(This article originally appeared on LinkedIn Pulse and can be found here.)
A reality of business is accepting that an unexpected change in customer preference has the potential to kill your brand overnight. As a result, Customer Experience (CX) has proved to be a vital part of an organization’s success in the areas of purchase intent and retention. While theories differ among CX professionals, the data shows that personalization is the key to a great customer experience. However, personalization isn’t easy to accomplish, as it requires you to map out every touchpoint in your customer’s journey and to identify customer personas. This post will detail how to differentiate your CX strategy through better data sources, different key performance metrics, and a data-driven approach to provide the experience each customer expects.
Improving CX Data Quality & Variety
Customer Experience professionals rely heavily on the use of surveys to collect data. This requires a variety of customers to talk to. It’s easy to talk to your own customers, but to reach a larger sample size you also need to venture to your competitor’s customers, as well as potential customers. This can be rather time-consuming for CX teams, so hiring panel companies has been the typical method used to conduct surveys. This reliance on panels introduces serious data quality concerns. These survey takers are motivated by incentives and are often used repeatedly, increasing the risk of involving professional survey takers that don’t represent the average customer. A large percent of your responses will more than likely be counterfeit. While panel companies aren’t the only way to find survey takers, all approaches to this process come with trade-offs.
Whether you use panel companies or social media and other digital advertising, you need to test data quality. An effective strategy to clean your responses is to utilize the information provided by the IP address of your survey taker. Most survey tools allow access to the IP address, which can be used to find a geolocation, then view public information associated with it such as real estate records and google maps. Additionally, a name or email address of the individual greatly speeds the validation process, as they can be used to find individuals online and verify their responses are genuine. This process involves multiple points of consideration before concluding that a survey response is invalid (for more details on this complex process, click here). Besides simply optimizing survey data, CX professionals should also be analyzing much more observational data. You can include data from your CRM about purchase process/patterns, marketing information your customer has seen and will continue seeing, and if you’re part of a SaaS organization include events data and app reviews. By combining these additional data types with your survey responses, you’re capable of fully understanding customer expectations.
Breaking Your NPS Dependency
Net Promoter Score (NPS) is by far the most frequently used metric on most CX teams because the most powerful marketing channel available is still word-of-mouth recommendations. While NPS is a good metric to track, too many organizations use it as a “silver bullet metric” when it’s actually relatively low-value, as it utilizes subjective, self-reported data. One of the innovative alternatives is Customer Lifetime Value (CLV), a high-value metric that will help you navigate through key business decisions. In addition to CLV, customer satisfaction is an often overlooked metric. Customer satisfaction was originally the driving metric in the CX world before NPS stole the limelight. However, in recent years, customer satisfaction has been experiencing a resurgence as a primary performance metric. This revival is mainly being pushed by Product Managers wanting to prioritize their product roadmap based on what makes the highest value customers happy.
Thinking Like A CX Data Scientist
The reason CX data is often neglected by executives is the way it’s presented. A common complaint being that the information presented is usually uninsightful because it’s too simplistic. Beyond simply reporting average values, your CX team needs to do further analysis so you can deliver insights pulled from the shape and spread of your data. The most dynamic way to differentiate your CX strategy with data-science is to use multiple variables concurrently, starting with tests of relationship/difference and finishing with predictive models. In practice, this requires going beyond just t-tests and crosstabs to clustering methods and predictive models that are constantly subject to real-world validation. This requires hiring a wrangler type data scientist (skip to 17:30 in this video to learn what a “Wrangler” is) and is a necessity to guarantee the results from your work are valid.
As a final reminder, to ensure that your reporting process is effective you need to present your findings in a clear, visual fashion. Even if you’re performing the deepest analysis possible using perfectly clean data, all your work will be totally useless if you can’t communicate insights properly. Begin implementing simple frameworks. Since you’re already in the process of mapping the customer journey, this is an easy place to start. SWOT is another useful framework, as it requires you to take all your data and layout in simple terms the most efficient steps available to your team to create a personalized experience to retain your customers as well as motivating them to recommend you to their family, friends, and colleagues.
Want to discover more about what CLV provides? Read our post on how to get Better Insights Through CLV.