> ## Documentation Index > Fetch the complete documentation index at: https://docs.voxworks.ai/llms.txt > Use this file to discover all available pages before exploring further. # Behaviours (Busy and Stop) > The Behaviours fixed section: the busy and stop routes that end a survey part-way through, what counts as each, their outcomes and priorities, why Always fires is normally left off, and the behaviour step levels. ## What are Behaviours? Two things can end a survey part-way through, whatever question is being asked: the respondent is busy, or they want to stop. The **Behaviours** fixed section is where you describe what counts as each. It is described as *"Busy and stop — what counts as each, in the respondent’s words."* The pane adds: *"What each route does is fixed — busy ends the call as callback\_requested, stop as refused. What counts as one is survey content."* Open **Behaviours** in the **Outline** on the [Questions tab](/surveys/questions-tab), edit the two groups and click **Save**. The confirmation is *Behaviours saved*. *** ## The two routes | Group | Default priority | What it does | | ----------------------- | ---------------- | ------------------------------------------------------------------------------------------------------------------------------- | | **Busy / call me back** | 20 | Ends the call with outcome `callback_requested`. The respondent hears the `callback_requested` closing text, then the farewell. | | **Stop the survey** | 30 | Ends the call with outcome `refused`. The respondent hears the `refused` closing text, then the farewell. | Both are live for the whole call, from the consent question to the last question. When either fires during a question, that question is recorded as `abandoned`, and the session ends as **terminated** in [Results](/surveys/results) with the outcome code saying which. The busy route does not ask for a callback time. A time is asked only when the respondent says "not now" to the consent question at the start of the call (see [Gatekeeper](/surveys/gatekeeper#3-consent)). Whether and when the contact is called again is decided by the batch the survey runs in — see [Calling Rules & Retries](/batches/calling-rules). Write the closing texts for both codes in the [Closing](/surveys/closing-and-outcomes) section — for example *"No problem, we'll try you another time."* for `callback_requested` and *"No problem at all, I'll let you go."* for `refused`. *** ## Fields Each group has: | Field | What it holds | | ------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Priority** | From 1 to 999. A lower number ranks higher. Busy and stop rank above every [FAQ](/surveys/faqs) (41–99). | | **Accept** | *"What counts as this — one condition per line, in the interviewer's words."* | | **Often heard as** | *"Recogniser mishears only. These are never seeded as keyterms."* | | **Reject** | *"What must not match here, and which route it belongs to instead."* | | **Always fires** | *"Skips the check that decides whether the caller was answering the question instead. For a route where a miss costs more than a wrong fire. Normally left off here: a “no” or a “not right now” to a question is what this route misreads."* | ### Example Accept lines **Busy / call me back** * *Says it is a bad time, or asks to be called back later* * *Says they are driving, at work or about to go out* **Stop the survey** * *Asks to stop, or says they do not want to continue* * *Asks to be taken off the list or not called again* ### Reject lines that matter A short answer to a question can sound like a behaviour. *"No"* to *"Do you use public transport?"* is an answer, not a request to stop; *"Not right now"* to *"Are you currently working?"* is an answer, not a request to call back. The interviewer already checks each reply against the question just asked and treats a direct answer as an answer by default, and it automatically keeps busy and stop from firing on a refusal of one question (*"I'd rather not answer that one"*) or on a voicemail greeting. If you still see a behaviour fire on answers, add **Reject** lines naming them, for example *"A one-word answer to the question, such as yes, no, never or unsure"* or *"Declining to answer one question — that is a skip, not a stop"*. ### Why Always fires is normally off **Always fires** takes a route out of the answer-or-detour check, so it fires whenever its Accept lines match. For busy and stop that is almost always wrong: the words that trigger them are the same words respondents use to answer questions, and a false stop throws away the rest of the interview. Leave it off unless a real call shows the route being missed. See [Answers given early and interruptions](/surveys/riders-and-interruptions). *** ## Behaviour step levels The pane ends with its own **Step levels** group: *"The gates the busy and stop closes are spoken under — their own, not the survey's. The reason is already recorded and the call is ending, so the line is delivered and the wait after it is short."* | Level | Unset here | | ------------- | --------------------------------------------------------------------------- | | **Eagerness** | High | | **Idle** | Low | | **Thinking** | The survey's own level from the [Agent](/surveys/agent#step-levels) section | Each can be set to **Survey default**, **Low**, **Medium** or **High**. *** ## Requests to hear the question again There is no survey-level behaviour for *"can you repeat that?"*. Every question handles a request to hear it again itself, re-reading it in different words; each re-read counts towards the question's **Ask** cap (see [Avoiding repetition](/surveys/repetition)). If an imported survey defines a survey-level behaviour for this, it is removed on import with a warning, and the question handles the request itself.