2. Confidential - for limited circulation only
Agenda
• Relational Solutions/Mindtree Background
• Case Studies
• Identify Costly Replenishment Issues
• Reclaim Lost Sales Dollars
• Set Alerts to avoid Out-of-Stock
• What Makes this so Difficult?
• Go Beyond OOS
• What to do Next!
3. Confidential - for limited circulation only
Mindtree at a Glance
3
We engineer meaningful technology solutions to help
businesses and societies flourish.
$715M
annual revenue
17,000
Mindtree Minds
42
offices
17
countries
65%
US
25%
Europe
10%
APAC
22.5%
YOY growth
Top 7
IT company in India
5. Confidential - for limited circulation only
Global Presence
Basel, Switzerland
Brussels, Belgium
Cologne, Germany
Dublin, Ireland
London, UK
Paris, France
Solna, Sweden
Vianen, Netherlands
Europe Asia
Dubai, UAE
Kuala Lumpur, Malaysia
Singapore
Sydney, Australia
Shanghai, China
Tokyo, Japan
North America
Chicago, IL
Cleveland, OH
Dallas, TX
Gainesville, FL
Los Angeles, CA
Minneapolis, MN
New York, NY
Ontario, Canada
Phoenix, AZ
Scottsdale, AZ
San Jose, CA
Schaumburg, IL
Redmond, WA
Warren, NJ
Company HQ Delivery Center Delivery & Training Center
India
Bangalore
Bhubaneswar
Chennai
Hyderabad
Pune
5
6. Confidential - for limited circulation only
#1 SAP HANA implementation partner #1 Duck Creek implementation partner
Acquisitions in 2015-16
6
Leader in demand signal repositories,
data integration & COE for analytics
(predictive & prescriptive)
Platinum Salesforce implementation partner
7. Confidential - for limited circulation only
Our CPG Focus
7
Digital Marketing
Demand Chain
Management
Predictive Analytics
Trade Promotion
Management and Post-
Event Analytics
Channel Collaboration
Demand Signal
Repository & Data
Ingestion
BI and Analytics
Distributor Management
Sales Force Automation
CRM & Master Data
Management
8. Confidential - for limited circulation only
Agenda
• Relational Solutions/Mindtree Background
• Case Studies
• Identify Costly Replenishment Issues
• Reclaim Lost Sales Dollars
• Set Alerts to avoid Out-of-Stock
• What Makes this so Difficult?
• Go Beyond OOS
• What to do Next!
9. Confidential - for limited circulation only
Aligned Shipments with
Sales
Improved Vendor
Managed Inventory
CPG Case Study:
Demand Signal Repository
9
Sales Performance Tracking
Velocity And Distribution Report
Cross Retailer/ Cross Channel
Reports
Customer Profitability Analysis
11. Confidential - for limited circulation only
Providing
Recommendations on
Suggested Order
Quantities
Cloud enabled
collaboration platform
for 18,000 distributors
in 22 countries
Reduced order-to-cash
cycle by 60%
Savings $15M on store
audits in 3 years
11
CPG Case Study:
Winning the New Age Consumer
12. Confidential - for limited circulation only
Unified Digital
Marketing platform
70 unique brands
182 countries
1600 digital properties
40% cost savings
50% less time-to-
market
Consumer Goods Giant Engaging with Consumers
Faster and Better Across all Digital Channels
12
14. Confidential - for limited circulation only
The Perfect Store
• The Rep is adding to the
recommendations on our
Suggested Order Quantity.
• Learned process that
ultimately results in the perfect
store
15. Confidential - for limited circulation only
Agenda
• Relational Solutions/Mindtree Background
• Case Studies
• Identify Costly Replenishment Issues
• Reclaim Lost Sales Dollars
• Set Alerts to avoid Out-of-Stock
• What Makes this so Difficult?
• Go Beyond OOS
• What to do Next!
16. Confidential - for limited circulation only
Lowes
Walgreen’s
Costco
Shipments
Other POS data
EDI/FF/txt/Access, etc
Wholesaler/Dist
IMS
Submit Reports
Most time spent gathering disparate
data
•G ather ing
•Cleaning
•Integr ating
•Justifying
?
Inconsistent reports lead
to questionable decisions
Kroger/Market 6
Wal-Mart, Shiloh
Retail Link, EDI/AS2
?
