> ## 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. # Results > Read a survey's results in the workspace's Results mode: the filters, the Overview metric cards, sessions over time, terminal statuses, the contact funnel, largest drop-offs and data quality, and the Questions, Outcomes and Routes views. ## Opening Results Open a survey and click **Results** in the **Setup | Results** switch in the workspace header. Results has six tabs: **Overview**, **Questions**, **Outcomes**, **Routes**, **Responses** and **Export**. This page covers the first four; see [Responses and export](/surveys/responses-and-export) for the other two. Results refresh on their own every 30 seconds while the page is open. A **Live** pill shows *Updated (time ago)* and *(n) sessions in range*. *** ## Sessions and calls Results are built from **sessions**. A session is one respondent's interview: it starts when a call gets past the [Gatekeeper](/surveys/gatekeeper) into the questionnaire and records every answer from then on. A call that never reaches the first question — no answer, voicemail, a screener, or a respondent who declined or asked for a callback at the greeting — files **no session**. Those calls are counted only in the [Contact funnel](#contact-funnel), which is built from call records. Each session has a **status** — its lifecycle — and an **outcome code** — the business result: | Status | Meaning | | ---------------- | -------------------------------------------------------------------------------------------------- | | **In progress** | The interview has started and has not ended. | | **completed** | The respondent finished the survey. | | **screened out** | A screen-out on a question ended the interview. | | **terminated** | The respondent stopped, or asked to be called back, part-way through. The outcome code says which. | | **abandoned** | The interview was under way and the call dropped or timed out. | | **failed** | The session could not be recorded properly. | See [Closing and outcomes](/surveys/closing-and-outcomes#how-outcomes-appear-in-results) for how outcome codes map to statuses. *** ## Filters The filter bar applies to every Results tab: | Filter | Options | | ------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Version** | **All versions**, or one version (*v3 · published*). Answers from other versions are matched to questions by question key. | | **Range** | **Last 24 hours**, **Last 7 days** (default), **Last 30 days** or **All time**. Times are in your browser's time zone, shown beside the filter. | | **Channel** | **All channels** or **Voice**. | | **Status** | **All statuses**, or one session status. | | **Outcome** | **All outcomes**, or one of the codes seen. | | **Include test sessions** | Off by default, so sessions from calls that ran an unpublished version — such as [test calls](/calls/test-calls) on a compiled draft — do not skew your figures. | **Reset** appears once you change any filter. *** ## Overview ### Metric cards | Card | Hint in the app | | ---------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ | | **Dialled** | *"Every call made with this survey's compiled scripts in the range. Connected means the call was answered and lasted longer than zero seconds."* | | **Started** | *"Respondents who entered the questionnaire. A call that never reached the first question files no session."* | | **In progress** | *"Sessions whose stored status is in\_progress. Not “active now”."* | | **Completed** | *"Sessions whose stored status is completed."* | | **Completion rate** | *"Completed ÷ started. In-progress sessions stay in the denominator."* | | **Screened out** | *"A screen-out rule on a question ended the interview. Refusals, callbacks and no-contact at the greeting are not sessions."* | | **Terminated** | *"The respondent stopped or asked to be called back part-way through. The outcome code says which."* | | **Abandoned · failed** | *"Abandoned: the interview was under way and the call dropped or timed out. Calls that never reached the first question file no session."* | | **Avg duration** | The average interview length across the finished sessions shown. | ### Sessions over time A daily chart of **Started** against **Terminal** (finished) sessions across the range — at most the last 30 days, and the last 14 days for **All time**. *"The most recent bucket is still open — treat today as partial."* ### Terminal statuses How sessions split between in progress, completed, screened out, terminated, abandoned and failed, each as a share of started sessions. The card notes *"Lifecycle status and business outcome are separate dimensions"* — use the **Outcomes** tab for outcome codes. ### Contact funnel The funnel counts **calls, not sessions**, from the call records — so it includes the calls that ended at the gatekeeper. Its stages are **Dialled**, **Connected**, **Entered survey** and **Completed**. The test-session switch does not apply here. Below the stages, two lists break down where calls fell away: * **Not connected** — dials that were never answered. * **Connected, did not enter** — answered calls that ended before the first question. Each lists calls by their recorded call outcome, such as **No answer**, **Voicemail**, **Not interested**, **Callback requested**, **Wrong number** or **Caller hung up**. See [Call Statuses](/calls/statuses) for what each call outcome means. ### Largest drop-offs The questions where the most interviews ended, shown as *(key) — trace ends here*. A question high on this list may be confusing, too long, or sensitive. ### Data quality *"Diagnostics — nothing is cleaned up silently."* The card flags: * *"(n) answers recognised below 0.6 confidence."* * *"(n) superseded answers retained — only final answers are counted."* * *"(n) terminal sessions carry no stored answers."* * *"(n) answers reference a question outside this version."* * *"(n) sessions have been in progress for over 24 hours."* With nothing to report: *"No data-quality issues detected in this cohort."* *** ## Questions One card per question (Display questions are left out), headed *Question (n) · (key)*, with the prompt, how many answered, and the top 12 answers with their counts and percentages. Only final answers are counted — an answer the respondent corrected is replaced by the correction. For a [battery](/surveys/question-types/battery), each rated item is counted as its own answer. *** ## Outcomes The **Outcome codes** table lists each code with **Outcome**, **Sessions** and **Share of coded**, followed by *"(n) of (m) sessions carry no outcome code."* — usually interviews still in progress or dropped part-way through. *** ## Routes **Route and drop-off** — *"Sessions reaching each ordered section, inferred from stored answers."* Each section shows how many sessions reached it and what share that is, and how many ended there. Use it to see where [routing rules](/surveys/routing-rules) send respondents and where they stop.