Collect Tab Overview

Written By revealai

Last updated 6 months ago

The Collect Tab is where you’ll find all of your interview collection data. In this article we’ll review the sections and functionality of this tab.

Preview vs. Live Interviews

Preview Interviews are any interviews that are captured while the chatbot is set to “Testing” or a user has taken a live interview with the appended url parameter or ?preview=true. For example: https://interview.dev.getreveal.ai/i/revealai-demo?preview=true

Live Interviews are any interviews that are taken while the chatbot is in “Collecting” mode. Excluding interviews that are explicitly declared to be a preview as outline above.

As discussed in other articles, Preview Interviews don’t qualify for analysis. For this reason, it’s imperative that the chatbot is switched to Collecting. (note: there are ways to qualify interviews for analysis that we’ll discuss later, but it’s tedious and better to avoid.)

To help make sure the chatbot is in the correct mode, there will be a notice at the top of the tab showing that the chatbot is in Testing mode and action button to launch / move it into Collecting mode (see screenshot below). This notice will not be visible if your chatbot is in Collecting or Closed mode.

Once you have interviews that fall into each of the categories above (Preview and Live), the “Show Preview Interviews” toggle (see screenshot below) can be used to show and hide the preview results from the tab. When enabled, preview interviews will chart as a separate, yellow line.

Interview Metrics

The Collect tab compiles a handful of metrics to help provide a snapshot of what’s going on with your data collection. These are:

  • Interviews Started: Total number of interviews started

  • Interviews Active: Number of interviews currently in progress

  • Interviews Qualified: Number of interviews to be included in analysis

  • Interviews Complete: Number of fully completed interviews

  • Median Duration: Median interview completion time (Incomplete interviews are excluded)

Data Filtering

There are two primary was to filter your data:

  • Date Picker: Use the picker to select a start and end date

  • Chart-based Selection: Click on a date to set as a start date or drag across a select time frame

The filters will apply to the entirety of the tab, including the Interview Metrics outlined above, chart data and transcripts. The “Reset” button will return the tab to its default state.

Responses and Transcripts

Responses By Questions provide a glimpse at each question asked. This included the dropoff rate for each questions (how many participants stopped at that question) and the number of completed responses.

Transcripts provide detailed information about each response. They can be searched by Tracking ID, if one has been provided. Other data provided includes: Timestamp, Duration, Eligible for Analysis, Status.

The entirety of the transcript may be accessed by clicking the “Chat” icon to the right of the table. The transcript modal provides the following functionality:

  • Copy Transcript Link: Copies a shareable transcript link to the clipboard

  • Copy Specific Question or Response Link: Copies a shareable link to a specific spot in the interview

  • Include in Analysis: Toggle whether an interview is included in analysis (see)

Controlling Which Interviews are Analyzed

Each interview has a property which dictates if it is included in analysis runs. Reveal automatically sets this value based on some parameters, but if you need to manually make changes you can.

By default Reveal will include any interview that received user input while the project is in collecting mode. Interviews taken in preview mode are not included in analysis (since they’re usually for testing purposes). However, sometimes you might forget to launch the project and all your responses are marked “Preview”, or maybe you have some responses you’d like to disqualify based on some outside metrics of your own.

You can select interviews manually, or by tracking id if you are utilizing that feature. Once selected you can mark them to be included or excluded from future analysis runs.