Payroll Remittance Case Study

Payroll Remittance & Time Tracking Cycle Time
Sigma
Background-Problem:
Before
= 1.0
 Non-exempt employees have to submit time-tracking information
twice, first using a web enabled tracking form that records the time
spent on customer work requests and the second using an excel
payroll form which tracks the hrs worked each week. Cycle time is
on average 23 min. weekly per person. Non-exempts have
complained about the non-productive rework of using 2 forms.
 Project to target labor efficiency savings by decreasing wasted time
spent by non-exempt employees when remitting time to payroll
using 2 different medias. Allowing the increase in productivity willProcess Capability Analysis for Time
increase the amount of billable time available each week.
Sigma =
USL
3.3
Process Data
After
 Improving quality by providing all USL
time-tracking
staff
with
a
15.0000
Target
*
centralized process for remitting time
to
payroll
LSL
*
Mean
8.2731
 Goal:
Sample N
26
StDev (ST)
3.34791
StDev (LT) cycle5.14295
 Reduce payroll submittal & time tracking
time by 50% (unit of
measure: minutes), Upper Specification
Limit (USL) > 15 minutes.
Potential (ST) Capability
Cp
*
 Approach: DMAIC
CPU
0.67
CPL
*
 Data Collection: Cycle time for payroll
remittance
Cpk
0.67 was measured
Cpm
*
-10
-5
0
5
10
15
20
25
over 1 month (2 cycles) with a two
week pay period,
non-exempt
Overall (LT) Capability
Observed Perf ormance
Expected ST Perf ormance
Expected LT Perf orm
employees only.
 Results:
Pp
*
PPM < LSL
*
PPM < LSL
*
PPM < LSL
PPU
0.44
PPM > USL
115384.62
PPM > USL
PPM > USL
95
 Root Causes: Submitting time-tracking
in two forms
results
in
After Web-Form
Payroll22253.27
Implementation,
PPL
*
PPM Total
115384.62
PPM Total
22253.27
PPM Total
95
Ppkproductivity. 0.44
increased cycle time and decreased
mean cycle time was reduced from 23.17
minutes to 8.27 seconds. The result is an
 Solution Selection: Web tracking form feeding Oracle payroll
database solution was developed to track employee time and submit annual savings of $700K of increased
employee productivity
to payroll. Cycle time and accuracy was measured during the first
two weeks of rollout.

Call Center Backup Slides
Digitization
Process Start: Non-exempt
Payroll &
Time Tracking
employees enter time tracking info
into payroll excel form
Process Stop: Payroll Submits
Payroll for check issue
Cycle Time
Process
Requirement:
Defect
Definition:
Unit Definition:
Unit of
Measure:
Data Type:
Payroll submission cycle time <=15
minutes
Payroll submission cycle time
> 15 minutes
1 payroll form for 1 employee for 1
payroll period (2 weeks)
Minutes
Continuous
Background Process & Root Causes
Web form is not
meeting the
requirements of all
stakeholders,
resulting in rework

Machines
Non-Exempts enter time tracking info
into payroll excel form
Start
Payroll compiles and
enters into Oracle
Different Excel Form used (Format
Erroneous Information
ADP Payroll
System
Web based application
Non-standard process
applications
Inaccurate information being entered in excel forms
Timesheet Processing
Manually compiling/reconciling/ verifying staff timesheets
Too Many manual process steps
involved with payroll submission
Info is located in more than 1 place
New employee profile not setup correctly
ADP System Down
Number of hands involved
with payroll submission.
New Staff Untrained
Non Exempt Employees
Locating Information
Manually Entered
HR provided incomplete info
Admin’s & Managers
Manually compiling/reconciling/ verifying staff timesheets
Payroll Dept.
Exempt PTO must be
entered manually
Estimating time vs. recording actuals
More Than 1 Start/Stop
Manual processes in compiling and verifying
Too many start/stops lead to
consolidated time tracking
Poor Operational Definitions
Measurement
Why is payroll cycletime >
15 minutes?
Different versions of Excel forms
E-mailed
Phone calls; Tickets; Customer requests
Distractions
Excel Form
Manually entered
Putting off timesheet until last minute
Rounding errors
Operations
Stop
Having complete time in more than 1 form.
Customer programs use different forms
Separate e-mailed forms by
individual staff
Website down
Regional Admins compile
time tracking reports
People
Methodology
Manually entered
Program
Accounts
Admins/Mgr. compiles Excel form
and review both Web & Excel
forms
Payroll Submits Payroll for
check issue
Missing Information
Requires correct
format
Non-Exempts enter time tracking
info into web-form
Forms
Uses military time
Define – Problem Statement
Stakeholder analysis was performed to understand buy-in needs of all
groups
Quality Functional Deployment (QFD) was used to prioritize each identified
critical-to-quality (CTQ) involved in the process
Payroll submission cycle time <= 15 minutes
was selected for project Y
+
+
+
Focused on steps involving
excel form, compiling and
Importance
Importance Ratings
Ratings
Strong
Impact
99
data entry into Oracle
Strong Impact
Moderate
55
Moderate Impact
Impact
system.
Weak Impact
1



