The URL can be used to link to this page

Home My WebLink AboutSBStat 2023 - Customer Service Q3 SBStat | Customer Service 2023 Quarter 3 | Monday, November 6, 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 Q3 + October 2023 6 CoSB Post Call Survey Results | Summary Refreshed as of 3/29/2023 7 High response rate last quarterPercent of calls lower in August due to Build the Budget Steps taken to improve performance: Call monitoring More 311 liaisons are transferring residents to the survey than last quarter Implemented optional voicemail at end of survey Ongoing low-call rating analysis Additional steps to take: Continue to train newly hired customer service liaisons for Ask Shawn, Cathleen for their thoughts 8 Low-rated Call Follow-ups Initial Findings Goal: Better understanding of underlying reasons behind low ratings i.e. qualitative analysis through follow-up surveying Methodology: Gather list of phone numbers by each category for each week (through the dashboard), and ensuring there is no duplicate calls For the last four weeks, two completed survey calls for each of the following categories: General low customer satisfaction rating Low timeliness rating Unresolved call rating 9 Low-rated Call Follow-ups Initial Findings Initial Findings (15 surveys) 7 customers either claimed they had a good experience with their 311 call or were hesitant to report any issues with City customer service A mix of low timeliness ratings centered around timeliness of City unit responding to an ongoing service request and queuing for call Other feedback: A resident had technical issues reaching 311 within City limits A resident requested for street resurfacing, but potholes were not solving underlying street issue (1600 Jackson block) 10 Deep-dive analysis & discussion Diving deep into a few key initiatives being undertaken to improve city performance 311 Service Portal Request Analysis Streets SLA Performance Analysis 11 311 Service Portal Service Request Analysis 12 Problem Definition Problem Statement The City needs to engage with residents digitally. Residents have a preference to request services and engage with content online without necessarily having to speak with a person or during business hours. The City has not formally marketed the online 311 Service Portal or investigated current 311 Portal usage. How might we… Inform residents of the 311 Portal Increase Service Requests originating from the Portal Decrease number of calls to 311 Outcome Metric(s) Number of Portal Service Requests Number of 311 calls % of Service Requests from Portal from each Council District 13 Top services requested online 1. Christmas Tree Pickup 1347 portal requests, 9 phone requests 43% of all portal requests Accounts for bump in portal requests in Q1 and partly in Q4 for last two years Promoted on City Facebook in late December, January 2. Pothole in Street 794 portal requests, 1627 phone requests Seasonal: requests coming largely in Q1,Q2 Popular requests for both options, but most people still call to report 3. Traffic Concern Form 726 portal requests, 604 phone requests Jump in 2023 Q2 after speed humps launch Promoted on City Facebook for residents to fill out Confusion with Traffic Enforcement form 14 When do residents use the service portal? 311 operating hours Total Portal Service Requests by Hour (Eastern Time) Total Portal Service Requests by Day of the Week Residents mainly request services during business hours Drop off in requests after hours Not able to see correlation between types of service requested between business hours and after hours Portal requests are down by roughly 30% on weekends 15 Requests coming from Portal are concentrated in Council District 3 and 4 Location of service requests by origin, Jan 2021-Oct 2023 Blue (low frequency) -> Red (high frequency). Note: marks where the City team is directed to go and not address of requester. Portal Phone 16 Feedback from Service Portal Residents can currently submit feedback for the service portal on a Microsoft Form Most complaints center on difficulty of navigating the portal i.e. can’t find right article, contact, or request form Some complaints concern inaccurate information in articles e.g. links are broken, new version of article has not been approved 17 Methodology Each time a form is filled out in CRM by a liaison or on the service portal by a resident, the request origin is recorded. Timeframe of analysis: Jan 2021-Oct 2023 Using this view in our reporting server for our analysis, we can see: Request origin Service form used in CRM Generated service request id When it was created 18 Findings 311 Portal use has grown slightly over last three years As of October, 9% of service requests originate from the Portal in 2023 Q1 sees highest portion of service requests originating from the portal 33 service forms are available externally for residents, but vast majority are for three types services. Count of Service Requests by origin Proportion of Service Requests by