In today's business landscape, there is a growing focus on improving customer experience (CX), with companies increasingly turning to CX data, such as Voice of the Customer (VoC) surveys and online reviews, to guide their strategies. However, many businesses struggle to interpret and effectively use their CX data to drive tangible improvements through effective actions. This situation raises a critical question: How can businesses not only collect but also accurately analyse and effectively act upon their CX data? The answer lies in developing a deeper understanding of customer feedback, identifying key trends and underlying root causes of negative experiences, and quickly adapting corrective strategies to meet customer expectations.
How signals found in CX data can be transformed into actions
With the advent of AI using Large Language Models (LLM) and the ability to fine-tune these LLMs to fit specific purposes, we now have countless opportunities to explore vast amounts and varieties of open-ended CX data to mine value-creating insights.
In our work with clients, we often face the challenge of analysing CX data in a way that uniquely facilitates a deeper analysis of the underlying root causes of negative experiences. Furthermore, it is essential to measure the company's performance against broader industry standards for a meaningful comparison. The objective is to identify opportunities for improvement and recommend actions that address negative experiences.
Enter the Marketing Mix–Marketing Funnel Matrix
Many companies form their market plans and strategies by employing traditional marketing tools. These include the elements of the marketing mix (product, price, promotion, place, people, process, and physical evidence) and the stages of the marketing funnel (awareness, consideration, conversion, retention, and advocacy) incorporated within the context of the customer journey. While there is implied connectivity between the elements of the marketing mix and marketing funnel , each is usually treated independently when planning market strategies. Upon introducing CX data into the analysis process, the integrative complexity of planning significantly increases. This begs the question: what if all three perspectives could be combined into a unified framework? And, if so, would it make experience analysis simpler? A mix-funnel matrix (MFM) would reduce the complexity of CX analysis and bring actionable clarity. As the term matrix suggests, the MFM is a two dimension grid structure that graphically illustrates intersecting values. And through the use of AI, the MFM aggregates and classifies terms found in customer comments and appropriately fits those at the intersection of the various marketing mix elements and marketing funnel stages. Thereby forming actionable insights so companies can better learn and respond to opportunities for improving the customer experience and journey.
Customer Experience Meets AI
The analytic power of fine-tuned LLMs helps discern specific aspects and opinions contained in customer comments. For example, a customer comment about a Tavern/Pub located in Maryland read “I have been to the Pub many times, and the food is always excellent, especially the corned beef. The cook always makes everything to perfection, just love the food.” Another reads “Terrible service, they were out of two of my favorite craft beers, just because they are convenient to my apartment, there are better places to go”. The pretrained, fine-tuned LLM would extract aspects and populate the matrix as follows:
For example, for the ‘terrible service’ aspect and opinion terms, management would review the aggregation of similar data points filling the matrix cell to understand the range of experience signals related to service issues. The opinion term modifies the signal level; the LLM determines the sentiment strength of the term and assigns it a numeric value. If the sum aggregate value of instances indicates service to be in a critical state, then drilling deeper into the data could reveal the root causes of the signal. By analysing experience signals at the intersection of each mix-funnel dimension, management can determine what action-management responses to take to address the situation.
The Unified Framework
The example given was based on an individual Tavern/Pub, but the intent of the mix-funnel matrix is to provide various levels of aggregation, from brand to micro-industry, comparative for benchmarking purposes. As aspects and opinion terms are collected for multiple Tavern/Pubs, these are aggregated to illustrate experience levels among all Tavern/Pubs within a geographic area. Through sized icons, experience levels visually show the importance of any particular intersecting dimension. Looking at the intersect of ‘consideration’ and ‘place’, the large green icon may represent an excellent signal regarding ‘convenience’ being important to ‘selecting’ a Tavern/Pub, whereas the smaller dark red icon may be a critical signal related to ‘parking’ being a factor in choosing a Tavern or Pub. The following figure conveniently aggregates the data and displays it in an easy-to-understand matrix, clearly highlighting the key areas for action.
The Mix-Funnel Matrix (MFM) allows for a very quick identification of the intersections of marketing mix elements and funnel stages that have the predominant contribution and are prevalent drivers of the total negative customer experience.
To act, or not to act?
Finding nuances in touchpoint data is a complex exercise usually relegated to data analysts and scientists for interpretation. The mix-funnel matrix (MFM) reduces the complexity of CX analysis and brings actionable clarity to the data. This simplifies the decision-making process and the identification of effective corrective actions.
Value creation isn’t found in just the derivation of data; it is also found in the speed of analysis and the ability to quickly make actionable decisions. The MFM concept is found to be generalizable across all industries where CX data is available and collected. Achieving excellence in customer experience is possible through continuous improvement that focuses on tackling the root causes behind negative customer experiences, as identified by the MFM analysis.
The authors
Emil Tsankov, CEO & Cofounder Research Metrics. A global SaaS company that empowers organizations worldwide to achieve excellence in customer experience. Established in 2004 in Toledo, USA, the company has international offices in Varna, Bulgaria, and Mumbai, India.
Blaine Stout PhD, Consultant to Research Metrics. An independent consultant working with companies on operations and information management research, applications, and data analytics.
Comments