Process Requirements
Non-Exempts enter time tracking
info into payroll excel form
Non-Exempts enter time tracking
info into Web form
Admins and/or Mgr. compiles
Excel form and review both web &
Excel
Payroll compiles and enters into
Oracle
Regional Admins compile time
tracking reports
Payroll Submission Cycle Time
5
<= 15 Min
9
9
9
9
1
4
Payroll Submission Usability
(Simple)
3
> 95%
9
9
9
9
3
Availability to access payroll
submission form
5
100%
9
1
1
1
2
Ability to Edit for Changes for pay
period
5
100%
9
1
9
9
Employee recieves paycheck
5
100%
9
9
9
Complete MGMT reports
3
100%
9
9
1
Payroll Submission Completeness
5
100%
9
9
9
9
36
Payroll Submission Accuracy
5
100%
9
9
9
9
36
64
56
Weak Impact
1
Stakeholder Analysis
6
CTQ Prioritization
Payroll Submits Payroll for check
issue
CTQ Priority

5
5
4
1
0
0
0
Strongly
Against
Moderately
Against
Neutral
0
Moderately
Supportive
Strongly
Supportive
Totals
37
1
37
12
28
72
20
9
9
10
36
28
28
Measure





Descriptive Statistics
Descriptive Statistics
2 operators collected 2 data
points (For 4 different weeks)
on 4 different payroll forms (8
Trials)
Variable: Total Time
Anderson-Darling Normality T
Anderson-DarlingA-Squared:
Normality T est 1.1
A-Squared: P-Value:
P-Value:
Web time tracking week
submissions & the payroll
forms were used, collected
separately and then combined
into the total cycle time
Each operator
recorded
0
20 the
same cycle time for 1 operator
to complete the payroll cycle
time process. (Repeatability)
Variable: Total Tim
40
0
20
60
40
60
95% Confidence Interval for Mu
No variation was found
95% Confidence Interval for Mu
between operators using the
computer clock time stamps.
(Reproducibility)
15
20
30
1525
20
25
MSA needed to be conducted 3
times, each requiring updates
Confidence Interval for Median
to the operational95%
definitions.
95% Confidence Interval for Median
Mean
StDev
Variance
Skewness
Kurtosis
N
Minimum
1st Quartile
Median
3rd Quartile
Maximum
1.190 0.0
0.003
Mean
23.17
StDev 23.1786 15.37
Variance15.3721 236.3
236.300 1.294
Skewness
Kurtosis 1.29442 1.740
1.74046
N
28
Minimum
2.00
2.0000
1st Quartile
13.00
13.0000
Median
18.50
18.5000
3rd Quartile
31.7500 31.75
Maximum70.0000 70.00
95%Interval
Confidence
95% Confidence
forInterval
Mu for
17.2179
17.2179 29.1392 29.13
30 Confidence
95%
forInterval
Sigma
95%Interval
Confidence
for S
12.1535
12.1535 20.9235 20.92
95% Confidence
forInterval
Median
95%Interval
Confidence
for M
13.0000
13.0000 27.5517 27.55
Analyze
Run Chart for Total Time by Week
70
Week 2
Week 1
95% Confidence Intervals for Sigmas
Factor Levels
No special causes
impacting the data
over time
60
50
Total Time
Test for Equal Variances for Total Time
Antenna
Web
Excel
Payroll
40
0
50
30
20
100
150
F-Test
Levene's Test
Test Statistic: 0.108
Test Statistic: 21.054
P-Value
P-Value
: 0.001
: 0.000
10
Boxplots of Raw Data
0
5
15
Time used via web vs.
excel forms
statistically different
Antenna
25
Observation
Number of runs about median:
Expected number of runs:
Longest run about median:
Approx P-Value for Clustering:
Approx P-Value for Mixtures:
12.0000
15.0000
6.0000
0.1239
0.8761
Number of runs up or down:
Expected number of runs:
Longest run up or down:
Approx P-Value for Trends:
Approx P-Value for Oscillation:
Payroll
15.0000
18.3333
4.0000
0
0.0612
0.9388
Root Cause / Sources of Variation
1.Too many manual process steps with payroll submission
2.Manually compiling/reconciling/ verifying staff timesheets
3.Having complete time in more than 1 form
4.Different versions of excel payroll forms
10
20
30
40
50
60
Total Time
Remote (Blue) take 6
minutes longer than
Campus (green)
locations
14.5
20.5
70
Improve
Pilot Plan:




