Customer experience is becoming an increasingly popular topic, with many forward-thinking businesses thinking it can set them apart in today’s highly competitive landscape.
And they’re right — a recent study by Forrester showed that organizations with a cross-team CX approach were nearly twice as likely to have exceeded their top 2017 business goals. Compared to their competitors, experience-driven businesses grew their revenue 1.4 times faster and had 1.6 times the customer lifetime value of other companies.
Think back to a time that you had a great experience with a brand. Was it the variety in products that made the experience great? How about the price, or the speed of delivery? Did you interact with the brand through their website, through their social media, through email, or through multiple channels? Chances are you remember some, but not all, of the details. But, every detail counted and added up to the summative experience that you remember.
So with customer experience being more valuable than ever, and a customer’s perception of their experience being based on a number of different variables throughout their journey, you might think that you should be optimizing all of those different variables. But if you’re like most businesses, you’re going to be at least somewhat resource-constrained. You’re going to have to make choices on where to allocate your efforts based on which optimizations will matter the most.
Fortunately, there’s a method to Customer Experience Management. To figure out which optimizations will matter the most, start by asking your customers. Through surveys, you generate valuable data about what customers think about your company. Then, analyze the data to extract insights, make decisions, and get to work on the improvements that will elevate your customer experience and increase the success of your business.
Survey for Net Promoter Score (NPS)
One of the most popular measures of customer experience is the Net Promoter Score (NPS), which is a measure of loyalty to your brand and an effective method to determine customer likelihood to repurchase, refer friends and family, and resist market pressure to deflect to a competitor. NPS is found to correlate with revenue growth and NPS leaders tend to grow at twice the rate of their competitors.
With NPS, you’re asking one simple question: “How likely are you to recommend us to your friends and family on a scale of 0-10?”
Based on their answers, customers can be sorted into three categories: promoters, passives, and detractors. Promoters give a 9 or 10 rating and are highly likely to recommend you. Passives give a 7 or 8 rating and are generally satisfied but may be vulnerable to the competition. Detractors give a rating of 6 or below and are not likely to recommend you.
Your overall NPS score is the percentage of promoter responses minus the percentage of detractor responses. Scores can range from -100, where everyone surveyed is a detractor, to +100, where everyone surveyed is a promoter. Generally speaking, a positive score greater than 0 is considered to be good, and a score above 50 is considered to be excellent, but there is some variation depending on benchmarking per particular industry.
Not all customers are the same, nor are they necessarily using the same service or product. For that reason, break NPS results into segments, for example purchasers vs. non-purchasers, to uncover more specific insights and take more meaningful action.
Composite Scorecard Index
Due to its simplicity, NPS is great for providing a broad high-level overview of customer experience to top management, but a multi-dimensional weighted Composite Scorecard Index can be more reliable and less volatile. Plus, customer experiences aren’t composed of isolated experiences. There’s interplay between the attributes, and this makes traditional regression analysis far less predictable.
Ask your customers about their experience across a set of attributes. Customize these attributes and their weightings to fit your specific needs and combine them to produce one score.
In our experience, combining multiple attributes into a composite delight score can perform significantly better in predicting customer recommendations and retention. Plus, the very specific information it provides can be highly actionable. It doesn’t even require dramatic action — even improving on small things can often lead to large improvements in customer experience.
Surveying customers produces data and understandings about what’s important to customers and how they perceive your product and service. A path analysis can help you go further, extracting deeper insights through statistical regression.
Path analysis uses structural equation modeling techniques to analyze several variables at once and provide estimates of the magnitude and significance of connections between a dependent variable and one or more independent variables. Using it, you can find how independent variables directly and indirectly affect a dependent variable, for example how speed impacts NPS.
To start a path analysis, you first need a hypothesis about the relationship between your variables. Then, you can run a statistical analysis and an output path diagram, which illustrates the relationship between variables as determined by the analysis.
It’s important to note that while path analysis is useful for evaluating your causal hypotheses between variables, this method can’t prove the direction of the causation, only the correlation and strength of the relationship.
However, when you discover the degree to which different variables have an impact, you can be even more efficient in your efforts to improve the customer experience.
The influences you discover in path analysis can also be used to forecast future outcomes. Extrapolating from existing data, you can model future scenarios and choose the best current course of action in order to optimize your outcome.
For example, knowing the importance of various attributes of your experience, you can run a “What If” scenario to predict the potential contribution that various improvements would make to an end goal like raising NPS and elevating customer experience. For example, what if you increased choice by 30%? What impact would that have?
In this way, you prioritize your efforts for the biggest gain and increase the predictability around the impact of decisions. In the past, the only way to understand how changes would affect profitability would be to execute them and then wait and see. Now, advanced data models can take the guesswork out of ROI predictions.
Besides predicting measures like NPS, these methods can be used to forecast customer retention or traditional financial metrics like average revenue per customer. They can also be used to anticipate customer needs and determine when they’ll need attention, allowing you to proactively reach out and minimize churn.
87% of customers want to be proactively contacted in matters related to customer service, and 73% who are proactively contacted and had a positive experience felt a positive change in perception towards the brand. Predictive engagement is worth the effort.
Taking it Further
All of this work can be done by an analyst, but a tool like CloudCherry makes it exponentially faster and easier to develop weighted scorecards, run analyses, and create predictive forecasts. Putting these results in the hands of practitioners allows businesses to make strategic decisions faster and more effectively.
That said, in order to use data to its highest potential, you need human input. Getting the most from your data requires having an intuition and awareness about who your customers are and what’s important to them. It requires asking the right questions, making creative hypothesis, and understanding what you’re seeking from the data.
And that requires comfortability with and knowledge of data practices, honed through research and directed learning. Download our e-book, The Data Science of CEM: Analytics for Customer Experience Management, for more information about working with customer experience data and to equip yourself to make better data-driven decisions.Download The Data Science of CEM