Miami-Dade Water & Sewer FLORIDA BENCHMARKING CONSORTIUM APRIL 21, 2016 Miami-Dade Water & Sewer Using Data to Reduce Call Wait Times Presenters: Lourdes Avalos Michael Amador Jason Mena

MD-WASD Facts 2016 Miami-Dade Water and Sewer is the largest water and sewer utility in the Southeastern United States, serving nearly 2.3 million people a day Provides wholesale water service to 13 municipalities, as well as water and wastewater service to the unincorporated areas of the County Receives an estimated 48,000 calls per month 78 dedicated call center agents

Process Flow Chart 3 Mention 12 (24) call types/skillsets. Q1: % of Calls resolved in IVR Mention 12 (24) call types/skillsets. Q2: % of Calls Answered within 2 Minutes of Queue 3

Call Types Customers call for assistance with: Start new service (CM) Stop service (DM) Request Extension on bill (EXT) Transfer service to new location (XFER) Tampered meter (CCI) Underground Leak (UGL) Conservation (CONSV) Disconnection (CONP) Billing Related Inquiries (HB) Lien Issue (LIEN) Operator (OPER) Other (Other) Each call type is provided in either English or Spanish. 4

Problem Identified Data for our Q2 indicator showed a sizeable gap. 5 Source: ASE Scorecard 5

Develop Project Charter 6

Develop Project Timeline Plan The project’s timeline followed a traditional DMAIC approach to problem solving. 7

Monitor Team Progress A DMAIC Checklist was used to monitor progress and identify analytical tools used. Team identified an indicator; developed a Flowchart and a Spreadsheet Spreadsheet Line Graph Histograms and Paretos Single Case Bore Analysis and Fishbone Countermeasures Matrix; Barriers and Aids; Action Plan Before and After Line Graph Additional Bar Graphs Process Control System; Flowchart 8

Data Inaction: 9

Data In Action: 10

Stratify the Problem The team stratified Call data from the month of May 2014 (38,084 calls) and found: Explain intervals vs. calls. So the team stratified the call details for the 14,382 calls answered after 8 minutes (“late”). 11

Call Wait Time is attributed to overall call Handle Time, Call Volume, and Staffing. Handle Time was addressed in a prior project, and some staffing changes are currently being implemented, so the team narrowed in on Call Volume. In looking at the Volume of late calls (14,382) the team discovered that: PROBLEM STATEMENT: The High Bill Skillset Calls account for 1/4 of all Late Calls (>8 Min. Answer Time) 12

Identify Potential Root Causes Root cause analysis was performed with a fishbone diagram to identify reasons for the voluminous number of late High Bill calls. A B C D 13

Identify and Select Countermeasures The team brainstormed many countermeasures and narrowed them down to these for evaluation: Countermeasures Matrix   Problem Statement Verified Root Causes Countermeasures Legend: 3=Moderately 5=Extremely 2=Somewhat 4=Very 1=Little or None Ratings Effectiveness Feasibility Overall Take Action? Yes/No The HIGH BILL Skillset Calls account for 1/4 of all WASD Late Calls (>8 Mins. Wait Time) A - The standard for staffing/ training at call center is out of date. Provide HIGH BILL Skillset training as part of Level 1 training (initial set of skillsets) to all new calltakers, thereby increasing the pool of capable agents on that skillset. 3  5  15  Y  B - The standard for providing information on credit adjustments is insufficient. Provide an example on the website of a sample pool or leak credit scenario, showing allowable ccfs credited and $ credit due AND/OR add an insert to bill with rate structure * 5 15 Y C – No standard exists to relabel bill. Redesign the bill with a REVISED heading by coding a reissued bill printed within the same billing cycle OR: stamp the revised bill with a handheld stamp. 3 D – No standard exists to allow for modification of IVR. Ensure new Avaya phone system configuration will perform review/qualification of account to determine whether account is eligible for a self-service payment extension. 25 Y Phase 2 The team selected countermeasures for implementation. 14

Additional Recommendations: Relabel HIGH BILL to just BILLING call type Consider 'staffing pod' for BILLING calls and allocate appropriate average # of Call-takers. Identify dedicated staff to handle BILLING calls as their top priority, and then other calls as available Ensure (or prioritize training for) Call-takers that had OTHER and OPERATOR Skillset type also have BILLING Skillset Track # of BILLING callback requests (impacting available staff and ports) Consider acquiring license for workforce optimization software module to assist with staff scheduling 15

Develop and Implement Action Plan The team implemented an Action Plan for their Countermeasures. 16

Additional Call Center Improvements: 311 Skillset Sharing Partnership Project Employee Incentive Programs (team competitions, parking privileges, pizza parties, perfect attendance recognition, etc.) Overall Reduction of Backlogs in Corrective Billing Adjustments Reserved Agents Pool 311Direct phone app “One-Stop-Shop” Website Application Enhancements 17

Q2 – % of WASD Calls Answered on time (within 2 Mins. of Queue) Review Results The team collected indicator data and reviewed results of its implemented countermeasures. Q2 – % of WASD Calls Answered on time (within 2 Mins. of Queue) Target = 80% * New management team begins May 2015 The team will continue the implementation and monitoring of the countermeasures. 18

Percentage Of Calls Answered Within Two Minutes (Target 80%) March 2015 vs March 2016 19

Average Wait Time March 2015 vs March 2016 ASE Target Three Minutes

Calls Answered March 2015 vs March 2016

2016… 22

Lessons Learned Next Steps: 1) Data, data, data! Having the right data to work with is crucial. 2) Having a functional IVR reporting system helps us understand the true customer experience, which will help us identify additional fixes needed. 3) DMAIC Process is systematic and effective for problem solving. The team learned to trust the process and “let the data take you to the root causes”. Next Steps: * Continue to monitor the countermeasures and performance results. * Analyze Call Queue data and continue to address long wait times. * Continue to monitor call stats after workforce optimization software is implemented to compare performance changes. 23

MD-WASD Secret Sauce To Success: TEAM!