origin for each quarter Q1 Portal Request Benchmarks 2022 Q1: 833 (17%) 2023 Q1: 802 (15%) 19 Key Takeaways The 311 Portal is still being used by some residents even without extensive promotion. From preliminary data, proportion of 311 portal usage is lower in West Side. Many service requests that are available in the 311 portal are not being used by the public. The 311 Portal has potential for residents to file requests after hours. Potential Performance Targets for 2024 Q1: 20 Best Practices Review Facebook to highlight popular services that can be requested by residents (Christmas Tree Pickup) Boston residents can download the Boston 311 App to report and track service requests on mobile Idea: Assess percentage of residents who would request services on their phone Fort Wayne categorizes service request types visually without having to scroll down to find Screenshot from a City facebook post last year directing residents to fill out the form on the 311 Portal 21 Proposed action plan Observations Possible Actions Potential for public to use 311 Portal more Set Performance Targets for (5-20% increase in portal service requests) Use Analytics to monitor online service portal usage (CRM request totals, Google Analytics) Success of Christmas Tree Pickup Portal form and identification of areas with low Portal usage Create marketing plan for 311 Portal Identify candidate service requests to be made into public form (e.g. sewer back up) Residents using the Portal after hours 22 Taking Action 23 Streets SLA performance analysis Review of performance data captured from CityWorks and CRM for Potholes and Alleys 24 Problem Definition Problem Statement The City has not evaluated Streets service request performance using CityWorks data before. This is a problem for City leadership since the Streets team primarily uses Cityworks to close out work orders and does not report performance data into the 311 CRM, so it is not clear if the City has good performance targets (SLAs) for priority service requests like filling a pothole or alley maintenance. How might we… Set feasible performance targets for filling potholes and alley maintenance? Identify areas of improvement for Streets service requests? Outcome Metric(s) Higher percent of service requests completed on time Less unresolved service requests 25 Why this matters For residents, resurfacing and repairing roads is the top priority in Build the Budget online survey and voicemails. ‘Pothole in Street’ and ‘Alley Maintenance Request’ service forms were filled out 840 and 601 times, respectively, in a 12-month period Two of the top fifteen services requested. Both service requests are integrated with CityWorks. 26 Pothole/Alley Current State Simplified Process Map Definitions: Service Request (SR): Request from resident for a City service Work Order (WO): Authorization for a crew to be sent to complete work or respond to request Today’s analysis uses the timestamps coming from CRM and Cityworks joined data in the reporting server Calculated completion time = Cityworks WO Completion timestamp – CRM service request creation timestamp We are not using actual CityWorks SR completion since a supervisor inspects the streets days or weeks after the work was completed. Time frame: January 1, 2022 -> October 30, 2023 with focus on Spring (Mar-Apr) and Fall (Aug-Sep) high-activity months. Analysis done in Python/Pandas and Excel Mapping done in Power BI Maps 27 Overview of Completion Data Pothole requests created and completed Alleys requests created and completed 74.8% Actual Completion Rate 87.1% Adjusted Completion Rate Actual: Percent completed/canceled out of all CRM requests. Adjusted: Percent completed out of requests in CityWorks. Higher due to a handful of CRM->Cityworks integration failures that are being addressed. 28 Time to Complete Request 27 days to complete a pothole request on average 60 days to complete an alley request on average Factors causing low timeliness: Requests made during off season accumulate We can see that potholes are addressed in significantly less time during major patching weeks in Fall and Spring Invalid requests are not closed immediately Request time outliers pushed the mean time Pothole median: 13 days, Alley median: 53 days Increased frequency of pothole requests in Summer 2023 were not addressed until late Fall For Alleys, major spikes in requests in spring don’t get addressed until August or July Not included: Streets repairs many potholes not reported via 311. Avg. Number of days to complete a pothole request by month 29 On-time percentage given current SLA framework 13% Pothole requests completed within 48 hours 4% Alley requests completed within 48 hours Unsurprisingly, most pothole requests completed within 48 hours are for potholes reported on arterial roads Discussion: Does Public Works or other departments use different benchmarks for pothole completion time depending on street type? Streets Service Requests by Cityworks Completion Time Category See Appendix for how each category is defined. 