Working Capital - > 90 Day Delinquency Project – 0307-49-001 Heath Peacock Finance Tempe, AZ - CCC
Jan 21, 2016
Working Capital - > 90 Day DelinquencyProject – 0307-49-001
Heath PeacockFinanceTempe, AZ - CCC
3
Project Definition
Problem Statement:There is an opportunity to reduce delinquency for Trade Collection accounts with > 90 day delinquency in Tempe, AZ CCC. Excess delinquency leads to increased borrowing costs, reduced cash collected and additional reserve requirements.
Project Definition:Implement a series of improvements designed to increase collector and management effectiveness, enhance system utilization and reporting to reduce 90 day delinquency.
Project BenefitsReduce borrowing costs, increase cash and reduce reserves
$288,000 in reduced borrowing costs
7
Measurement System Analysis
Data type – Normal, continuous based on P = .536 Data Source – Data pulled directly from ARMS Goal – Verify Month end reported data is accurate Procedure – Pulled random sample of delinquency
bucket data and audited Sample – 135 random accounts, 15 in each
delinquency bucket Result – 100% accurate
Measurement System is Accurate!
8
DPMO and Sigma Level
Initial Capability Analysis Outcome Upper Specification Limit - 6.5% Sigma is short term based on 11 data points at month end DPMO – 350,000 based on delinquency in excess of Upper
Spec Limit Sigma Level – 1.88 Average delinquency – 10.28%
9
Data Collection Plan
Obtain calls by Collector – # of Calls, both incoming and outgoing
Delinquency Bucket by Collector Number of Accounts and invoices by Collector Time employed by IKON by Collector Sample data from IOSC collections history – number
of calls, Promise to Pay percentage and Percent Current
Leveraging IOSC Collection Experience
11
Tempe Data Correlation
1023.25
463.75
2746.5
1161.5
311.25
119.75
1248
526
Months with IKON
Customers Assigned
Total Invoices
Incoming Calls
External Calls
Correlation
Some Correlation
Minimal Correlation between Key Process Input Variables
12
Tempe Data Regression Analysis
Months with IKON – R-Sq (adj) – 0.4% Customers Assigned – R-Sq (adj) – 0.0% Total Invoices – R-Sq (adj) – 1.2% Incoming Calls – R-Sq (adj) – 0.0% But… We know, I know, the one thing that will most
definitely affect over 90% is the number of external calls by collector…
13
Tempe Data Regression Analysis
15001000 500 0
0.3
0.2
0.1
0.0
External Cal
Ove
r 90 p
erc
S = 0.0503701 R-Sq = 0.4 % R-Sq(adj) = 0.0 %
Over 90 perc = 0.0773945 - 0.0000103 External Cal
Regression Plot
!!!!!!
14
Tempe compared to IOSC
Compared Collector Activity to IOSC Collector Activity
Ho = Call volume is same Sample Size 175 and 168
over same time frame Mean – 32 and 45 Standard Deviation – 14 and
10 2 Sample – T test, P Value –
0.00Avg CallAvg Call
80
70
60
50
40
30
20
10
0
Tempe - IOSCOutbound Calls
Reject the null Hypothesis
15
Next Steps
Data is pointing out that the collectors may be on the phone but are not resolving the customers issues.
Establish new data collections Who are the collectors calling What issues are keeping them from collecting
Questions?