HomeMy WebLinkAboutSBStat 2023 - Customer Service Q2
SBStat | Customer Service
2023 Quarter 2 | Thursday, July 13, 2023
City of South Bend
1
Empower
departments
Design
services
Focus on
data
Improve
residents’ lives
The purpose of SB Stat is to bring the most powerful people in the City – the Mayor, Department Heads, and key staff – into a room to use data and take action on some of the City’s most
pressing challenges
2
Customer Service Portfolio Summary
Check CS projects in DR excel
3
Today’s Agenda
4
Customer Service KPI Summary
Type of performance measure reference - https://www.mtas.tennessee.edu/reference/types-performance-measures
5
Customer Service KPIs
Q2 2023
6
Customer Service KPIs – Equity & Access (draft)
Q4 2022 to Q1 2023
By poverty level
0-10% persons in poverty
11-50% persons in poverty
51-100% persons in poverty
https://www.census.gov/quickfacts/fact/table/southbendcityindiana/AGE295221
7
KPI Review
Percent of calls opted into survey has returned to 2022 levels
Steps taken to improve performance:
Recurring meetings to discuss previous month’s post call survey data.
Reminded 311 and Water Works customer service staff to redirect residents to surveys.
Edited 311 welcome message to inform callers that they will not be placed out of the queue if they decide to opt-in to the survey.
Additional steps to take:
Add voicemail option to survey
Manual transfer of residents to survey
Train Morris Box Office staff to redirect callers to survey
Ask Shawn, Cathleen for their thoughts
8
CoSB Post Call Survey Results | 311
Refreshed as of 3/29/2023
9
CoSB Post Call Survey Results | 311
10
CoSB Post Call Survey Results | Utilities
Similar but Timeliness is slightly lower than 311
11
CoSB Post Call Survey Results | Utilities
Averages increased over last 6 months
12
CoSB Post Call Survey Results | Morris
13
Investigating factors behind lower-than-expected timeliness ratings
14
Extra trash pickup, utility account update, and litter services are seeing disproportionally more low timeliness ratings.
There is a greater proportion of low timeliness ratings (1,2,3 or 4s) for the highlighted articles compared to the corresponding overall customer satisfaction rating
Initial thoughts: Some respondents are tying service request completion time or call outcome dissatisfaction to timeliness of the call– rating it lower than intended otherwise. This
can still be useful to diagnose process inefficiencies.
Extra pick up, Account balance, and Update utility account information are the most frequent call topics (17% of all calls) among callers who call 10 or more times.
Discussion:
How can we best ensure callers are satisfied for the listed low timeliness services?
Is this something we can improve customer service teams or the business units?
15
Proposed Action Plan
Findings
Possible action plan
People who provide low ratings for one question are likely to provide low ratings for all the questions
Average call time of resident rating 1 is 92 seconds longer than resident rating a 5 in the first half of 2023
Timeliness rating decreased partially due to additional 311 prepend announcements.
(Total prepend time until caller was placed on queue was around 90 seconds in May/June.)
Implement voicemail box for post-call surveys to better understand caller frustrations.
Ideally, reduce call times for articles with low timeliness ratings (Ideally 6 minutes).
Add queue time to customer service KPIs and future analyses
Limit usage of 311 prepend announcements
311 welcome message should be placed before all other announcements.
Total prepended announcements should not exceed 60 seconds.
Any additional messages should be relegated to hold time.
1st item: Probably due to dissatisfaction with the entire call
Also although this metric is to measure queue time, we see it correlate with other call factors such as call time in general and by service request
16
Which services are repeat callers requesting?
17
22% of calls are from one-time or anonymous callers
Surveys from each caller frequency types are roughly represented except for anonymous/blank.
Low variance of customer satisfaction and timeliness ratings by call frequency (4.43 – High freq avg, 4.58 – One-time avg).
18
Residents are repeat calling “public” service requests
Which services are receiving the highest proportion of repeat callers?
Neighborhood service requests e.g. Loose dog, Alley maintenance, NSE structure violations.
Exceptions include Compost Availability, and engineering concerns.
Program and bill inquiries and “directly impacting” service requests such as trash pickup and utilities have little or no repeat callers.
Residents that file service requests that are not a burden on a single individual/household have trouble tracking work that is rectifying reported issue.
311 Alley grading call note example. 60% of alley calls are repeat.
Note: Roughly 50% of calls received on CISCO since April 2022 don’t have an attached KBA either because the call was from an anonymous caller or inconsistency with joining data between
CRM and UCCX.
19
Recommendations
311 operators could record how many service request look ups are requested by residents on CRM
Build awareness of service tracking once it is live through 311
Resident who filed a service request, which includes sensitive information can be notified via email or phone
Conduct process mapping sessions for NSE violation disputes and identify possible process improvements
Improve data quality between CRM and CISCO databases through a shared key
Build and publish a public tool for residents to look up non-sensitive service requests i.e. Pizza Tracker
20
Does overall customer satisfaction vary by KBA?