?
? ?
?
?
What Makes Analytics So Difficult?
Kmart
IRI
Nielsen
Third Party Data
NPD
Weather DunHumby
RSI
EDI
17. Confidential - for limited circulation only
• G a t h e r i n g
• C l e a n i n g
• I n t e g r a t i n g
• J u s t i f y i n g
Why POS Data Cleansing?
Property of Relational Solutions, Inc. By Janet Dorenkott June 2015
Big Data, Omni-Channel
& New Data Sources
End User
Requests Change
Data Frequency is
Inconsistent
Data Formats Vary
From Source to Source
You Are at the Mercy of
Retailer Decisions
Constantly Changing
Conditions
POS is Not Always Available
& May Cost Money
Everyone Wants to See
Data Differently
Different Sources?
Different Reliability
Every Retailer Provides
Different Data Elements
Lack of Integration &
Manageability
18. Confidential - for limited circulation only
Every Retailer Provides Different Data Elements
Date Time Units Price Store # In Transit OH Inv OH Qty In
Warehouse
On
Order
9/16/2015 08:12:00 150 $3.20 #442 30 20 10 20 50
Date Time Units Store #
9/16/2015 10:14:06 122 #210
Retailer A
Retailer B
19. Confidential - for limited circulation only
Aligning Different Week Start & End Dates
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday Saturday
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
20. Confidential - for limited circulation only
Different Sources, Different Formats
EDI 852
EDI 867
SAP
Retailer
Portals
AS2
Oracle
TXT
SQL
Excel
CSV
How Do I get
all these
sources into
1 common
database
type?
FTP
Access
21. Confidential - for limited circulation only
Different Sources, Different Reliability
Missing
Data
EDI
Inconsistencies
UPC
Issues
Portal
is Down
Re-Casts
• Data was supposed to come in from CVS today! Where is it?
• RetailLink is down!
• Walgreens is padding UPC codes in the front with 2 digits!
• Food Lion is padding UPC codes in the back with 1 digit!
• Kroger wants us to buy data from Market 6!
• Dollar General was missing 2 days from last month!
• Target EDI came in 4 hours late!
22. Confidential - for limited circulation only
Different Retailers, Different Hierarchies, Different Vernacular
Internal
Category
Brand
Sub-Brand
Item
Product #
Retailer A
Category
Sub-
Category
Brand
Product #
Product ID
Sku
Retailer B
Category
Sub-
Category
Item #
Item Name
Sku
23. Confidential - for limited circulation only
Frequency of Feeds Vary
EDI
Daily
Syndicated
Data Monthly
ERP Data
Every Second
Currency
Conversion
Daily
Daily POS
Once a Month
RMA
Annually
Plan Data
When
Available
Weather Trends
Quarterly
24. Confidential - for limited circulation only
Constantly Changing Market Conditions
New Competitors New Retailers
Trends
New Laws
Acquire a Company
Economy Management Direction
National Conditions Natural Disaster
New Corporate Direction
Retailers Partners
Company switches
data providers
The company adds a new data source
Retailer Contracts
New CEO
25. Confidential - for limited circulation only
You Are Subject To Retailer Decisions
Make Portal
Changes
Start Sending
Inventory
New POS Source
Category Focus
Changed
New Buyer
Update Contract
Restrictions
Policy
Changes
More Data Offered
Opened New Stores
Acquired Another
Retailer
26. Confidential - for limited circulation only
End User Requirements Change
I Need to see
Target compared
to Walmart
Include
Shipments in
this Report
How do I
compare my
sales with
plans?
We just
bought new
demographic
data
I need to
calculate
potential
impact
I upgraded
our Nielsen
contract
We switched
from SAP to
Oracle
I want to
compare this
with weather
trends
I want to compare
POS from the
retailer with POS
from IRI
27. Confidential - for limited circulation only
POSmart
Consistency
• Easy Access to
Information
• Improved Productivity
Streamline Data
Gathering
Synchronize, Integrate &
Validate POS
Wal-Mart/Shiloh
EDI /AS2/Retail Link
Promo/Forecasts
Costco
Nielsen/Spectra/RSi
Shipping
Kmart
Kroger
IRI/Mkt6/NPD
EDI, AS2
Walgreens
Wholesaler/Dist
Home Depot
Productive Meetings
Integration & Harmonization
28. Confidential - for limited circulation only
Agenda
• Relational Solutions/Mindtree Background
• Case Studies
• Identify Costly Replenishment Issues
• Reclaim Lost Sales Dollars
• Set Alerts to avoid Out-of-Stock
• What Makes this so Difficult?
