Extensive Survey Results in a Power BI report, allowing you to slice and dice survey data
Jump to Overview | Samples | Variables | Alert Functionality | Power BI Reporting | Related Alerts | Glossary of Terms | Webinar
Overview
Overview
We have a survey dashboard embedded in our website once you log in; however if you really want to slice & dice survey responses, then our Power BI Dashboard version is the way to go. You can create an Executive Dashboard that includes survey results along with other financial benchmarks.
* * *
Sample
Sample
* * *
Variables
Variables
N/A
* * *
Alert Functionality
Alert Functionality
-See here for getting Power BI in place.
-Click here to download the file
For ECi Hosted clients: We are working with ECi to open up our access so we can set an ODBC connection for you. If you are interested, then please provide us the public static IP address for your company network so that ECi can whitelist it to allow us access.
* * *
Power BI Reporting
Power BI Reporting
Use the (1) Reporting Period date slider in the top-right corner to change date filters and/or (2) the Filters button to open the filters panel. Every page header also shows the Last Refresh timestamp, the Reporting period (start, end and number of days) and In view (surveys sent and submitted under the current filters):
Filter options here:
Insights panel - Click Insights in the top-right corner of any page to slide out that page’s own commentary. The panel has four blocks:
- How to read this page – the interpretation rules for that page.
- Key take-aways – the current numbers read back as narrative.
- Flags – signals that need attention, such as a falling response rate, a small sample or hidden detractors.
- Action items – what to do next.
Take-aways, Flags and Action items are calculated from the data in view, so they rewrite themselves whenever a slicer or filter changes.
Drill-through: Survey Details - Pages that list customers, techs, dispatchers, equipment or dimensions carry Surveys in period and Unsubmitted Surveys buttons. Select a row, then click a button to open the Survey Details drill-through for that selection. The drill-through keeps every filter from the page you came from; the back arrow returns you to it.
- Surveys in period – cards for Surveys listed, Customers and NPS on these surveys, then every answer with its question, comment, sentiment score and the sentiment terms that drove the score.
- Unsubmitted Surveys – every survey sent but not submitted in the period, with survey type and title, authentication ID, call number, account, contact and contact email.
View NPS / Tech rating / Dispatch Rating - Month over Month Trends
Breakdown three key sections
Example of Review Tech Rating optional filters to narrow down by Branch and a Make as example
Export results to excel option
View All results by Customer (all filter options available)
Analyze results by a specific question
Average score for a question
Highlight specific question to view those results only
Group Survey Metrics by various dimensions
Example below looking at results broken out by branch, then broken out by Contract Type
Search for key word in Comments submitted by Customers
Executive Overview - Ratings, response health, sentiment and survey coverage in one screen
Gives leadership a single-screen read on customer satisfaction — NPS, Rate Tech, Rate Dispatch and comment sentiment — together with response health and CM-call survey coverage, and shows what changed against the prior period. In under a minute a leadership reader knows whether satisfaction is improving or slipping, whether the ratings rest on enough responses to trust, and which detractors, techs, branches and survey-coverage gaps need follow-up first.
- The six cards compare the reporting period with the same number of days immediately before it. Green arrows are improvements, red are deteriorations.
- The NPS chart shows month-on-month NPS (solid) against the 12-month rolling NPS (dashed).
- Signals are computed from the data in view and ranked by how much attention they need. Click a branch, a tech or a customer to cross-filter the rest of the page.
The page has three tabs: Scorecard (six KPI cards, NPS and Response Rate trends, Signals that need a decision, Branch and Tech scorecards, and the detractors and passives to follow up), Period comparison (every headline metric against the prior period, with month-on-month vs 12-month rolling trends) and Where to act (detractors and passives to call back, CM calls that did not fire a survey grouped by reason, and the themes behind negative comments).
CM Calls Without a Survey - Corrective-maintenance calls that never fired a survey, and why
Finds the corrective-maintenance (CM) service calls that closed without firing a Service Call survey, and explains why each one was excluded, so you can separate exclusions that are working as designed from fixable gaps and correct the survey trigger rules.
- A CM call should fire a Service Call survey unless a rule excludes it. Reasons are derived from call status and the location’s survey history.
- Survey filters do not apply to this page. Call filters (call type, tech, dispatcher, SLA) do.
- CM calls are loaded for roughly the last 60 days only. The window is stated in the page header, and coverage is only measured inside it.
- What to do about each reason marks every reason as By design or Fixable and gives the action to take.
- Coverage: failed SLA vs within compares survey coverage on calls that failed SLA with calls inside SLA. If failed-SLA calls are surveyed less, NPS is biased upward.
The page has two tabs: Overview (coverage KPIs, the reasons, weekly coverage, and breakdowns by call type, tech and location) and Call list (every CM call without a survey, with the no-survey reason and days since the location was last surveyed).
Comment Sentiment - A lexicon-based sentiment score on every comment
Scores every survey comment as positive, neutral or negative, and shows the terms and themes that drive the score: whether comments are getting more or less positive, which themes drive negative comments, which techs attract them, and which promoters are hidden detractors that should be called back.
- Each comment is scored with the 113-term Sentiment Lexicon. A “not” or “never” before a positive term flips it.
- Comments with no matching term are “No signal” and are excluded from averages.
- Hidden detractors are promoters (NPS 9–10) who left a negative comment. Treat them as detractors and add them to the call-back list.
- Extend the Sentiment Lexicon table with your own product and service terms; scores update on refresh.
The page has two tabs: Overview (average and net sentiment, distribution, monthly trend, sentiment by NPS rank, top positive and negative terms, negative themes and sentiment by tech) and Comments (every scored comment, filterable by Sentiment and by question).
Key Influencers - Net Promoter Score, Rate Tech and Rate Dispatch
Find the KPIs and dimensions that make an NPS answer more likely to be a promoter, passive or detractor (or a Tech / Dispatch rating more likely to be 10, 8 to 9, or 0 to 7), and see how those influencers shift between periods. Each of the three pages has two tabs, and the Key influencers visual has its own Top segments view:
- The Key influencers tab uses the Power BI Key influencers visual over one row per survey. Pick an outcome to see which conditions make it more likely, and by how much.
- Top segments shows the clusters: combinations of conditions in which the outcome is most likely, with the population of each.
- By reporting period compares every condition between the reporting period and the prior period. KI Status reads New, Strengthening, Stable, Weakening or Faded.
- A lift shown as ∞ means nobody outside the group reached the outcome. That is a small-sample artefact, not a law.
- With fewer than about 60 responses, influencers below 1.3x are within normal noise.
Net Promoter Score
Rate Tech
Rate Dispatch
* * *
Glossary
Glossary of Terms
SCR - Survey Completeness - Quantity of surveys where all questions answered on the survey
QRR - Question Response Rate - Quantity of questions answered against total questions on all surveys
VR - Valid Responses - Quantity of times you are rated between 0 - 10 on the NPS question (as opposed to a blank entry on the NPS question)
Sentiment Score - Each comment is scored against a 113-term Sentiment Lexicon. A “not” or “never” before a positive term flips it. Comments with no matching term are “No signal” and are excluded from sentiment averages.
Hidden detractors - Promoters (NPS 9–10) who left a negative comment — the score says happy, the words do not.
* * *
Related Alerts
Related Alerts
ID109 - Service Call Surveys
ID220 - Supply Order Surveys
ID312 - New Equipment Install Surveys
See full list of all Power BI reports that we currently have available
https://support.ceojuice.com/hc/en-us/sections/360011259132-Power-BI-Reports
* * *
Webinar
Webinar
0 Comments