21
Residents record a lower satisfaction rating for billing-type calls, but can resolve the issue
Yes: variance of KBA customer satisfaction with more than 10 surveys: range from 3.83 to 5.0.
Eleven of the sixteen of the KBAs with <4.5 customer satisfaction have 85%+ resolution rate.
22
Deep-dive analysis & discussion
Diving deep into a few key initiatives being undertaken to improve city performance
Service Request Dashboard
Improving Solid Waste and 311 Coordination
23
Service Request Dashboard
Daily performance data on City service requests
Problem analysis+ process improvement track
24
Service Request Dashboard
Problem Statement
City employees do not have easy access to performance data on service requests from residents. There is missing data with the way City teams review service request processes and for
311 to track service request progress. These issues make it challenging to improve the customer experience for residents.
Dashboard Goals
Make service request performance data accessible so teams can make process improvements
Track service request statuses from a central location
Improve data quality
Sufficient data quality to launch CRM notifications for residents
Outcome Metric(s)
Number of business units utilizing data routine meetings
Number of service requests meeting SLAs
Percent of service requests statuses that are assigned correctly
Background: Delivery Associates for the last few months to build this dashboard. Business units, Mayor’s office
25
Discussion questions to be thinking about
We presented a new service request dashboard for City teams and a monthly data routine process –
What barriers do you see for your teams using this dashboard?
Any immediate services or research questions we should evaluate?
26
Background
A total of 58 eligible service requests for tracking
Not instant and SLA time is less than a week
The city is accountable for providing the service
Nine high priority service requests:
Significant request volume and resident interest
All listed service requests have a work order management system in use
Two data limitations are incomplete service statuses and incomplete breakdown of general service requests
We rated the current state of data quality in CRM on previously mentioned metrics:
% of service requests that are closed
Data quality of request status types that are aren’t closed:
Very Good (Detailed and accurate) is benchmark for public SR tracking-> Detailed enough so that the resident is informed of their service request AND instantly marked so that the resident
knows the request status is accurate
27
Top service requests’ current data limitations
Requires breakdown by KBA
28
Dashboard Walkthrough
Points to cover:
It is found on the Innovation and Technology Power BI workspace. If you need access to it, the link will be in the follow-up memo and if there are access issues please reach out to David
Finley
Please note that this dashboard is internal.
Data in reporting server and dashboard is refreshed every four hours
Header and Filters: Identical for each tab on the dashboard
Use drop downs to filter by status, KBA, team or date
Overview tab purpose: Useful for summary data for a service request type
You can edit the census map to show different socioeconomic indicators
Example usage: Show data quality difference between pothole and Solid Waste and how this could be useful for each division or department head
Individual Request tab purpose: Can be used when investigating a specific case. Let’s say Mayor’s office receives a complaint from resident. A DH can easily check the dashboard to view
call notes without having to consult crew or division heads. And can see if there is a pattern emerging from call notes
Can filter by service ID
Service type tab: Allows the user to compare overall service request completion rates, SLAs and more with other service requests. At a high level, it can help identify service requests
that are not meeting SLAs
Data Dictionary: You can use this tab if there is any confusion on jargon within the dashboard
29
Theory of Change
Monthly Performance Data Routine
30
Example Agenda
How many SLAs are meeting their 80% target?
Which SLAs are consistently exceeding their targets?
Discussion questions
Notes
Xxx
Reference Docs
Progress milestones
Action items
David and Becky work on creating clear service request definitions
David, Denise, and Max work on system integrations
Xxx
Delivery Associates’ example data routine template for teams to use
Monthly data routines can be 30 to 60 minutes long
Does not have to be a standalone meeting – can be supplemented to standard check-ins between department and division leaders
Next Steps
Project Timeline
32
Taking Action
33
Discussion questions before closing out the topic
We presented a new service request dashboard for City teams –
What barriers do you see for your teams using this dashboard?
Any immediate services or research questions we should evaluate?
34
Performance Improvement --> Streamlining Solid Waste & 311 Service Request Process
35
Defining the problem
Problem Statement
Service requests from residents that are made directly to the Solid Waste division are responded to more quickly than service requests made by residents to 311. This is a problem for
residents because there is inequity in how quickly the City is responding to their service request.
How might we…
…improve our process so residents do not experience inequity?
…improve coordination between 311 and Solid Waste?
Outcome Metric(s)
Average number of days to complete Solid Waste service requests made via 311
36
Background
Residents currently have multiple ways to request a service from the Solid Waste team, (1) through 311 via CRM and (2) through a Microsoft Form
37
Data Analysis
Work Orders from 311 charge codes take longer than those from non-311
charge codes
Data insights
All 311 charge codes, except for “311- Xmas Tree Removal”, include more work orders than their non-311 counterpart charge code.
Work Orders from 311-related charge codes take significantly longer than work orders from non-311 charge codes.
