CX in the age of AI: A leading outsourcer’s perspective – Interview with Cathy Jooste and John Samuel of CGS
September 27, 2023
When we talk about attention, we’re actually talking about engagement – Interview with Walter Flaat of dentsu Canada
October 3, 2023
This is a guest post by Akin Arikan, the author of Customer Experience Analytics and Multichannel Metrics.
Adrian points out that “CX should not be reduced to a KPI”, as fans of Punk CX will know. “Data are only a proxy or measure, of how we are doing helping our customers have a good experience.”
At the same time, data can trump opinion for decision-making, even the opinion of superiors in an organization. And while measurement seems like something better suited for nerds, what could be more punk rock in the corporate world than overcoming a higher-up’s objections and getting the company to do the right thing for their customers?
But for that, we need to show data that creates both customer empathy and shows the impact on the business. Especially, on digital channels, where you cannot even see customers with your own eyes.
So, what data is a good proxy for customer experiences? Where is the line between measurement skills that help vs. those that produce CX programs that are “overly technical, benchmarked, frameworked, measured, codified, certified, specialised and functionalised etc etc.”, as Adrian warns.
Outcome-oriented metrics are critical but are not a good proxy for experience
Outcome-oriented metrics such as revenue, journey completion rates, or website conversion rates are often seen as the crown jewel of key performance indicators on dashboards. There is a good reason for that, yet, at the same time they are not a good proxy for customer experience.
For example, on a banking website, conversion rates measure how many customers start a journey, such as a mortgage application, vs. complete that journey. The assumption is that the metric measures both the bank’s success with driving applications and the customer’s ability to complete those applications successfully.
Yet, what if the application forms were a giant PITA to complete? And, what if the only reason that customers struggled through them was that the bank offered a better interest rate? But that doesn’t mean the customer will ever recommend the bank to others. The conversion rate then just measures how successfully the bank is “buying” customers by paying them with better interest rates.
Same, for example, on a retail website where shopping can be a terrible frustration sometimes. Some customers suffer through it anyway if the retailer has offered steep discounts. They will probably, however, never return.
So, what measurements are a better proxy for customer experience and where do outcome-oriented metrics come back into the picture?
CX Analytics skills to the rescue!
Knowing how to use the new breed of CX Analytics will help here. What are CX Analytics? They are a range of relatively newer measurements – especially for digital channels – that help CX teams step into the shoes of their customer. They include the following:
- Qualitative data, e.g. from website and app feedback
- Quantitative data, e.g. from so-called digital experience analytics, i.e. session replays, journey analytics, heatmaps, performance analytics, etc.
Unlike traditional scores and metrics, the immersive, visual nature of CX Analytics can help the team see experiences from the customer’s perspective, e.g.
- Session replays can recreate individual customer experiences while excluding any sensitive data from being captured. This can reveal the underlying reasons why customers are blocked and “rage clicking”, e.g. technical errors, slow speeds, or experiences that are simply confusing
- Journey analysis can show where customers are going in loops on the site or app, unable to find what they are looking for
- Frustration scoring can surface and rank-order the journeys that cause the most struggle
- Heatmaps can show why online forms or retail checkout are frustrating some customers, among other things.
Equivalents are emerging also for non-digital channels, e.g.
- Brick-and-mortar store foot-traffic visualizations can show how shoppers are flowing through store isles and how long they are queuing in front of registers.
- Contact center sentiment analysis based on AI can detect, categorize, and quantify customer frustrations
How do Customer Service and CX Teams combine qualitative and quantitative CX Analytics?
Teams regularly hear of things that are reported to be broken, whether it is through customer feedback, customer service complaints, or anecdotally from the boss’s spouse who tried something on the app or website that didn’t work.
But as explicit as customers can be in their feedback, they are rarely descriptive enough for teams to understand what actually happened, and what behaviors led up to the issues. Feedback comes in such as: “I guess you don’t want my money. This is the worst website ever. I hope you lose your job!”. Yet, recreating these frustrating experiences is often challenging.
Sometimes, weeks and months go by with advanced customer support teams unable to recreate issues even though additional customers keep complaining.
That’s because if the experience were broken for all customers, it would be easy to see. But typically it’s only broken for pockets of customers, e.g. on certain browsers, devices, or based on certain browsing paths. Without stepping into their shoes, it’s like tapping in the dark as to why it’s happening.
That’s where CX Analytics skills come in, for example, by replaying the website and app sessions where customers left feedback.
Why outcome-oriented metrics are still needed?
Sometimes, seeing the website session replay of a single customer who repeatedly tried to checkout or tried 14 times to complete their travel booking is enough for a team to drop their lunch sandwich and go fix the problem.
But for most companies, there is simply too much feedback to address, and millions of session replays to watch. The largest online businesses receive hundreds of thousands of feedbacks per year It is simply not feasible to act on every one.
That’s why quantifying the impact of issues on both experience and business outcomes is the way to prioritize which ones to fix first. CX Analytics skills help with that too, e.g. by knowing how to …
- go from a single session replay to quantifying how many other customers suffered the same bug or same confusion
- quantify the journeys that lead to the most feedback
- isolate the form fields that are most confusing and frustrating
- find the pages with slow speeds that get customers disgruntled
- and correlate all this to the loss of customers and revenue that ensues
Where is the rub with missing skills?
Too often, it is only analytical or technical personnel that is trained to use CX Analytics data. That doesn’t work anymore, because customer experience is too important to be the job of only analysts.
Creating better experiences also requires many teams to collaborate, e.g. Customer Service, CX/UX, Marketing, Ecommerce, Product, and Technical Teams. That’s why knowing how to use CX Analytics is a mission-critical skill for all those teams today.
Training and enabling these teams to get closer to their customer and to understand their needs is something that will help the Punk XL agenda.
Where do you start?
A “rip and replace” approach can be risky. It can introduce as many new issues as you are trying to fix. As Adrian has pointed out, it’s more about evolution vs. transformation. What are the redesigns, new experiences, or issue fixes that are most likely to move the needle, step by step?
- To see how your company compares to the benchmark in your use of CX Analytics across teams today, consider completing the CX Analytics skills self-assessment on my website and get personalized tips (totally free and no registration required.)
- When choosing a CX Analytics software or services provider, ask providers how they will enable each of the teams and roles that should benefit from hands-on access to data.
- When operationalizing the use of CX Analytics in your company, make a plan that can eventually grow to cover all roles in all teams, not just analysts and technical roles.
For real-life use cases on digital channels, cheat sheets, and examples from dozens of companies, check out my new book on customer experience analytics, published by Taylor & Francis group earlier this year.
This is a guest post by Akin Arikan, the author of Customer Experience Analytics and Multichannel Metrics.
About Akin
Akin is the author of two books
Say Hi to Akin on X (formerly known as Twitter) @AkinArikan and feel free to connect with him on LinkedIn here.
Credit: Image by Gerd Altmann from Pixabay