• Go Beyond OOS
• What to do Next!
29. Confidential - for limited circulation only
Military,
Distributor
Budget
3rd Party
GL
Forecasts
Reports
Scheduled
& Cached
Shipments
SourcesRetailers
WalMart
CVS
Costco
Kroger/DunH
Meijers
Sam’s
Distributors
POSmart
Integrate, Validate,
Synchronize & Manage
BlueSky
Over 200 KPI’s
Or Other BI Tool
DW
POS
Integrator
BIS
EDI, Text, Flat
Files, Access,
Retail Link,
Partners
SAP, JDE, DB2,
Oracle, JDA
Users access via
web
3rd Party Data:
AC Nielsen/IRI
Spectra/NPD,
Market6, RSi
Forecast
Shipments, Promo,
Vendor
TradeSmart
POSmart
“Smart” Architecture
OtherTPM
BlueSky Analytics
Power User
BlueSky Viewer
Casual Users or
BlueSky XL Users
Control
Center
30. Confidential - for limited circulation only
Deployed RSI
TradeSmart- Post-event
analytics solution
6% sales increase
1% profit gain
Go Beyond POS Integration:
Post event Trade Analysis to Optimize Spend
30
Promotion Execution
Baseline Sales
Uplift & ROI Analysis
Optimized Spend Allocation
Cannibalization & Halo Effect Analysis
Pantry Stocking/Post Event Dip
Trade Promotion Optimization
31. Confidential - for limited circulation only
Agenda
• Relational Solutions/Mindtree Background
• Case Studies
• Identify Costly Replenishment Issues
• Reclaim Lost Sales Dollars
• Set Alerts to avoid Out-of-Stock
• What Makes this so Difficult?
• Go Beyond OOS
• What to do Next!
32. Confidential - for limited circulation only
Next Steps:
Complete Our “Getting Started” Guide
Coordinate a Demo for your Peers
Start with key retailers and syndicated data providers
Automate standard reports
Start Providing new Insights
Predictive Impact & Prescriptive Recommendations
33. Confidential - for limited circulation only
Follow our Relational Solutions Training Blog
www.relationalsolutions.com/blog
Join Demand Signal Repository Institute Group
Follow Relational Solutions
Connect with Janet Dorenkott and Karen Kurtzweil
Watch our Training & Video’s on Relational Solutions Channel
Follow @POSmartBlueSky & @JanetAtRSI
Like Relational Solutions
Contact Janet Dorenkott 440-899-3296 x225 Janetd@relationalsolutions.com
Contact Karen Kurtzweil 440-899-3296 x235 Karenk@relationalsolutions.com
Stay Connected & Informed!
34. India | USA | UK | Germany | Sweden | Belgium | France | Switzerland | UAE | Singapore | Australia | Japan | China
Editor's Notes
Client food company with products ranging from peanut butter and jelly, coffee beans, cake mix, cooking oil and even pet food (a recent acquisition). We built their Demand Signal Repository for their VMI team. We worked with a wide range of data sources including direct downstream point-of-sale retailer data, EDI VMI data, multiple IRI databases, Promotion data from Siebel, as well as internal master data, shipments and order data. The data sizes are in the terabytes. They needed to know things like
How do our sales align with shipments?
What are our inventory turns? How does that compare with our competition?
Where are your items selling? Are there geographies that sell your products better/worse than other geographies? What drives that?
Are you growing or shifting volume? How does your price compare across geographies/channels and are you competitive?
They needed to understand On-shelf Availability
Tracking OSA
OOS Report
Predicted Out Of Stock
Phantom Inventory & Zero Scans
Root Cause Analysis
This retailer had a need to help their retailers understand what would sell. Cross Selling – We Use “R” (open source predictive analytics model/scripts). It looks at one store and the stores around it to see if it accurately reflects the products/proper assortment, mix and volumes. It takes into account store size and what various retailers are selling. For example, Target is carrying a product and selling $100K/month, but Walmart, right down the street, isn’t even carrying it. A rep could use this information to tell Walmart that they are losing business. For this calculation to be made, cross reference is needed.
