Role
Service quality
How many calls and cases do you review in a month against how many there are? If it is a few dozen out of several thousand, your assessment describes a sample while you are accountable for the standard across the whole team.
Intro call
Four questions we would start with
How many cases do you actually review in a month?
If it is a fraction of a per cent, conclusions about service quality are statistically thin - and yet conversations with people are held on the basis of them.
How do you know two assessors' scoring is comparable?
Without calibration two auditors score the same call differently. The agent sees it and loses trust in the whole system.
Who notices first that the standard is starting to slip in one area?
If nobody does until the quarterly review, the problem has time to settle in and spread through the team.
What do you not do, even though you know it has to be done?
Usually feedback on time. The score appears a month after the call, when the agent no longer remembers it - and then it teaches nothing.
We take over the reviewing, you keep the conversation with the team
Reviewing cases against formal criteria, pointing out the cases worth scoring and putting the results in order can be taken off the desk. Judging the quality of a conversation, and the feedback, stay with a person.
- This is a development tool, not a scoring tool. - The distinction has to be made at the outset and has to be known to the team - otherwise service quality turns into surveillance and agents start playing to the score.
- The substantive judgement stays with a person. - The system points out the cases worth watching and checks the formal criteria. Whether the conversation was good is settled by the person who listened to it.
- Feedback has to be fast. - A score a month later teaches nothing. Shortening that time matters more than increasing the number of calls scored.
Scope
What can be taken off service quality work
Eighteen example tasks. Each one has a human approval point. Below is a selection, and the scope we fit to your process.
The standard and the criteria
- Writing the scoring criteria down in a form the team understands
- Examples of cases that meet and do not meet the standard
- Calibration of scoring between auditors (run by: the specialist)
- A register of changes to the criteria
Reviewing cases and calls
- Pointing out the cases worth watching: unusual, long, ending in escalation
- Checking the formal criteria across a wide sample
- Detecting cases where the client came back repeatedly
- A summary of the topics where the standard holds least well (assessment: the specialist)
- Pointing out cases where a procedure or a permission was missing
Feedback on the standard
- Material for a team conversation about the standard (the conversation: the specialist)
- A summary of solutions worth repeating, not only of shortcomings
- Examples of cases to discuss at a calibration session
- Reminders about scheduled reviews
- A register of what was agreed and how far it was delivered
Training and measurement
- Pointing out the topics that need team training
- Examples of good solutions to use in training
- Quality scores set against client satisfaction
- A periodic report (conclusions: the specialist)
A day's work
Today
- You pick the cases to review at random or by the time available
- The score appears weeks after the call
- Two auditors score similar cases differently
- You give feedback in a batch, once a month
- You do not know whether the quality score translates into client satisfaction
- Examples of good solutions are lost in the case history
With AI assistance
- The cases that reach review are the ones genuinely worth watching
- The score is closer to the call, so it has a chance of teaching something
- The criteria are written down and calibrated, so the scores are comparable
- Conversations with agents have material and dates
- You can see whether the quality score moves with client satisfaction
- Good solutions go into training rather than into the archive
- Coverage by scoring
the conclusions rest on a wider basis
the formal criteria are checked widely, the substantive judgement deliberately
- Comparable scoring
agents trust the score
the criteria are written down and calibrated between auditors
- Time to feedback
the score teaches, because it arrives while the call is remembered
the review stops waiting for a free week
- The link with satisfaction
it is known whether the standard relates to what clients think
both measures are set side by side rather than kept apart
We start with the criteria, not with a tool.
- Step 1
We write the criteria down and calibrate
With the team, not from behind a desk. The exercise usually reveals that the auditors understand the standard differently.
- Step 2
Pointing out the cases to review
Instead of drawing at random - unusual cases, long ones, with escalations and with clients coming back.
- Step 3
Shortening the road to feedback
Material and dates for the conversations. This is the change the team feels fastest.
- Step 4
The link with satisfaction and with training
Summaries and examples for training. It needs a few months of data.
Common questions
Questions that come up at the first meeting
Does less work on reviews mean a smaller quality team?
Service quality teams are usually small and look at a tiny sample of cases - not because the rest does not interest them, but because they cannot get through it. Taking over the formal review lets the same team's attention reach more cases. The substantive judgement stays entirely with people, which is what the paragraph below is about.
Does this fall under the AI Act?
Systems used to monitor and assess employees are named expressly in the AI Act as high risk - the same category as recruitment, with obligations from day one, not above some scale. That is why the scope stops before that line: we measure the process and the topics of cases, not people. The summaries say in which kinds of cases the standard holds least well and where a procedure is missing - they do not say which agent comes out how. If you want to assess the work of particular people, that is a separate rollout, separate obligations and a separate conversation, including with the workforce.
Still have a question that is not here? Book a call
Next step
How long passes between a call and the feedback on it?
We will show you what service quality work looks like when the cases that reach review are worth watching and the feedback comes back while the call is remembered.