Big Data

Report
PepsiCo & Safeway
A “Big Data” Collaboration To
Reduce Out-Of-Stocks Using
Visualization Techniques
Carl Graziani
SVP Supply Chain, Safeway Inc.
John Phillips
SVP, Customer Supply Chain &
Global GTM, PepsiCo
There Is A Lot Of Data For Collaboration
S
Safeway Data Sharing Programs
Data Sharing Programs In Place With CPG
Vendors In Marketing & Supply Chain
Supply Chain Data Sharing
Shopper Insight /
Loyalty Data Sharing
Share
POS And Inventory Data
Marketing Data At Household
And Segment Level
Aimed At
Reducing OOS And Inventory,
Increasing Sales
Decisions On Assortment,
Pricing, And Promotion
Used By
Customer Supply Chain Teams
Customer Marketing Teams
Typically Cost
Nothing To Participate
A Fee To Participate
S
Safeway Data Visibility Program
20 Vendors Are Now
Receiving Data From Safeway
 Collaborative Process With
Safeway Supply Chain To Request
Firm Orders To Reduce OOS’s,
Distribution Voids & Pre-Event
Allocations
 Working With PepsiCo & Deloitte
On A Data Visualization Program
 Collaborative Process With
Safeway Marketing Groups For
Specific Competitive Responses
 Vendors Are Beginning To Report
Fourth Quarter Benefits Back To
Safeway
S
Data Visibility Core Competencies
Innovative
Collaborative
Comprehensive
Strategic
• Relatively New Program Even Though Data Sharing Is
Not A New Concept
• Opportunity For Increased Collaboration Between
Supplier & Retailer
• Real Time Visibility Can Influence Many Areas Including
Assortment, Inventory, Distribution & Promotion
• Vendors To Take The Lead In Driving Insights Through
Leveraged Data Sharing
• Program Enhanced Through Vendor Feedback
S
PepsiCo & Safeway Are Collaborating Further
P
P
360° Retail Execution™ Delivers “Big Data” For
Driving Performance
 Every Item / Every Store / Every Day
 31+ Retailers Sharing Daily Data
 53,234+ Retail Stores
 130 Million Saleable Units Every Week
 Enterprise Program Driven From Center
 Annotated With Attributes & Hierarchies
 Activated With Account Teams, Supply
Chain Field Execution
P
PepsiCo Believes In The Power Of Data &
Analytics To Drive Supply Chain
 Near Real-time Data & Dashboards
 Identifies Actual & Predictive OOS &
Overstock Issues At SKU/ Store Level
 Enables Root Cause Analysis
 Actionable Tasks Prioritized By Profitability
 Drive Sales & Execution
‒
‒
‒
‒
‒
‒
New Product Introductions
Closing Distribution Voids
Promotion Execution & Effectiveness
Store Merchandising & Replenishment
Order & Shipment Forecasts
Retail Pricing Compliance
P
Demand Signal Repository (DSR) Overview
Retailer
Shares POS
Data
Account Team
Shared Scorecards
DSR
Cleanses &
Stores Data
Improved
Shopper
Experience
Dashboards
BI Tools
Advanced
Analytics
OOS
Phantom Inventory
NPI
In-Store
Execution
Promo Execution
Supply Chain
P
Examples Of Driving Value Through Shared Data
Filling Distribution Voids Through Scripted
Replenishments
Field Teams Are Leveraging Gap Scans
Increasing Forecast Accuracy & Driving Supply Chain
Efficiencies Through True CPFR & VMI
Improved Forecasting Approaches Are Resulting From
The Safeway / PepsiCo Partnership
P
Filling Distribution Voids Through Scripted
Replenishments
PepsiCo 360° Analytics Reveal D-Voids
 Item-Store-Day Analysis
 Planogram Compared To Sell-thru
 Missing Items Identified
 Potential Lost Sales Calculated
PepsiCo & Safeway Resolve D-Voids
 Jointly Develop DC Force-Shipments
 Determine Which Products
 Agree On Quantities Needed
P
Filling Distribution Voids Through
Scripted Replenishments
Opportunities Identified
 12 Brands Analyzed
 Weekly Store Lost Sales Amounted To
