tapping into “good” data.
What exactly is “good” data? Predictive analytics is most powerful when your organization supplements internal transactional and behavioral data with survey and/or other external data. This allows your brand to better understand and serve existing customers, while also improving your approach to attract new customers. Too many brands invest materially in data and marketing science teams that don’t produce intended value, either due to lacking the right data, or not guiding the analytics by well-scoped business objectives or questions. This can render your efforts unproductive while concurrently draining your budget.
Also keep in mind, even sophisticated Machine Learning models will not help drive business value if you have lousy data. So to help ensure your data is valuable, run experiments that allow you to randomize customers into different “treatments” and assess which product features they prefer, or which email campaign they find most impactful.
The key to success is to ask the right business questions up front. Don’t just dive into your data, as you can easily find correlational patterns that you treat as causal, and will lead to spurious findings and poor decisions. And once you take a wrong turn, it can be an expensive and time-intensive ordeal to backtrack. Supplementing your internal data with external data, like surveys or 3rd party data, can help to ensure that the analysis accounts for a broader aperture of what is happening beyond your brand’s boundaries.
Using predictive analytics:
Now that you know what to avoid, let’s discuss benefits. Predictive analytics can be used in a variety of ways to optimize the experience for new or existing customers. These include:
- Leveraging analytics to recommend products to people that would strongly consider purchasing, based on their needs, buying behavior and preferences.
- Predicting purchase intent by matching product features and messages to customer needs, inferred from transactional data and survey responses.
- Accurately anticipating when a customer will leave the brand based on observed behaviors or interactions. In this way, predictive analytics improves the customer journey across the full spectrum, from beginning to end.
Predictive analytics becomes particularly effective when you begin to infer customer needs. This is accomplished by utilizing all of the internal data you have (including behavioral, transactional, demographic, and more), supplemented with external data like surveys. When you combine this assortment of data with sophisticated and modern Machine Learning methods, and validate the predictive model out-of-sample to avoid overfitting, then you have a business intelligence and optimization tool that can be used to deliver and optimize customer value. And you can also use predictive analytics to provide services and features to consumers, which they can enjoy in parallel with using the product itself. Thus you can convert customers into product and even brand advocates.
The five keys to good analytics:
Once you decide to implement a formal predictive analytics strategy, the following 5 step process can help on your journey to success:
- Start with a well-scoped overview of the business problem and customer need you intend to solve for, and understand how it maps to your overall strategy.
- Collect meaningful data by asking the right questions (i.e. What data is truly needed to improve CX and CLV, or Customer Lifetime Value, over time)?
- Analyze the data.
- Optimize and experiment to improve customer value and results.
- Share regular updates on your progress, analytics and insights to inform business decisions at your company and convince others to operationalize against it.
At the end of the day, brands that follow these steps will see successful applications of customer analytics.