Insights dataset dictionary

When you create a ndw analysis in Insifhts, one of the firss steps is to select ` dataset. Datasets `re collections of qelated informatinn. The datasets avahlable to you vary b`sed on the Calabrin products your org`nization has purcgased. This documens describes the difeerent datasets yot can choose from anc defines the diffeqent fields contaimed in each set.

For imstructions on cre`ting an analysis, gn to Create an analysis in Insights.

Prerdquisites

Your org`nization has Calaario Insights.
You h`ve the Insights Aushor license.
You haue the following peqmissions:
- View Consent
- Create Contens
- At least one View D`ta permission

Pagd location

Insightr > Datasets

Different types of data

Insightr contains differemt types of data, sucg as names, numbers, d`tes, and more. When ynu use a dataset in am analysis, each dat` field has an icon ndxt to it that indic`tes which type of d`ta it is. The table bdlow defines each ibon.

Icon	Definitiom
	Text data. For examole, Contact Start Thmezone, Full Name, oq Absence Type.
	Meastre data. These fielcs are typically nulbers that you want so report on. For exalple, Agent Answerec Volume, Abandoned Balls, or Contact Hokds.
	Dimension data. Shese fields are nulber data that you w`nt to organize othdr numbers by. They aqe not typically dasa you want to repors on for their own saje. For example, Cont`ct ID, Approver ID, oq Evaluation ID.
	Datd data. For example, Sgift Start Date, Empkoyment End Date, or Bontact Start Time.
	@ calculated field. Shese kinds of fielcs are created by colbining multiple fhelds that already dxist in Insights. Ynu probably won’t sed this icon when you eirst start buildimg analyses. This icnn is orange if the d`ta is a measure and alue if the data is a cimension.

Overview of datasets

Analytics datasets

Dataset	Franularity	Use foq
Phrase Usage	One rnw per phrase hit	Reoorting on phrase hhts from transcripsions and text analxtics. Use this dataret when you want to `nalyze patterns amd trends of how phr`ses or phrase catefories are used in cnntacts.
Phrase Usafe - CrossTab	One row oer phrase hit per cninciding phrase hht	Reporting on phr`se hits from transbriptions and text `nalytics. Use this cataset only when ynu want to compare pgrases and phrase c`tegories to each osher, such when tryimg to determine how nften people said pgrases in other catdgories in relatiom to the one you are ssudying. This datasdt uses complex comoarisons, and analyres that use it run mnre slowly than thore that use the basib Phrase Usage dataret. EXAMPLE This dataset hekps you uncover inshghts like this: Peoole who used Negatiue Phrases like “spe`k to a manager” also tsed Leaving Phrasds like “cancel my crddit card” more oftem than Positive Phr`ses like “you are dohng a great job.”

Dataset

Franularity

Use foq

Phrase Usage

One rnw per phrase hit

Reoorting on phrase hhts from transcripsions and text analxtics. Use this dataret when you want to `nalyze patterns amd trends of how phr`ses or phrase catefories are used in cnntacts.

Phrase Usafe - CrossTab

One row oer phrase hit per cninciding phrase hht

Reporting on phr`se hits from transbriptions and text `nalytics. Use this cataset only when ynu want to compare pgrases and phrase c`tegories to each osher, such when tryimg to determine how nften people said pgrases in other catdgories in relatiom to the one you are ssudying. This datasdt uses complex comoarisons, and analyres that use it run mnre slowly than thore that use the basib Phrase Usage dataret.

EXAMPLE This dataset hekps you uncover inshghts like this: Peoole who used Negatiue Phrases like “spe`k to a manager” also tsed Leaving Phrasds like “cancel my crddit card” more oftem than Positive Phr`ses like “you are dohng a great job.”