Largest difference of 4.5 days between Appliance (13.1 days) and 311 Appliance (17.6 days) Work Orders
Hypothesis
Work Orders created via 311 calls may involve additional steps that delay work orders from being finished.
38
Theory of Change
39
Taking Action
40
Celebrating our values
This section highlights exemplary work happening in the City to improve performance that may otherwise go unnoticed
41
Celebrating our Values
First monthly data routines with all City call liaisons with 311, Utility, Morris
-->
Post call survey dashboard
42
SBStat | Customer Service
2023 Quarter 2 | Thursday, July 13, 2023
City of South Bend
43
Appendix section
44
22% of calls are from one-time or anonymous callers
Surveys from each caller frequency types are roughly represented except for anonymous/blank.
Low variance of customer satisfaction and timeliness ratings by call frequency (4.43 – High freq avg, 4.58 – One-time avg).
45
Share of service requests by medium to high volume repeat callers
On average, how many days back did a repeat caller last call 311 regarding the same article? 28 days
There might be inconsistencies in the data or callers misinterpreted the survey question
46
Breakdown of Q1 and Q4 ratings by caller type
47
Data routine folder
Data routine materials
Data routine template
Data routine tracker and process improvement idea list
48
DA.Digital suggests the City incorporate a new routine to review the dashboard. If the City does schedule a new routine, keep the following questions in mind:
Recommendations /
After researching best practices and analyzing initial dashboard data, DA.Digital has seven recommendations for the City to consider:
Recommendations
Prioritize addressing street light issues along Miami St. between St. Joseph River and US-20
Prioritize addressing general service (code enforcement) service requests in the southwest area
Analyze data at the service type and individual service request level to find patterns within specific departments
Provide additional CRM training for staff
Invest in three core skill sets (data visualization, data architecture and engineering, and operations) to maintain the dashboard and support the data system
Add logic to address “on hold” and “external time” in a future dashboard update
Embed the dashboard in routine meetings or develop a routine specifically for dashboard review
Develop a public facing version of the dashboard
Recommendations /
Delivery Associates recommendation slides
DA.Digital recommends prioritizing street light issues along Miami Street between St. Joseph River and US-20 as there could be a correlation with crime incidents
What does the data show?
A high concentration of service requests (22) for street light issues have been created in this area since January 1, 2023
9 of 22 (41%) of street light-related service requests are still open
All 13 completed requests were finished “Very late (>40% over)” and took an average of 472 hours (20 days) to complete vs 402 hours (17 days) for South Bend
Data from the Incidents + Crime Dashboard also shows a concentration of incidents reported in the same area
SLA Dashboard
Incidents + Crime Dashboard
Recommendations /
DA.Digital recommends addressing “General Service (Code Enforcement)” service requests in the southwest area as the region could be experiencing an inequitable level of service
What does the data show?
A large concentration of service requests categorized as the “General Service (Code Enforcement)” in the southwest area (highlighted in green)
34 of 62 (55%) General Service Code Enforcement service requests were completed very late ( > 40% of their expected SLA time)
45% of all General Service Code Enforcement service requests in this green area met their SLA vs 55% for all of South Bend
33 of 62 (53%) service requests are “Neighborhood Parking Guidelines” service type
The green area covers Census Tracts 22, 24, and 26
The average estimated 2021 Census median household income of this green area is $37,720 which is lower than the South Bend median household income of $46,002
Recommendations /
DA.Digital recommends analyzing data at the service type and individual service request level to find patterns within specific departments
What are the benefits?
Avoid comparisons between different departments since they have different processes
Allow departments to identify their own strengths and weaknesses
Create opportunities for departments to provide additional context to their performance for each Service Type and Knowledge Article
Provide a detailed analysis when viewing individual service requests
Help find patterns of quickly closed or overdue service requests
When selecting KPIs for data analysis at the service type and individual service request level, what framework should you use?
Analyze and prioritize data using the 3M framework. Is the KPI:
Meaningful for residents, departments, and leadership?
Measureable by the dashboard and other supporting data?
Movable the city has control of the process and its weakness?
Recommendations /
DA.Digital recommends that South Bend provides additional training for staff to support consistent use of the service requests. DA.Digital has developed an activity plan to help with
CRM training which has been shared
What is an activity plan?
An activity plan is a support material for a training, workshop, or important meeting. It is useful to make sure that the planner has considered all the relevant parts of the meeting
so you achieve its objectives.
What does it have?
A list of identified participants
Established objectives to meet the needs of the participants and get them to intended outcomes
An outline of interactive activities to help participants achieve outcomes
List of necessary support materials for the activities
A refined agenda
For more information on activity plans and how to make them, please refer to the slides in the appendix.
Recommendations /
Delivery Associates’ recommendations on when teams should update service request statuses:
55
Excel tables
Customer Service Stat Q2 tables.xlsx
Repeat callers by KBA [tab 1]
56