Whose forecast drives production? Did you make enough product to keep up with demand? Is there an opportunity to introduce a complimentary item to stores to grow sales ?
Suggested Order Quantity (SoQ)
Providing a SoQ will allow smarter orders to achieve optimal sales
The SoQ can be calculated using consumption data and ESR, and enhanced further with the inclusion of shipment and order data. (comparing suggested to actual and benchmarking achievement rates)
Use SoQ to feed the CPG mfg Ordering systems
Cross Sell
Identifying the best items to merchandise or promote together for the highest sales impact
Identify what other items a store has a high propensity to sell from CPG mfg that are not currently selling in the store but in the market
This customer had a Replenishment project that needed to understand
Order issues And Safety Stock requirements
What they Must Stock
Shrinkage And Unsaleables
Demographics Based Assortment
They wanted us to offer Suggested Order Quantities. We can suggest quantities for any item across the globe based on either POS, Shipments, or Orders. We can also provide seasonal information. SOQ in advance tells you, based on history and information leading up to the event, what should be ordered. Seasonal SOQ is Predictive. In order to calculate this, we use historical sales data and sales data leading up to the event.
This customer had a need for Demand Forecasting
In this case We Uses historical sales data to determine future demand
& to better project inventory needs
They needed to know At what price do sales fall off? At what price do sales accelerate? What is the optimal price in each geograhpy/channel?
Does your product sell better/worse during certain seasons or weather events or social media events? What is influencing sales?
If your product is not on the shelf at the retail store it has no opportunity to sell. Our solution enables you to track lost sales even without counting on retailer on hand data. We compare expected sales to actual sales and send alerts when stores are not living up to expectations and an out of stock is likely.
Root Cause Analysis
Put the logic/decision tree in place to distinguish between upstream supply, mfg distribution centers, retailer distribution centers, 3rd party distributors, and store level or demand issues that lead to an out-of-stock occurrence.
This logic can begin with consumption data but will be more robust once additional data is included.
We also do something called Virtual On Shelf Availability (OSA)
We don’t rely on the retailer inventory information to calculate this, especially for DSD (direct store delivery) manufacturers. We have an algorithm that learns. OOS is based on the estimated sales rate. We compare what shipped to what sold.
The Perfect Store – This is the analytics we use to determine whether the Rep is implementing the SOQ that we are recommending. This is a learned process that takes into consideration our recommendations and what was actually ordered by the rep ordered in the field. Here’s what we recommend, here’s what you did, here’s the result.
How many stores are authorized to sell your items? How many are selling? Are their stores that based on sales rate or other variables be eliminated from your product distribution list? Is the POG integrity kept or are there items that have been placed in the wrong location that take space away from items the consumer is looking for?
Perfect Store
The best mix of product assortment, placement, space and price by store.
Scorecard that measures how a store is performing vs KPIs (compliance to SoQ, cross sell achievement…)
Store Clustering
Grouping the stores based on common store and/or demographic characteristics.
Clusters can be built from sales performance, size, type, demographics, etc.
One or more of the above store characteristics can be used to build clusters.
The clusters will help determine the best mix of merchandise and the assortment quantities.
This slide depicts the data dilemma that consumer goods companies face. The data is there. You get it. You might even have the impression you’re getting too much data. But that isn’t the case. The issue is having that data in a usable and reliable format. The data comes in from so many different sources that it’s impossible to get the full value out of it. Most companies are using it just to produce the reports they need. They just don’t have the time to create new insights even though the data is there.
Each one of these source provide information and potentially reports. But analysts spend an average of 80-90% of their time gathering and pulling together data. Inevitably, there are discrepancies in the reports and analysts are then stuck spending more time trying to figure out where the numbers came from. Not only is this a very tedious task but when people’s numbers don’t match, there is a lack of confidence in the data and rightfully so. But it wastes not only the analysts time but it also leads to wasted time on managements part when they argue over who’s data is right and could potentially lead to costly and incorrect decisions.
On top of that since 90% of an analysts time is spent gathering, cleaning, integrating and justifying their reports. This is time that keeps analysts away from actually getting to analyze the data and gain new insights. This is why our enterprise solution offers companies a very fast ROI.
