| Trigger | Import a list, dial in bulk, read the script | Bound by rota; you cover only what you staff | The conversations customers start with you are one team’s job; your SOP decides who follows up, and when |
|---|
| Turns | One call, script branches, and it ends there | Real back-and-forth; how far it gets varies by person | Each stage has a goal and an entry condition; unmet, it doesn’t move on |
|---|
| Intent | Keyword matching; vague answers slip through | They read it, but each rep has their own bar | You define how intent is graded; the same scale runs every time |
|---|
| Follow-up | One call, then a person takes the rest | Runs on memory and stamina; most stop by the third try | Full context is kept, so it picks up over days and rounds |
|---|
| When it’s wrong | Edit the script, and hope nothing else broke | Reviews rely on retelling; calls go unheard | Checked on the review workbench first; replayed turn by turn after; configs roll back |
|---|