30 Increased Pothole requests in Summer 2023 Clusters of pothole reports along certain neighborhoods/roads: Far Northwest Neighborhood centered around College St. McKinley Ave between Ironwood and Twickenham River Park neighborhood Erskine Park Frequency and length of time of reported Potholes, May-July Red = More reports and time request was left open 31 32 Work order timeline takeaways SRs outside city limits were not closed (Most coming from Portal) Many duplicate service requests appeared for arterial intersections. During non-peak seasons if there is availability, Streets would select a handful of potholes to be repaired for that given day. During peak seasons, all eligible potholes were put to work orders at once and crews would go through each designated area over the course of a week or two 33 Proposed Process Improvements for Reporting 34 Taking Action 35 III. Changes to SBStat in 2024Starting in 2024, we'll be taking a less meeting-centric approach to performance management. We're going from this... Regular, quarterly meetings with a big group Focus on cross-departmental performance management projects directed by the Mayor's Office Work delivered and championed by I&T Overall, very strategically focused performance management program ...to this Automated email-based dashboards, reports (with an option to meet if needed) Mix of Mayoral projects and Department-specific projects Work delivered by I&T and championed by the Department Overall, more operationally focused performance management program 36 Why the Change? Efficiency. Less meetings! Less Decks! Less scheduling, more services! Better fit for our data culture. Unlike other cities, our data culture is more democratized across departments and at every level of government and we want the program to mirror and respect that. These conversations don't just happen on the 14th floor of CCB. 37 We'll do more of this >>> The UAP Dashboard is a monthly email that goes to a cross-section of Public Works, A&F, Mayor's Office and I&T to monitor the health of UAP and utility delinquencies. When delinquencies go too high or the UAP program gets close to red, we trigger a meeting. 38 Next Steps Department and Division Heads should think about your Department-specific goals for next year. What data and teams do you want to focus on? What change do you or one of you Directors/managers want to champion? We can give you the templates/maps/reporting you need to drive that change. 39 Celebrating our values This section highlights exemplary work happening in the City to improve performance that may otherwise go unnoticed 40 Celebrating our Values Increase in post-call survey uptake to beyond pre-Build the Budget levels! Kudos to Streets, Sewers, and Engineering teams for their willingness to engage in data routine meetings to improve City services reporting! 41 SBStat | Customer Service 2023 Quarter 3 | Monday, November 6, 2023 City of South Bend 42 Appendix section - Post-call survey screenshots -Streets analysis code 43 CoSB Post Call Survey Results | 311 Refreshed as of 3/29/2023 44 CoSB Post Call Survey Results | 311 45 CoSB Post Call Survey Results | Utilities Similar but Timeliness is slightly lower than 311 46 CoSB Post Call Survey Results | Utilities Averages increased over last 6 months 47 CoSB Post Call Survey Results | Morris 48 Completion Time SLA Categories When a service request has been completed, the completion time as a percentage of the stated SLA determines the time category: Very Early: 0-60% of SLA Time Early: 61-80% of SLA Time On-Time: 81-120% of SLA Time Late: 121-140% of SLA Time By default, a service request has an SLA time of 48 hours to complete. Example: If an abandoned vehicle service request is completed in 60 hours, it is classified as a Late service request (60 is 125% more than 48 hours). 49 Streets data analysis SQL Query to construct table for analysis SELECT sfd.name, slam.* FROM [CRM].[data_mart].[sla_master_view] AS slam LEFT JOIN CRM.data_mart.dim_service_form_definition AS sfd ON slam.service_form_definition_key = sfd.service_form_definition_key WHERE sfd.name IN /* Street SRs to filter for */ ( 'Pothole in Street', 'Alley Maintenance Request' ) AND slam.created_on < '2023-10-01’; Azure Data Studio Notebook link 50 Key Takeaways Time to complete service requests was higher in 2023 than 2022 partially due to increased pothole requests in summer 2023 Invalid service requests (outside City limits) are raising SLA times Alley requests are highest in Spring but are not resolved until Fall 51 Potential Future Data Projects Predicting which streets might have potholes from Street GIS layer, Street Condition Index, and Cityworks data Aggregate CityWorks pothole data by council district or internally-designated districts Adding “SR Closed” as a status in addition to “SR Completed” Discussion: Do any of these projects bring insight for your team or are there other questions to investigate? 52