Several Thousand Dollars Per Item
Actions Taken
 Safeway Pushed 2,339 Store/SKUs
Across 49 Items
 PepsiCo VMI Replenished DC Inventory
Results
 Recaptured Sales = $500K -$1.5M
P
Collaboration Has Led To A Change In Safeway’s
Internal Processes, Resulting In Benefits Along
The Entire Supply Chain
 Higher Forecast Accuracy
‒ MAPE Reduced 20%
‒ Bias Reduced 15%
 Improved Store In-stocks
 Less Days Of Supply
‒ DOS Reduced 15% YOY
 Shorter Order Lead Time
 Improved Service Levels
‒ +1.1% Service Level Improvement
S
We Have Data But Need Analytics & Visualizations
Companies That Excel In Advanced Analytics Also Excel In
Financial Performance With Profit Margins In The Range
Of 19 To 73% Higher Than Those Of Other Companies
Data To Enable
Decision Making
Increasing Employee
Expectations
Shortage Of
Analytical Talent
Analytical IQ For
Competitive
Differentiation
Increasing Customer
Expectations
Personalization And
Hyper Targeting
The Data Tsunami
Availability Of Data
Source: Jim Duffy and Scott Rosenberger, The Future of Consumer Products Companies: Technology – Gaining an Advantage with Advanced Analytics, 2007
S
Why Visualization?
Market Forces Highlight The Growth Of
Data, A Need For Talent, Changing
Expectations & Improving Decision Making
70% Of Our Sensing
Receptors Are Dedicated
To Vision
Certain Visuals Are More Impactful Than
Others Such As Relative Position,
Groupings, Shading, Etc.
fonts, weights,
sizes and colors
S
It’s All About The User Experience
We Need To Move From Rows And Columns To Something More Natural And Impactful
Yesterday
Today
These
Just asconsumers
consumersare
are
employees
being preconditioned
and need to
be
totrained
learn visually
the same way
fonts, weights,
sizes and colors
S
Companies Are Investing Significantly In
Visualization Capabilities
Deloitte Has Made Significant Investment In Our
Visualization
Because
We See
And Capabilities
Other Leading
Consumer
Companies
Such As P&G Data
VisualizationProducts
As A Critical
Step To Understanding
AreDeveloping
Also Making
Similar
Bets With
And
Deeper
Insights
“Business Spheres” (~50 Locations).
S
Familiar Visualization Examples
Source: SourceMap
(http://sourcemap.com)
Source: Lumino.so (http://www.luminoso.com/)
S
PepsiCo, Safeway & Deloitte
Visualization Design Session
On August 28th, 2012 PepsiCo And Safeway Came To The HIVE
(Deloitte’s Highly Immersive Visual Environment) To Design And
Rapidly Prototype New Ways To Visualize Key Challenges
S
Reduce Out Of Stocks & Improve Days Of Supply
Challenges And Questions To Address
 What Locations Are Causing The Most Significant Challenges &
What Are Those Challenges?
 What Causal Variables Are Impacting My Days Of Supply & Out Of
Stock Performance?
 What Brands Are Most Impacted By Out Of Stock & Days Of Supply
Performance?
 What Are My Total Lost Dollar Sales For A Particular Set Of Events?
P
Reduce Out Of Stocks & Improve Days Of Supply
Streamgraph
Force-directed graphs
Tree Maps
Sunburst
Word Tag Cloud
Bubble Chart
Many Eye Bubble Chart
Time Series Analysis
Parallel chord
Calendar View
Heat Maps
1. Consider
A Technique
To Visualize
The Data
Geo Spatial
Often Used For Highly Dimensionalized Data Sets
P
Reduce Out Of Stocks & Improve Days Of Supply
Rapid Prototype
2. Identify Casual
Variables That
Impact Out Of
Stocks And Days
Of Supply
3. Showcases
Trends
Between
Those Factors
P
Opportunity
What stores
aremeet
key drivers?
Store
does not
in-stock target
Reason
Code
NE
Not Enough
NO
Not Ordered
SA
Sold After
SI
Item doesWarehouse
not meet target
What items
key drivers?