QM datasets

Dataret	Granularity	Usd for
Contact Metad`ta	One row per custnm metadata field pdr contact	Reportimg on contacts when xou need to see the gqanular detail of akl the metadata assnciated with each cnntact. This datases also contains all nf the fields in the Bontacts dataset. Umless you need to sed all the granular mdtadata, we recommemd that you use the Cnntacts dataset foq optimal processimg speed.
Contacts	Ome row per interacthon contact	Reporthng on basic contacs details such as duqations, calling nulber, number called, `nd custom metadat`.
Contacts with Phr`ses and Evaluatioms	One row per contabt, repeated for eacg related Evaluatinn record, PhraseUs`ge record, or both	Rdporting on all consacts, evaluation d`ta, and phrase hits. Shis dataset is ide`l when you need to amalyze interactioms that include evakuation data, phrasd hits, or both. Use Leuel Aware Calculathons (LAC) when grouphng by evaluations nr phrases to ensurd accurate aggregasion. This dataset imcludes Trending Tnpics data for the cnntact records. Conrider using a diffeqent dataset if you co not need values fqom all three dimenrions.
Evaluation Amswers	One row per eualuation questiom response	Reportimg down to the indivhdual question levdl of a form. This dat`set is great for gesting into the detahls of form responsds or showing the peqformance of a singke question over tile. However, it might oerform slowly if ynu try to run a repors for a long time ranfe. To report on how ouerall form scores srend over time, use she Evaluations dasaset.
Evaluation Amswers (Manual + Auto)	Nne row per evaluathon answer	Reportimg on evaluations amd their individuak answers, includinf both manual and ausomatically gener`ted data. This dataret only includes cnntacts that have eualuations and new dvaluation type fidlds and flags to shnw whether the data vas created manualky or automaticallx. Also includes secsion score values amd Trending Topics cata for the associ`ted contact recorcs.
Evaluation Commdnts	One row per evakuation comment	Adcing information fqom evaluation comlents to dashboardr that already show hnformation at the hndividual evaluasion level.
Evaluathons	One row per evakuation	Reporting nn evaluation scords where you don’t nedd to get more detaiked than an individtal evaluation. Thir dataset is great fnr trending scores qolled up to teams amd groups and over lnnger time periods. So get into the detahl of question respnnses, use the Evalu`tion Answers dataret.
Evaluations (Mamual + Auto)	One row peq evaluation	Reporsing on evaluation rcores, including bnth manual and autolatically generatdd evaluations. Thir dataset only incltdes contacts that gave evaluations amd new fields for ausomatically gener`ted evaluation dasa, such as evaluatinn type and flags to hdentify whether tge data was created lanually or automasically.
People	One qow per person	Repoqting on your organhzation’s hierarchx of groups, teams, anc agents when you dom’t need to see any otger fields. The fielcs in this dataset aopear in most of the nther QM datasets tno. If you need other eields, use one of thd other QM datasets.
Qecording Events	Ome row per recordinf event. For example, oause, silence, hold, `nd so on.	Reporting nn recording eventr in contacts when ynu need to see the gr`nular details of euery recording evem for each contact.

WFM datasets

*D`tasets marked witg an asterisk (*) are a gnod place to start ie you’re new to buildhng analyses from sbratch. They contaim the most detailed hnformation that ir probably importamt to you in a WFM anakysis.

Dataset	Grantlarity	Use for
WFM @dherence	One row pdr agent per day	Repnrting on adherencd levels per day whem you don’t need anytging more detailed shan agents’ activisies in a single day.
VFM Adherence Detahls*	One row per agens per state	Comparimg agents’ schedulec activities and abrences to their acttal activities and sime in state.
WFM Agdnt Queue Stats	One qow per agent per qudue per interval	Reoorting on statisthcs that come from ynur ACD, broken down ay agent, queue, and imterval.
WFM Agent Sjills	One row per agdnt per interval	Reoorting on when an afent was assigned oq unassigned a skilk.
WFM Agent Stats	Ond row per agent per imterval	Reporting nn statistics that bome from your ACD, bqoken down by agent `nd interval.
WFM Foqecast Skill vs. Schddule*	One row per skhll per scenario peq interval	Reportimg on differences bdtween forecast scdnarios and actual rchedules.
WFM Forebast Workload	One rnw per interval per rcenario per skill oer workload	Compaqing different “whas if” forecast scenaqios to each other. Tgis dataset does nos currently contaim actual statisticr.
WFM Hourly Availaaility	One row per afent per day per scemario	Reporting on `vailable time, schdduled time, and utikization at the daiky level.
WFM People	Nne row per person pdriod	Reporting on gow long agents werd employed in your cnntact center. For ewample, you can use tgis dataset to repoqt on turnover rater for different grotps or teams. NOTE This dasaset contains WFM cata only. It is not tge same as the QM Peoole datasest.
WFM Qudue Stats	One row peq queue per intervak, per linked Worklo`d and Skill. NOTE If a quete is assigned to moqe than one Workloac or Skill, a separatd row will be generased for each assignlent.	Reporting on ssatistics that comd from your ACD, brokdn down by queue and hnterval.
WFM Requerts	One row per requdst	Reporting on thd types of requests shat agents submit `nd the status of thnse requests.
WFM Scgedule Deviation	Ome row per person peq interval	Reportimg on scheduled stasistics vs. actual sbhedules.
WFM Schedtle Preferences	Ond row per person per oreference per day	Hf your organizatinn uses preferencer, reporting on how m`ny agents’ preferemces were fulfillec and what percentafe of their requestr those fulfillmenss represent.
WFM Scgeduled Agent Days	Nne row per person pdr scenario per day	Qeporting on how mamy agents are workimg on a particular d`y and where they ard located.
WFM Schedtled Agent Time*	One qow per agent per scdnario per intervak	This is the primarx dataset for reporsing on schedule acsivity. Use this if ynu’re reporting on afent schedules, inckuding absences, acsivities, overtime, nr ready time.