Outsouce to a robust
payroll application

Call Into payroll system
using numbers entered as
Payroll Submission
Training calls for managers will be held and
feedback used to make any necessary
changes, prior to roll out
Create a single database
for single payroll submittle
(No web form)

Standardize Improved
Payroll Form
Open training calls will be held prior to roll
out, ensuring all employees are trained on
new web-based payroll submittal
Digitize web form and
excel payroll form

Baseline (AS-IS) Do
nothing
Enhanced web-payroll solution will be
implemented in Q1, 2003
Solution Alternatives
Importance Rating

Pugh Matrix
Cost
10
-
+
+
-
-
-
Error Checking
8
S
+
S
+
+
+
Fail Safe (Emergency Backup Solution)
10
S
+
S
+
+
+
Low Learning Curve (Simple to Use)
5
-
+
+
+
-
+
Real Time
7
S
+
S
+
+
+
Documentation, including process escalation
paths, training manuals, and process maps
will be sent out as a reference for all accounts
Editable Adjustments
7
S
+
S
+
-
+
Adjust for Holidays, Weekends and PTO
10
-
+
S
+
+
+
Overtime and On-Call
8
-
+
S
+
+
+
Helpdesk form will be updated to reflect any
additional employee information needed
MGMT Verification
10
S
+
+
+
-
+
Documentation & Training
9
-
+
S
+
+
+
Sum of Positives
0
10
3
9
6
9
Sum of Negatives
5
0
0
1
4
1
Sum of Sames
5
0
7
0
0
0
Weighted Sum of Positives
0
84
25
74
52
74
Weighted Sum of Negatives
42
0
0
10
32
10
Totals
-42
84
25
64
20
64
Regular team meetings will be scheduled in
order to document risks and process
deficiencies
Accounts with duplicate profiles will be
removed from database prior to roll out and
new ones created as necessary
Payroll accuracy will be gauged during the first
two pay periods after implementation
Concept Selection Legend
Better
+
Same
S
Worse
-
Key Criteria
Pugh Matrix used in evaluating key criteria to each proposed solution
Improve Continued
How was the solution determined:

Changing Systems & Structures
Boundaryless Question
Examples
Brainstorming with Subject Matter Experts
(SME) for identification of technology
improvements needed to remove the manual
reworked process steps
Vertical
Was a Pugh matrix or other tool used:

Both the Pugh Matrix (Criteria-Based Matrix)
and brainstorming with multi-functional team
pointed to web site real time coding
improvement opportunities

Stakeholder analysis for change supporters
and resistors with influence strategy plan
applied

Changing systems and structures analysis in
order to understand the vertical, horizontal,
internal & external change impacts

Failure Modes & Effects Analysis (FMEA) user
to capture & mitigate risks

Each program account in the country
worked slight variations to the payroll
process (Including web site tracking,
Different spreadsheets, etc…)
Managers required manual
spreadsheets from each employee, then
consolidated onto one spreadsheet and
validated by 1-4 managers
All programs now use the same
process, reducing variations and nonvalue added steps.
Horizontal
External
The redundant process steps had
Employees are now more productive
employees doubling or tripling the tasks with the removal of non-value added
How well have we
needed to submit time
steps, allowing for an increase in
removed barriers
customer CTQ's and value added
between our
support tasks
business, suppliers,
Manual process steps between business Digitized process steps have freed
customers, and
units lead to long processing cycle times. payroll and MGMT tasks to coping
competitors?
with escalations and faster
processing times!
Average loaded employee rate (salary +
benefits) used to compare pre & post cycle
times for productivity savings
Team Effectiveness Tools:
How well have we
removed the
vertical boundaries
Managers have a one-shop view of Time Reduced e-mail, reduced labor used
in our
Tracked + Payroll Hours Entered for All by MGMT and have combined time
organizations?
their reports vs. One-at-a Time
tracking compliance with Payroll
Remittance
How well have we
removed walls
between groups
within our
business?
Cost Benefit Analysis:

Assessment
Managers can real-time view an
100% improvement as Managers can
employees Time Tracking vs. a bi-weekly now adjust an employees time weekly
manually circulated spreadsheet
and address OT & PTO issues faster!
Managers can view all employees
digitally in one view, other managers
can also view account reports with
access to the web-based Site!
The changing systems and structures assessment
allows for the discovery of process change impacts
(both positive and/or negative) in all directions of
the process. Helping to identify resistance to change
and change risks across additional groups.
Control

The Failure Modes & Effects Analysis (FMEA –
“Risk Analysis”) will be a living document, used to
address any process problems during and after the
new system rollout.