WA are
Adjustment
Stocking Issue
WS
Warehouse
Short
Analysis
Review store
orders, forecast
assumptions, POS
Review display
inventory
Validate available
inventory
Action
Force-out
Product
Adjust Store
Inventory
Review Store
Ops
Ensure VMI
Reorders
Determine
Recovery
P
Reduce Out Of Stocks & Improve Days Of Supply
How This Visualization Works
Effect
Cause
Sun
Brand N
Day of Week
Brand
Item Velocity
DOS
OOS Reason
High
OOS Percentage
Lost Dollars
Wed
Brand 2
Med
Med
WH Short
Med
Med
Mon
Brand 1
Low
Low
Not Ordered
Low
Low
High
Stocking
Issues
High
High
Whereas Here There Is No Obvious
Trend…maybe You Have To Dig Deeper
Groupings Show Trends,
Even If Just One Color
P
Reduce Out Of Stocks & Improve Days Of Supply
Diving Down Into A Specific Product
Root Cause Identification: What are the key
reason codes for out-of-stocks in select
districts?
Corrective Action: If root causes can be
identified, can corrective actions be put in
place to reduce or remedy the out-of-stocks?
Please select the DISTRICT:
District 1
District 2
District 3
District 4
District 5
District 6
District 7
District 8
We are going to begin by
filtering our dataset. We
can hide in stock data
Please select the TRADEMARK:
Product 1
Product 2
Product 3
Product 4
Product 5
Product 6
Product 7
Product 8
Product 9
Product 10
Product 11
Product 12
Product 13
3.39% OOS rate.
Let’s dig in
further
XX District: District 6
YY Trademark: Product 6
P
Reduce Out Of Stocks & Improve Days Of Supply
Change Thread Coloring To Develop Insights
So far we:
1. Filtered to show only OOS products
2. Changed color to reflect OOS Reason Code
More “stocking issues” than expected…if we highlight these threads we can
see them more clearly
P
Reduce Out Of Stocks & Improve Days Of Supply
Following A Thread To Develop An Insight
Interesting, almost all of the stocking issues
are from store 2272. Need to get a macro
view across all products
Highlighted stocking issues
P
Reduce Out Of Stocks & Improve Days Of Supply
Zoom Out And See Bigger Picture
Let’s back up and look at the big picture:
• All products
• All OOS
• Entire district
Let’s dig into store 2272
Over $4000 in lost sales for
XX
this district over this period
YY
District 6
Product 1, Product 2
Product 3, Product 4, Product 5, Product 6,
Product 7, Product 8, Product 9, Product 10
P
Reduce Out Of Stocks & Improve Days Of Supply
One Location Is Causing A Significant Number Of Lost Sales
Highlighted threads related
to store #2272
About 1/5 of the lost sales
dollars for this region come
from one store! That’s 3X
higher than the average
XX
District 6
Product 1, Product 2
YY
Product 3, Product 4, Product 5, Product 6,
Product 7, Product 8, Product 9, Product 10
P
Reduce Out Of Stocks & Improve Days Of Supply
Develop Actionable Insight
Removed all other stores to
focus analysis on store #2272
Highlighted threads related
to store #2272
1/4 of the issues for this store
occur on Saturday. We are
exploring solutions with
PepsiCo for alternate delivery
or incremental
storage
XX
District 6
Product 1, Product 2
YY
Product 3, Product 4, Product 5, Product 6,
Product 7, Product 8, Product 9, Product 10
P
What We Learned: Collaboration Is Essential
 Retailers / Supplies Share The Shelf
 The Magic Comes From Sharing
‒ Data: Must Be Free & Open
‒ Insights: Joint Interest In Analysis
‒ Actions: Aligned On Plans / KPIs
 Data Visibility & 360 Retail Execution
Building “Big Data” Muscle
 More Data Streams Are Coming With
Digital Couponing, Etc.
 Data Has Value Through Collaboration
S
Safeway & PepsiCo Will Build On Successes Using
“Big Data” & Visualization Techniques
 Supply Chain Remains An Opportunity For
Improved Productivity Within CPG
 Data Sharing Provides A Foundation For
Retailer/Supplier Collaboration
 New Data Visualization Techniques Will
Make Use Of Data More Intuitive
 CPG Industry Needs To Develop Analytical
Competencies in Their Supply Chains
P
Any Questions?
P
P

similar documents