What this episode is about
Over the years my team and I have looked at hundreds of QA forms from clients, and when we were recording this, about another hundred in just a couple of months. Some are good, some are just okay, and some customers don't have one at all. This is the order we follow when we build one from scratch.
Decide what the form is for first. Agent performance, compliance, insight from your calls. It probably shouldn't be just one of those. And keep it short. Big forms usually ask the same thing from three angles, because every new manager added two or three questions of their own.
Opening: was the agent ready, was the tone right, did they verify properly. Two or three questions. Etiquette is the core: true empathy, call control without being pushy, professionalism that fits your brand, active listening, and hold and dead-air procedures. Then the body of the call, which depends on whether you're sales, retention or support, and then a proper closing.
The one most people skip is a small analytics section of three or four yes/no questions. Did they cancel? Was price the reason? Did they mention the Facebook ad? Your QA team, human or AI, can be your analytics tool. For scoring, a simple scale is fine, and I'd use auto fails for must-read disclosures instead of weighting every section.
Key takeaways
- Know what the form is for: performance, compliance, insight, or a mix.
- Less is more. Cut redundant questions that piled up over the years.
- Opening and verification should take two or three questions.
- Call control, real empathy, listening and hold procedures are the core of the form.
- Add three or four unscored analytics questions, and use auto fails for required disclosures.
In Tom's words
“I'm a huge believer that less is more”Tom Laird · 06:50
“So you could almost use your humans as an analytic tool”Tom Laird · 06:10
“I would prefer you to do auto fails.”Tom Laird · 21:43
Transcript excerpt
These excerpts come from an automatic transcript and are lightly edited for readability (filler and obvious transcription errors removed). For exact wording, listen at the timestamp.
[06:50]I'm a huge believer that less is more. When I see huge QA forms, a lot of the questions are kind of redundant, right? They're asking the same thing in different angles, and it can be kind of confusing for AI, and it can be confusing for a human to really get to the core. And a lot of times this happens because a lot of people have an input into the QA form. What normally happens is, as somebody takes over QA, or takes over as a call center manager, they add their two cents to the form.
[12:04]The third is just kind of the blanket: did they use proper verbiage, tone, professionalism. And this is different for everybody, because I think this is a culture question, right? If you're at a bank you have to be prim and proper, right? Yes sir, no sir. But if you're working for maybe a skateboard shop, it's, what's up, man? Are they actively listening? Are they not talking over the customer because they're getting bored? Did the customer say something earlier and the agent just totally forgot about it? That's so annoying to a customer.
[18:22]Normally that's kind of where we would end, right? I think having an analytics section, a small section of maybe three or four questions: did the customer cancel? Was it too expensive? Did they say anything about the product design? Did they say anything about an advertisement we had on Facebook? All these things that we normally ask agents about in a focus group, start to ask those questions on the form.
[20:47]I would prefer you to do auto fails. So again, if a specific disclosure wasn't read, or they didn't verify appropriately, they get a zero for the call or for that section. I would prefer you do that than go through and weight every single section. And there's no reason that you can't set up what I just talked about on Excel, on a Google Sheet, or even a Google Form.