Zenylitics is a managed conversation intelligence firm that applies speech analytics and analyst review to contact center data. In this engagement, Zenylitics measured call center dissatisfaction at a national legal intake operation and traced it to 11 root causes.
Industry: Legal services, mass-tort intake and client services
Challenge: Calls were going badly at scale, with no way to count how often or explain why
Result: A 71% drop in the specific driver Zenylitics and the client targeted, measured over a three-month window, alongside broader year-on-year declines in overall dissatisfaction
Company Overview
The client is a national mass-tort legal intake and client-services operation, handling a high volume of inbound calls from prospective and existing claimants. The contact center function sits at the front of the client relationship, since a claimant’s first experience with the operation shapes whether they proceed with a claim and how they feel about the process afterward. Leadership’s objective was to understand what was actually happening on calls that went poorly, and to fix it. Zenylitics has delivered monthly readouts to this client without a break since August 2024.
The Challenge
Everyone in the organization sensed that some calls were going badly. What nobody could say was how many, which calls, or why. The operation had no systematic count of dissatisfied interactions and no defined list of the reasons behind them. Leadership was managing a real problem using instinct rather than data, which meant coaching, staffing, and process decisions were all being made without a clear picture of what was actually driving claimant frustration.
Why the Existing Approach Failed
A feeling is not a metric, and an unmeasured problem cannot be prioritized against other operational demands. Without a count, leadership had no way to argue for resources to fix the issue, since there was nothing to point to beyond anecdote. Without named causes, any fix attempted would have been a guess, since a team cannot coach toward a specific behavior change if nobody has identified which behaviors are actually causing the dissatisfaction in the first place. The organization needed both pieces, the scale of the problem and its specific drivers, before any coaching or process change could be targeted rather than speculative.
The Zenylitics Approach
Data Layer. Every call across the intake and client-services operation, rather than a manually selected sample, so the resulting picture reflected the full population of interactions.
Analytics Layer. Every call scanned for dissatisfaction signals, applied consistently at scale. Across 2025, that scan covered 165,424 calls and flagged 3,965 of them as dissatisfied, giving the organization its first real count of the problem it had only sensed before.
Human Intelligence Layer. Zenylitics analysts defined and agreed 11 root causes with the client, then listened to and attributed every flagged call by hand each month against those causes. The causes were also categorized into two groups, issues the contact center could actually influence through coaching, such as agent behavior on the call, and issues rooted in the client’s own broader claims process, which coaching could not touch. That split did more than organize the findings. It kept the analysis, and the client’s own expectations, honest about which parts of the problem coaching could realistically move and which parts sat outside the contact center’s control entirely.
Action Layer. Each month’s ranked causes became that month’s coaching plan, with the current top driver going directly to the team responsible for addressing it.
Key Insights Discovered
The 11 root causes, once ranked, showed the organization something it had never had before, a monthly, evidence-based list of exactly what to work on next. Rather than treat every driver as equally coachable, the ranking made the distinction between controllable and uncontrollable causes visible and actionable. Settlement time, for example, reflects the client’s own claims process and does not move regardless of how well an agent handles the call. Agent behavior, by contrast, is something a contact center can directly train and improve.
That distinction led to the engagement’s clearest finding. Agents who used a specific rebuttal technique when handling a recurring objection saw measurably different outcomes than agents who did not. Calls where a rebuttal was used carried 23 percent less dissatisfaction than calls without one, and scored meaningfully higher across the specific behaviors behind that difference, with ownership language up 20 percent, empathy up 30 percent, and confident language up 35 percent on rebuttal calls compared to calls without. This gave supervisors something concrete to coach toward instead of a general instruction to “handle objections better.” One identifiable behavior, easily coached, connected directly to a measurable shift in claimant sentiment.
Business Outcomes
Agent-behavior complaints, the specific driver Zenylitics and the client targeted through coaching, fell 71 percent over a three-month window from November 2025 to January 2026, moving from 21 percent of dissatisfied calls down to 16 percent and then to 6 percent. This is the clearest proof that the coaching intervention worked, since it tracks the exact behavior category the program was built to change.
Onshore dissatisfaction, measured as a year-on-year rate comparison (January to May 2026 compared with January to May 2025), declined 20 percent. This is a broader measure than the targeted agent-behavior driver, covering the full range of dissatisfaction causes rather than one specific category, and reflects the cumulative effect of the monthly coaching cycle across the whole onshore operation. Offshore dissatisfaction also declined over the same period, by 5.3 percent.
Dissatisfaction across all calls, also measured year on year (January to May 2026 compared with January to May 2025), declined 7.5 percent while call volume rose. This is the widest lens available, spanning every root cause including the ones coaching cannot influence, which makes it the most conservative of the three figures and the hardest one to move.
Across the three measures, the narrowest and most coachable metric moved the most and the broadest metric moved the least. Settlement time complaints, which coaching cannot influence, held at a similar rate over the same period, while agent-behavior complaints fell significantly. The behaviors the monthly coaching cycle targeted are the behaviors that moved.
Why This Matters
This operation went from managing a complaint problem on instinct to running a monthly cycle where evidence, not guesswork, determined the coaching priority. The distinction between controllable and uncontrollable causes mattered as much as the count itself, since it kept the resulting program focused on the behaviors coaching could actually change rather than chasing outcomes no training session could ever move.
A count shows that a problem exists, and a ranked list of causes shows what to do on Monday.
FAQs
What causes call center dissatisfaction in legal intake?
Call center dissatisfaction in this legal intake operation came from 11 root causes that fall into two groups. The first group covers causes coaching can influence, such as agent behavior, and the second covers causes tied to the client’s own claims process, such as settlement time.
How does speech analytics reduce call center dissatisfaction?
Speech analytics flags dissatisfied calls so analysts can attribute each one to a root cause and turn the ranked causes into monthly coaching plans. In this case, the program scanned 165,424 calls in 2025 and flagged 3,965 as dissatisfied.
How much did agent-behavior complaints fall?
Agent-behavior complaints fell 71 percent over three months, from 21 percent of dissatisfied calls in November 2025 to 6 percent in January 2026. This measure tracks the specific driver the coaching program targeted.
Request a Demo to see how speech analytics can identify the causes of call center dissatisfaction in your contact center.