Call outcomes and dispositions

The outcomes that have fixed words#

The first four are refusals or non-events, and _capture_lead creates nothing from any of them. That rule is stricter than it looks: a busy outbound list can produce hundreds of rows that are technically calls and contain no information, and a pipeline that treated them as contacts would be measuring dialling rather than interest.

no_answer and not_reached are the pair most often conflated, and they point at different problems entirely — one at the list being called, the other at the machinery doing the calling. Calls that were never a conversation takes that distinction apart properly.

Disposition belongs to a person#

Disposition is not a second guess at the outcome. It is the business's own reading: whether the call mattered, what it was really about, what should happen next. It is written by a human, at whatever moment they decide, and it is never written by the engine or by a provider callback.

That separation is structural rather than stylistic. Carriers deliver status callbacks after a conversation is over — sometimes minutes after — and if the two shared a column, the machine's late write would win by arriving last. Keeping them apart lets a late webhook update duration, provider status and cost without touching a colleague's judgement.

Where outcomes surface#

The Phone screen

Rows grouped by what they were, so refusals do not pad the count of conversations.

Line health

Repeated not_reached results past the failure threshold inside the health window raise a Needs You item, because they indicate the line rather than the callers.

Follow-up drains

An outcome decides whether a promise was kept. A call-back that never rang has not been made, whatever the attempt count says.

The #/data grid

Calls are one of the thirteen sheets, so outcome is filterable and exportable alongside every other field.

Cost

Every outcome is charged, including refusals — briefly, from the call's own duration when no usage was reported.

Counting calls honestly#

1. Exclude line_off and after_hours from any measure of demand you handled — then count them separately as demand you refused.

   Result Two numbers that mean different things, instead of one that means neither.

2. Split the remaining failures into no_answer and not_reached.

   Result The first is about who you are calling; the second is an engineering problem with a Needs You item behind it.

3. Compare the conversation count against the carrier's own completed calls for the same window.

   Result A gap is a missing row, which is a routing or verification defect rather than a reporting quirk.

The engine's outcome is the honest denominator for all of this, which is why it is kept out of a person's reach. If a colleague could edit it to reflect what a call felt like, the one column that reliably describes the wire would stop describing it.

Questions#

Related#

• Calls that were never a conversation
• The call record
• Who ended the call
• Phone line health
• Leads captured from a call
• Headline numbers that stop at the page size