This slide depicts the data dilemma that consumer goods companies face. The data is there. You get it. You might even have the impression you’re getting too much data. But that isn’t the case. The issue is having that data in a usable and reliable format. The data comes in from so many different sources that it’s impossible to get the full value out of it. Most companies are using it just to produce the reports they need. They just don’t have the time to create new insights even though the data is there.
Each one of these source provide information and potentially reports. But analysts spend an average of 80-90% of their time gathering and pulling together data. Inevitably, there are discrepancies in the reports and analysts are then stuck spending more time trying to figure out where the numbers came from. Not only is this a very tedious task but when people’s numbers don’t match, there is a lack of confidence in the data and rightfully so. But it wastes not only the analysts time but it also leads to wasted time on managements part when they argue over who’s data is right and could potentially lead to costly and incorrect decisions.
On top of that since 90% of an analysts time is spent gathering, cleaning, integrating and justifying their reports. This is time that keeps analysts away from actually getting to analyze the data and gain new insights. This is why our enterprise solution offers companies a very fast ROI.
POSmart streamlines the cleansing and integration of all these various data sources and puts them into a common repository where the data is clean and reliable. It’s also in a format that’s designed for easy reporting. Reports can be automated to send to users and power users also have an easy to access database where they can do ad-hoc querying. The result is far more productivity both from the analysts position as well as more productive management meetings. In addition, you have more access to reliable information than you’ve ever had. This data will help you know more about your business, provide more value to your management, understand your customer’s better and provide more insights to generate more profits.
The POSmart architecture is open. The data base is independent and can reside on Teradata, Oracle, DB2 and most others. If you can imagine everything on the left side of the white line exists at all our customers. This is just a sample of data sources, but you get the concept. On a nightly basis our BlueSky Integration Studio looks for new data. Any new data is then brought into our staging area where we cross reference the retailer data with your internal data, apply business rules, we align internal data with external data and we load the data into the POSmart data model where end users can easily access and query the data. Reports that were scheduled to run are then cached on the server so when end users come in in the morning, all their reports are already there and waiting for them. In addition, analysts are able to create new reports.
CPG companies spend as much as 20% of sales funding trade promotion activity. We generate baseline sales to understand lift on promotion as well as evaluate total cost to achieve that lift to uncover ROI for both the retailer and CPG company.
Do all stores need a secondary display location when on promotion? Do all stores achieve incremental sales on promotion?
Post Promotion Analytics
Allows analysis of trade spend ROI
Aligns trade funds with consumer events
Trade Promotion Optimization
Takes Post Promotion Analysis further with predictive analytics which identifies optimal promotions to run
The client estimated that implementing a post event trade spend analysis tool would increase sales by a MINIMUM of $15 mil/yr (0.32% of gross sales), which translated to a $2.5 mil/yr increase in profit gain (or 0.054% of gross sales), easily covering the cost of the implementation in the first year.
The above ended up being a very conservative estimate. The realistic estimate of sales increase was 6% ($276 mil/yr), which would proportionately translate to $46 mil/yr in profit gain (or 1% of gross sales).
Promotion Execution
Baseline Sales
Uplift & ROI Analysis
Optimized Spend Allocation
Cannibalization & Halo Effect Analysis
Pantry Stocking/Post Event Dip
Trade Promotion Optimization
How are consumers shopping?
Do you have a standardized calendar and KPIs to evaluate your total business using all available retailers?
Are you seeing volume shifts across channels? Does that align with your strategy/spend/efforts?
What is your total cost to serve a retailer?
Are you and the retailer both profitable?
So please join us next Wednesday, July 16th for our 2pm Demo. You can register by going to our website and clicking on events. You will also receive a follow up email with a way to register as well. In that demo we will cover POSmart and TradeSmart as well as PromoPro, baselines, explain why shipments are needed in more detail and go over some TradeSmart Case studies.
We would like to thank you all for joining us today. We invite you to follow our Relational Solutions training blog by going to our website at relationalsolutions.com. We also suggest you join the Demand Signal Repository Institute on LinkedIn. You can also see more training video’s on our Relational Solutions, YouTube channel. We ask that you follow Relational Solutions and Janet on Twitter, LinkedIn and Facebook and connect with us on LinkedIN. You are also free to contact us via phone and email. Thanks again for joining us and we look forward to your participation next Wednesday!