Documentation regarding the Web-Based Payroll
functionality will be dispersed to all employees to
reduce learning curve.

Escalation paths for the Web-Based payroll
problems will be formulated and documented prior
to rollout.

Training calls will be held prior to rollout for nonexempts and management.

Cycle time and payroll accuracy will be measured
during the first two weeks of rollout.

Eliminating the need to track time in two forms
reducing both cycle time and process variation

Centralization and further digitization of the
payroll remittance process will allowed for easier
adjustments in future maintenance

Reduced rework and increased staff productivity

Standard practices allow for simpler future
changes.

Web interface more user friendly than as-is
process
Variation by Operators
entering varying case
comments
Individual and Moving Range (I&MR) control chart
selected for continuous data with no subgroups.
Upper Specification Limit (USL) < 120 indicates process
control with mean of 12 & range 2 seconds.
Results
Was applying Lean Six Sigma helpful?

Defined the gap in performance and identifying
root causes

Solved root causes and validated solutions

Sustained improvements and shared lessons
learned
Financial Data Collection

10 weeks of data collected with web page Log
files created to count each additional sub case
created (does not count the original master sub
case assigned to engineers).

Web page case creation increased to 100% with
the full training of the engineers and usability
testing.

All financial savings validated with each business
division

Process improved from 1.12 to 3.3 Sigma

Human error eliminated with 100% automated
digitization of all transactional tasks

Standardized web submittals as single method
used with backup plan for phone should disaster
strike

Multi-generational project plan (MGPP) formulated
to measure and improve on remaining CTQ’s
(Critical-to-Quality) identified in Quality Functional
Deployment (QFD)
Baseline
1701
52
88452
Calculation Description
QTY
Average Qty Sub Cases Created (Weekly)
Multiplied by 52 w eeks to Annualize
Estimated Annual Average Subcases Created
Improved
1701
52
88452
Cycletime
120
260
260
0.043
0.072
0.072
1597.05
3808.35
4791.15
Average Engineer Labor Cycletime (Seconds)
Average Administrator Labor Cycletime (Seconds)
Average Call Center Analyst Labor Cycletime (Seconds)
Average Engineer Labor Cycletime (Hours)
Average Administrator Labor Cycletime (Hours)
Average Call Center Analyst Labor Cycletime (Hours)
Avg. Administrator Cycletime (Hrs) * Estimated Annual Average Subcases Created
Avg. Engineer Cycletime (Hrs) * Estimated Annual Average Subcases Created
Avg.Call Center Analyst Cycletime (Hrs) * Estimated Annual Average Subcases Created
55.29
0
0
0.015
0.000
0.000
0.00
1358.48
0.00
Average Loaded Pay Rates
$35
$47
Average Loaded Pay Rate (Engineer/Administrator)
Average Loaded Pay Rate (Call Center Analyst)
$35
$47
Cost
$55,896.75
$133,292.25
$200,000.00
$60,000.00
$225,184.05
$674,373.05
Engineer Indirect Labor Savings
Administrator Indirect Labor Savings
Quartly contractual penalties paid in 2002 (baseline) vs. 2003 (Improved)
Call Center Analyst Direct Labor Savings (2 staff reduced)
Call Center Analyst Indirect Labor Savings
Total Annual Labor Cost for Subcase Creations
22,000
4
88,000
$0.134
$11,748.00
Average Qty of Subcase Creation phone calls Annually (25% or 1700*0.25 = 425 calls w eekly
* 52 Phone
w eeks)Call (Averge 4 minutes)
Average Talk Time for One
$47,546.64
$0.00
$0.00
$0.00
$0.00
$47,546.64
Phone Cost
Annual qty of talk time minutes for subcase Creations
Cost for 1 call (minute)
Total Direct ITS Phone Savings Cost
0
0
0
$0.116
$0.00
Cost Savings
Total Web-Based Subcase Creation Project Indirect Labor Savings (Annual)
Total Web-Based Subcase Creation Project Direct Labor Savings (Annual)
Total Web Page Subcase Creation Project
Indirect (Labor ) + Direct Savings (Annually)
$626,826
$71,748
$698,574