Method notes
How we measure what customers say, and what the measurements cannot tell you. Numbers with their samples, and limits stated rather than buried.
- Practical guide
How to analyze a focus group: from the recording to a quote that holds up.
A focus group is not analyzed to see what comes up. It is analyzed against the questions the study has to answer, keeping track of who is speaking and citing every finding with its minute.
30 September 20266 min read
- Call analysis
Your QA sample rates agents well. It finds themes badly.
Both jobs use the same recordings, so it is reasonable to assume one sample serves both. It does not, and the reason is arithmetic rather than opinion.
9 September 20266 min read
- Method
Two review queues. Only one of them measures anything.
Every vendor says a human reviews a portion of the output. Almost none say how that portion was chosen, and the choice is what decides whether you get a number or a promise.
9 September 20265 min read
- Data model
The agent did not fail. The microphone did.
A quality system that cannot say "I could not tell" will say "no" instead, and the difference between those two answers is somebody's performance review.
9 September 20264 min read
The deep material lives with the product.
Survey Coder and Voice Capture each keep their own writing. These are the pieces worth starting with, and they open on their own sites.
Survey Coder
Open-end coding: codebooks, NPS verbatims, and what four AI layers plus a human reviewer actually do.
Voice Capture
Spoken answers inside a survey: how it works, what the measurement says, and how to add it to Alchemer.