Suppose you are driving your car on a dark and curvy mountain road, on a rainy night, with a very foggy front windshield, dim headlights, and a perfectly clear rear view mirror to guide your path forward? It’s not a comforting scenario.
Surprisingly, this is how many companies still operate when attempting to understand and improve their customers’ experiences and the resulting customer satisfaction (or JD Power/Net Promoter) ratings.
They know a great deal about the past but very little about the future; they know a lot about the average customer but very little about the individual customer.
Even worse, with status-quo research, there’s danger around every corner but very little in the way of early warnings or accident avoidance strategies. As a result, relying entirely on traditional customer satisfaction tracking research is an extremely expensive way to chart your past performance, while leaving you with insufficient intelligence to proactively and systematically improve your customers’ experience, future market share, brand equity, and customer satisfaction ratings.
At a minimum, the problems with traditional customer satisfaction studies are threefold:
- They are not forward looking
- They rely almost exclusively on survey/opinion data
- They only measure the average satisfaction of selected segments of customers
The untraditional approach: traditional tracking research, big data, predictive analytics
Here’s the good news: when big data and predictive analytics are added to traditional customer satisfaction research, you suddenly have the power to proactively deliver a better customer experience, and in turn, drive better satisfaction ratings, customer retention, revenue, and other performance metrics.
Here are 5 of the many innovative approaches to integrate into your customer experience and satisfaction strategies:
1. Append big data to research results: Start by identifying additional data that may help you better understand and predict future satisfaction results. Some sources include third party data (Acxiom, for example), web log data, usage and transaction data, call center data, and even unstructured data from social media (e.g. using text mining).
Append this data to your existing research data. Track this data over time and build a variety of time series data sets and derived variables for modeling (see below).
2. Understand customer journeys (or paths) to purchase and other drivers: Traditional satisfaction research tends to focus on discrete touchpoints or drivers. While useful, this approach fails to identify the most common journeys or multi-touchpoint and multi-channel paths that your customers follow. For example, one very common journey might be to attempt a purchase, look for customer support online, call your customer support center, and comment on their experience (good or bad) via social media (Tweets, Facebook posts, blogs, or star ratings).
These and other journeys should be mapped and understood as part of your satisfaction research. They are often highly predictive of customer satisfaction and customer retention.
3. Leverage predictive analytics: Predictive models can tell you where your satisfaction ratings are going, why, and by how much (overall and for each unique customer).
So when JD Power tells you where you have been, your model will tell you where you are going, which gives you time to spot the key drivers and get to work on making real improvements. Or when Net Promoter Score (NPS) research gives you promoters vs. detractors, your model will tell you where your NPS is going in the future and exactly what is driving the promoters and detractors (and what you must do to earn more of the former).
4. Simulate success to prioritize improvements: When prioritizing potential satisfaction improvements, a robust simulator is often your best friend. While complex to design and build, simulators are extremely valuable when determining which improvement initiatives deserve the highest priority and the biggest investment.
The right simulator will be based on your predictive model and will help you rank initiatives based on cost of implementation and likelihood of success by segment/individual customer. This approach gives you objective and statistically-proven metrics to pick from competing alternatives—and that’s far superior to deciding based on political power and rhetoric.
5. Make global and customer-specific improvements: Traditional customer satisfaction research seeks to identify global (or macro) drivers and solutions. For example, if a false perception exists about product quality/reliability, and this misperception is driving down satisfaction ratings, the company will focus on communicating the truth about its strong product quality/reliability (and rightfully so).
The addition of big data and analytics unlocks the potential to communicate with customers on a one-to-one basis, using their unique drivers of satisfaction (or barriers to the same). When appropriate, companies can continually score their entire customer base and communicate in very targeted and powerful ways to improve their satisfaction.
Clearly, there’s much to be gained from using big data and predictive analytics: discover more drivers, identify meaningful journeys, predict future results, objectively rank alternatives, and make macro and micro improvements.
Together, these enhancements can lead to significant gains in customer satisfaction results over time and with a better ROI. (See my related post "The New Insights Imperative: Market Research + Big Data + Predictive Analytics" for more on this topic.)
While successfully integrating these 5 approaches isn’t easy, the results are almost always worth the effort. The next time you find yourself on that winding and treacherous road, you’ll appreciate your crystal clear windshield and your super bright halogen headlights leading you safely to your destination.
Now, that’s a better scenario.
Take Action Today >>
At Data Decisions Group, we are leaders in the integration of consumer data, behaviorial data, and market research data. Contact us today to discuss your market research and data needs.
Click below now to request more information or a conference call with our experts.