Insights dataset dictionary

When you create a new analysis in Insights, one of the first steps is to select a dataset. Datasets are collections of related information. The datasets available to you vary based on the Provider products your organization has purchased. This document describes the different datasets you can choose from and defines the different fields contained in each set.

For instructions on creating an analysis, go to Start an analysis in Insights.

Prerequisites

  • Your organization has Provider Insights.
  • You have the Insights Author license.
  • You have the following permissions:

    • View Content
    • Create Content
    • At least one View Data permission

Page location

Insights > Datasets

Different types of data

Insights contains different types of data, such as names, numbers, dates, and more. When you use a dataset in an analysis, each data field has an icon next to it that indicates which type of data it is. The table below defines each icon.

Icon Definition
Text data. For example, Contact Start Timezone, Full Name, or Absence Type.
Measure data. These fields are typically numbers that you want to report on. For example, Agent Answered Volume, Abandoned Calls, or Contact Holds.
Dimension data. These fields are number data that you want to organize other numbers by. They are not typically data you want to report on for their own sake. For example, Contact ID, Approver ID, or Evaluation ID.
Date data. For example, Shift Start Date, Employment End Date, or Contact Start Time.

A calculated field. These kinds of fields are created by combining multiple fields that already exist in Insights. You probably won’t see this icon when you first start building analyses. This icon is orange if the data is a measure and blue if the data is a dimension.

Overview of datasets

Analytics datasets

Dataset Granularity Use for

Phrase Usage

One row per phrase hit

Reporting on phrase hits from transcriptions and text analytics. Use this dataset when you want to analyze patterns and trends of how phrases or phrase categories are used in contacts.

Phrase Usage - CrossTab

One row per phrase hit per coinciding phrase hit

Reporting on phrase hits from transcriptions and text analytics. Use this dataset only when you want to compare phrases and phrase categories to each other, such when trying to determine how often people said phrases in other categories in relation to the one you are studying. This dataset uses complex comparisons, and analyses that use it run more slowly than those that use the basic Phrase Usage dataset.

EXAMPLE   This dataset helps you uncover insights like this: People who used Negative Phrases like “speak to a manager” also used Leaving Phrases like “cancel my credit card” more often than Positive Phrases like “you are doing a great job.”

QM datasets

Dataset Granularity Use for
Contact Metadata One row per custom metadata field per contact

Reporting on contacts when you need to see the granular detail of all the metadata associated with each contact. This dataset also contains all of the fields in the Contacts dataset. Unless you need to see all the granular metadata, we recommend that you use the Contacts dataset for optimal processing speed.

Contacts One row per interaction contact Reporting on basic contact details such as durations, calling number, number called, and custom metadata.
Evaluation Answers One row per evaluation question response Reporting down to the individual question level of a form. This dataset is great for getting into the details of form responses or showing the performance of a single question over time. However, it might perform slowly if you try to run a report for a long time range. To report on how overall form scores trend over time, use the Evaluations dataset.
Evaluation Comments One row per evaluation comment Adding information from evaluation comments to dashboards that already show information at the individual evaluation level.
Evaluations One row per evaluation Reporting on evaluation scores where you don’t need to get more detailed than an individual evaluation. This dataset is great for trending scores rolled up to teams and groups and over longer time periods. To get into the detail of question responses, use the Evaluation Answers dataset.
People One row per person

Reporting on your organization’s hierarchy of groups, teams, and agents when you don’t need to see any other fields. The fields in this dataset appear in most of the other QM datasets too. If you need other fields, use one of the other QM datasets.

Recording Events One row per recording event. For example, pause, silence, hold, and so on.

Reporting on recording events in contacts when you need to see the granular details of every recording even for each contact.

WFM datasets

*Datasets marked with an asterisk (*) are a good place to start if you’re new to building analyses from scratch. They contain the most detailed information that is probably important to you in a WFM analysis.

Dataset Granularity Use for

WFM Adherence

One row per agent per day

Reporting on adherence levels per day when you don’t need anything more detailed than agents’ activities in a single day.

WFM Adherence Details* One row per agent per state

Comparing agents’ scheduled activities and absences to their actual activities and time in state.

WFM Agent Queue Stats One row per agent per queue per interval

Reporting on statistics that come from your ACD, broken down by agent, queue, and interval.

WFM Agent Skills

One row per agent per interval

Reporting on when an agent was assigned or unassigned a skill.

WFM Agent Stats One row per agent per interval

Reporting on statistics that come from your ACD, broken down by agent and interval.

WFM Forecast Skill vs. Schedule*

One row per skill per scenario per interval

Reporting on differences between forecast scenarios and actual schedules.

WFM Forecast Workload

One row per interval per scenario per skill per workload

Comparing different “what if” forecast scenarios to each other. This dataset does not currently contain actual statistics.

WFM Hourly Availability

One row per agent per day per scenario

Reporting on available time, scheduled time, and utilization at the daily level.

WFM People

One row per person period

Reporting on how long agents were employed in your contact center. For example, you can use this dataset to report on turnover rates for different groups or teams.

NOTE   This dataset contains WFM data only. It is not the same as the QM People datasest.

WFM Queue Stats

One row per queue per interval

Reporting on statistics that come from your ACD, broken down by queue and interval.

WFM Requests

One row per request

Reporting on the types of requests that agents submit and the status of those requests.

WFM Schedule Deviation

One row per person per interval

Reporting on scheduled statistics vs. actual schedules.

WFM Schedule Preferences

One row per person per preference per day

If your organization uses preferences, reporting on how many agents’ preferences were fulfilled and what percentage of their requests those fulfillments represent.

WFM Scheduled Agent Days

One row per person per scenario per day

Reporting on how many agents are working on a particular day and where they are located.

WFM Scheduled Agent Time* One row per agent per scenario per interval This is the primary dataset for reporting on schedule activity. Use this if you’re reporting on agent schedules, including absences, activities, overtime, or ready time.