For the past 30 years, we have not calculated first contact resolution (FCR) or digital containment rates correctly.
There are hundreds of articles and posts on how to define FCR, including:
- Asking customers whether their issue was resolved
- Polling frontline employees to see if they solved the customer’s issue
- Tracking whether the customer contacted you again within 7 or 14 days.
None of these metrics is correct.
Misleading Metrics
Often, customers do not know if their issue is resolved, and survey response rates are unreliable. Asking frontline staff adds bias, especially when FCR impacts their evaluations. Tracking follow-up contacts is problematic since new contacts may be unrelated.
Digital solution containment rates are a more recent phenomenon, and they are also prone to significant error. Once again, you can ask customers if their issue was resolved in the app, Chatbot, Agentic AI bot, or other digital tool, or you can calculate the percentage of customers who do not require assisted support as being contained or resolved.
The problem, as I’ve indicated in earlier articles and in my 2022 book with David Jaffe, The Friction Organization, is due to Silent Suffers (which I will capitalize for emphasis).
Academic researchers into Customer Complaint Behavior (CCB) have proven that non-complainers outnumber complainers by as much as 20 to one, with a rough average of six to one. While we obsess (quite correctly) over handling customers who do reach out to contact us, whether to complain or to ask routine questions, many times more customers won’t even bother. Why? They give up, don’t know where or how to complain, are hesitant to voice a complaint, or fear retribution.
And, one more reason: Being reasonably satisfied with their interactions, which we define as Happy Campers. However, research has also shown that Silent Sufferers outnumber Happy Campers by three or four to one, meaning that the majority of non-complainers are not happy and don’t bother to complain.
In the commercial world, this leads to an 88% churn rate. In essence, customers leave as soon as their contract expires or another option appears, and they don’t tell you. No exit interviews! Or, Silent Sufferers will buy less over time, refuse new offers or upgrades, and still not tell you. For government services, where there’s essentially no other choice, it could lead to financial or health problems, as well as continued anxiety and upset with governmental services.
Let’s do some quick math. If you think that your FCR is 72%, this means that 28% of the time, customers did not get the answer that they wanted. However, using the Silent Suffer to Happy Camper to Complainers (S2:HC:C) ratio, FCR is more likely to be 45 to 48%, a significantly worse number to imagine. If you believe your digital containment rate is 40%, at least on average, this means that 60% of the time, customers do not get the answer they want and require subsequent live assistance. Using the S2:HC ratio, the digital containment is probably a much weaker 18 to 24%.
What to Do?
First, don’t rely on surveys or broad-brush estimates, such as no contact within 7 to 14 days of a previous contact means the issue was resolved.
Next, identify and reach out to the non-complainers, especially those who are the Silent Sufferers, using AI and LLMs.
Finally, realize that FCR and digital containment rates, both associated closely with customer satisfaction and loyalty, are much worse than we’ve been calculating and reporting for decades. We need new solutions, and we need them fast!
This last point should send shivers up the spines of CX analysts, CX practitioners, and C-level executives. It may explain why overall customer satisfaction, NPS, customer effort, and other scores are so low and, in many cases, declining. Applying the S2:HC:C ratio to survey response rates means that the actual satisfaction, loyalty, and effort scores are much lower than are being reported and celebrated.
CSAT Illusion
Years ago, we did a project for a high-tech company that was proud that their newly-launched product already had a customer satisfaction rate of 81%. They asked us to determine how they could increase CSAT above 90%. After conducting many interviews at their head offices, we fanned out across multiple context centers to get a feel for what was really happening at the coal face with frontline employees supporting their customers.
What we discovered shocked everyone. Because the company’s process only sent post-contact surveys after closed cases, agents who were hesitant about their performance or their resolution capability would simply not close the cases. This meant that there was a built-in bias in the customer satisfaction score. We recalculated that the starting CSAT rate was not 81%, but rather below 70%. This meant there was a much bigger gap to reach the desired 90% score, and it also led to changes in their practices and policies.
Listen to the Voice of Data
Today, we can use the S2:HC:C calculations with what my team and I are calling Voice of Data or VOD. We no longer need to rely on expressed or shared VOC since, as I’ve attempted to prove in this article and elsewhere, it’s not a reliable indicator of customer satisfaction or loyalty.
Stay tuned. I will cover VOD, how to identify and turn around Silent Suffers, how to calculate more accurate FCR and digital containment rates, and how to use AI and LLMs to increase both of these important metrics in the weeks and months to come.