- 1 day ago
- 4 min read
The industry built quality measurement around one assumption: if it isn't recorded, it can't be scored fairly
Every serious quality management programme in customer service starts from the same foundation. A call gets recorded. A transcript gets generated. An evaluator scores it against a rubric, and if a frontliner disputes the score, the recording settles the argument. This is why call centre quality assurance became the industry standard it is today. The proof already exists. Nobody has to trust a human observer's memory or judgment, because the tape does the trusting for them.
This foundation is so old that most organisations no longer notice it's an assumption at all. It just feels like how quality measurement works.
That assumption is also why physical, face-to-face service was left out
Walk onto a retail floor, a bank branch, a hospital reception desk, and there's no recording. No transcript. Just a live moment between a customer and a frontliner, and whatever a supervisor happens to observe, if anyone is observing at all.
The industry's response to this has not been to build a way to measure it. It's been to leave it alone. Every major quality assurance vendor now markets "omnichannel QA," and it's worth being precise about what that actually means. It means voice, chat, email, social. It does not mean the counter, the branch floor, or the person standing in front of a customer with no camera and no transcript running. That environment was never treated as unmeasurable because it matters less. It was left out because nobody could prove, the way a recording proves, that the score given was fair.
The gap isn't a technology gap. It's a trust gap
This is the part worth sitting with, because it changes what the actual problem is. It would not be difficult, technically, to send a supervisor onto a retail floor with a scorecard and have them evaluate five interactions a day. Organisations already do exactly this, informally, in mystery shopper programmes and spot checks. What's missing isn't the ability to observe. It's a way to make that observation trustworthy enough that a frontliner accepts the score instead of contesting it, and an organisation trusts the data enough to act on it.
A recording earns trust through evidence. An unrecorded, live, subjective moment has to earn trust a different way.
What we built instead of a recording is transparency
Our answer to this was not to find a way to record the floor. It was to replace the recording with immediacy. When a frontliner is evaluated, they see their own score right after it happens, not at the end of a month, not buried in an aggregated report they never open. The proof isn't a tape they could review if they contested the outcome. The proof is that they saw the scoring in near real time, against a scorecard they helped build, before it ever became a number that only management could see.
This matters more than it sounds like it should. A score a frontliner discovers weeks later, with no memory of the interaction it refers to, invites suspicion. A score they see immediately, tied to something they just did, invites correction and conversation instead. Transparency doesn't just make the system feel fairer. In an environment with no recording to fall back on, it is the thing that makes the score defensible at all.
This is also why the same evaluation, tracked over time, becomes something worth watching
Once a frontliner's quality scores exist consistently, sampled daily, visible to them, alongside the customer's own satisfaction scores collected separately, a pattern becomes visible at the individual level that neither number shows alone. A frontliner whose internal quality scores are consistently strong but whose customers report lower satisfaction is a different situation from one where both numbers are weak, or where a customer's rating looks like an outlier against everything else on record. Watched at the level of one person over time, the relationship between how well someone is following the standard and how the customer actually experienced it becomes something to investigate, not something to write off as noise.
Most organisations already sense that internal quality scores and customer satisfaction don't always move together. What's rare is a system built to notice that gap at the individual level in the first place, rather than treating the two numbers as separate reports that happen to sit on the same dashboard.
The same transparency that makes the measurement possible is what makes people want to use it
Here is the part that surprised us. Ask supervisors at large, established organisations to adopt a new internal quality system, and the resistance is predictable. It adds work. It feels like another way to be judged. Ask managers at a company just beginning this process, and we've seen the opposite reaction, genuine eagerness to get started.
The difference has not been the size of the organisation or how much bureaucracy exists. It's been what the evaluation is used for, and whether the people being measured can see it happening to them. A system introduced as surveillance, something done to frontliners for management's benefit, gets resisted, reasonably. A system introduced as support, where the person being evaluated sees their own score as fast as their supervisor does, gets adopted with far less friction.
Transparency, in other words, isn't a courtesy layered on top of the measurement. It's the same mechanism doing two jobs. It's what makes an unrecorded moment measurable in the first place, and it's what makes the people being measured trust the system enough to act on it instead of fighting it.
The question this leaves the industry with
Call centres got measured properly because the recording made objectivity possible. The floor got left behind because nobody found a substitute for the tape. If transparency, not recording, turns out to be what earns trust in an unrecorded moment, the real question isn't whether the physical service floor can be measured as rigorously as a call centre.
It's why the industry assumed, for this long, that it couldn't be.
