Test Lucy. Tell us what happened.
Ten scenarios. Call, then log it here — takes about two minutes per call.
Run any or all of the scenarios below, in any order. After each call, scroll down and log what happened — anyone on your team can use this same page.
- Use a real number you don't mind a text landing on.
- One call, one form. Log it right after you hang up while it's fresh.
- There's no wrong answer — if something sounds off or breaks, that's exactly what we need to hear.
＋Faster option — combine into 6 calls instead of 10▸
Every scenario below still gets checked — several just share one phone call instead of each getting its own. Use this if you want to cover everything in fewer calls.
Call 1 — solo
Scenario 1 only. Everything else assumes this one passed clean.
Call 2 — combined
Scenario 2 + Scenario 4 together — open as a returning caller, then get impatient partway through.
Call 3 — combined
Scenario 3 + Scenario 6 together — a real emergency, called in rattled and hard to follow.
Call 4 — solo
Scenario 7 only — needs the transfer to genuinely fail, so it can't share a call with Call 3.
Call 5 — after Call 1's booking is in
Scenario 5 — call back to reschedule/cancel the appointment Call 1 just made.
Calls 6a–6c — quick
Scenario 8's three hang-ups. Under a minute each, no real conversation needed.
Tack onto Call 2
Scenario 9 (light banter) fits naturally once that caller's already relaxed and mid-booking. Scenario 10 (gallows humor) fits into Call 3 or 4, whichever feels more natural in the moment.
The first 8 are the core test — tap one open to see exactly what to say. The last 2 are bonus, lower-priority, run them if you have time. (Or use the combined call plan above to cover everything in about 6 calls instead of 10.)
1New caller, routine booking▸
What to do
Give a name, number, email, and address, say you're the homeowner, describe the drip. When Lucy offers times, ask for a different window than the first one offered before you accept one — that's the specific thing this call is testing.
What we're checking
- The booked time matches the window you actually accepted — not the first one offered and turned down
- Your info was taken once, cleanly, no repeats
2"I'm already in your system"▸
What to do
Open with that line, then give your info fresh when Lucy asks.
What we're checking
- Lucy still gets your info and books cleanly, even though you said you're a returning caller
- No error, no dead end
3Emergency + escalation▸
What we're checking
- Lucy recognizes the emergency and treats it that way
- You get transferred to a real person (or told clearly if no one's reachable)
- The call gets flagged urgent on our end either way
4Impatient / in a hurry▸
What we're checking
- Lucy acknowledges once and moves straight to booking — no long back-and-forth
- She still gets every required piece of info (name, number, address, email) and still checks if it's an emergency — impatience skips small talk, never the important stuff
5Reschedule or cancel▸
What we're checking
- Lucy does NOT say the change is already made
- She takes down what you need and says it's going to the team to confirm — this is the one most likely to reveal a gap, so don't be surprised if it does
6Panicked / hard to follow▸
What we're checking
- Lucy stays calm and doesn't rush you
- She still gets everything she needs, even if a question has to be repeated
7Escalation, nobody answers▸
What we're checking
- Lucy tells you plainly she couldn't reach someone right now and that it's been flagged urgent — not silence, not a dead end
8Hang up mid-call (do this 3 times)▸
There's no pass/fail here — we just want to know what's left behind when a call drops. Hang up abruptly at each of these three points, one call each:
Variant A
Hang up partway through giving your info, before anything's been booked.
Variant B
Hang up right after you've given all your info, before a time gets picked.
Variant C
Hang up right after a time is confirmed, before Lucy finishes wrapping up.
Log for each
Just describe what you noticed, if anything — did it feel like the call just vanished, or did it seem like enough was captured that someone could follow up?
9Bonus — light banter, caller warms up▸
What we're checking
- Personality is allowed to show up without derailing the required flow — Lucy can be warm without dropping a required step
10Bonus — stressed but human, gallows humor mid-pressure▸
What we're checking
- Permission to be warm never leaks into skipping urgency handling — the emergency screening and escalation still happen, humor or not
Got it — thank you.
That goes straight to the team building your agent. Anyone else on your crew can use this same page to log the next call.