Professional Documents
Culture Documents
Understanding Analytics
AI and ML
Analytics Types, Data Types
Using analytics for Decision Making
What is our approach
Tools
Data Volume
Coding Language
What is our approach
Understanding of Concepts
Examples and Cases
Application from decision making view point
Strategic vs Operational application of analytics
Analytics in Business Scenario
Business Analytics
• Business analytics refers to a broad use of various quantitative
techniques
• Statistics, data mining, optimization tools, and simulation
• Supported by the query and reporting mechanism to assist
decision making
• Closed-loop framework for continuous process improvement
through monitoring and learning.
• Business analytics also helps the decision maker predict the
future business activities
• Based on the analysis of historical patterns of past business
activities.
Goal of Business Analytics
• Gaining insights into business practices and customer
behaviors:
• Business analytics is designed to transform unstructured, non-
standardized big data originated from multiple sources into meaningful
information helpful for a better business decision.
• Improving predictability:
• By deriving insights into customer behavioral patterns and market
trends, business analytics can improve the organization’s ability to
make demand forecast more accurately.
• Identifying risk:
• With growing complexity and uncertainty resulting from the
globalization of business activities, risk cannot be managed without
identifying it and then preparing for it. Business analytics can function
as an early warning system
Goal of Business Analytics
• Improving the effectiveness of communication:
• With the query and reporting mechanism of business analytics, quick
reporting procedures, user friendly report and scenario development
can be made.
• Enhancing operating efficiency:
• By aiding the decision maker in understanding the way business works,
business analytics can decrease the chances of making poor investment
decisions and misallocating the company’s resources and thus would
help improve the company’s operating efficiency.
Analytics in Business World
Dynamic Pricing
Product Recommendations
Analytics in Business World
They go so deep with utilizing data that they even follow if famous musicians are
playing at the Madison Square Garden so they could adjust their rates at their nearby
hotels.
To create a better customer experience, facial recognition check-ins so their guests don’t
need to wait at the reception desk anymore, and the hotel gathers even more valuable
information.
Amazon Echos into the rooms – this allows guests to make Alexa handle everything that
was previously handled by the reception staff. Now guests can get all the information
they want, while Marriott gets the knowledge of their customer’s preferences, needs
and potential concerns.
Analytics in Business World
They entered this saturated market of food delivery and, thanks to the data they
collected from being a taxi giant
They wanted to be recognized as a delivery that always brings food while it is still warm,
so they tried to model the physical world in a way that would allow them to be as
accurate as possible when predicting the time of food delivery.
To make this endeavor work, they also collected the data of how much time it usually
takes to prepare a certain meal, so they could pinpoint the exact time when the delivery
person should come and pick it up.
they are going deep with this that they even employed meteorologists to help them
predict what the weather will be like and how it will affect the delivery.
Analytics in Business World
Drive-thru with digital menus that change based on a variety of factors – from the time
of the day, to weather, and to historical sales data. That way they can offer their clients
a cold beverage on a hot day or maybe a coffee with their breakfast menu
Analytics in Business World
Starbucks uses customer data to recommend products to their loyal customers, create
better marketing campaigns and new menus, as well as decide where they’ll open their
next store. This system is so organized that it will offer their customers products based
on the season, weather and location they are at.
They also send out personalized emails with offers to customers who haven’t visited the
store in a while, so they can re-engage them or send them discounts.
Analytics Types
• Descriptive Analytics :
• What is happening
• Real time data, past data, visualization of future to some extent
• Does not explain root cause
• Diagnostics Analytics :
• Why is it happening
• Drill down to find factors
• Predictive Analytics :
• What is likely to happen in future based on previous trends/patterns
• Algorithms to make recommendations
• Data Quality, Missing Values determine accuracy
Analytics Types
• Prescriptive Analysis :
• Best course of future action
• Eliminate or by-pass issues
• Google map to avoid traffic congestion
• Cognitive Analysis :
• Combines a number of intelligent technologies
• AI, ML algorithms
• Apply human brain-like intelligence to perform certain tasks
• Draws conclusions
• Iterative and continuous learning, adaptation
Approach to Analytics – Case discussion
Data is going to keep growing and technology options will
follow the same trajectory. Organizations shouldn’t run
from this new digital reality, but learn to embrace it by
adopting and adapting their analytics strategies to remain
competitive. By applying the power of data and analytics
techniques such as machine learning, a firm can make
smarter, faster decisions for their business and its
customers, and actively disrupt their industry
A popular example would be Target's statistician Andrew Pole using personal
identifiable information (PII) to notice women on baby registries buying
large quantities of unscented lotion a the beginning of their second
trimester. The data also showed some time in the first 20 weeks women
purchased large amounts of calcium, magnesium, and zinc.
Using all of this shopping behavior data, Pole could then assign a pregnancy
prediction score to customers based on the purchase and purchase volume
of about 25 different products in-store
Whether you know it or not, you as a shopper are sharing highly detailed
buying patterns with retailers as you shop on a daily basis. You unknowingly
(or knowingly if you work in this field) provide this data through customer
IDs tied to personally identifiable information (PII) such as credit cards,
emails, and loyalty card numbers.
These national and global retailers are spending lots of time analyzing your
data to determine how to sell you even more. They leverage data points to
sell to you without even you knowing it.
Over time, the Target advertising team learned to find out consumers get really
freaked out if the store knows them too well, especially intimate details such as
being pregnant. The marketing department understood they needed to mix these
targeted advertisements in with others to lessen the "creep factor."
So now that same pregnant customer sees ads in their sales booklet for not only
diapers and cribs but also other foods, household items, lawn care items, and
clothes. The mix and match still allow for proper targeting but it's not as direct
and does not seem as invasive.
It's not enough to just collect this data from a variety of sources as an analyst.
Finding ways to tie databases together and spending the time to fully examine
the data to identify trends and commonalities is the key to generating these
types of insights.
Make sure your company is not only "data-heavy" but also "insight-heavy."
What is Supply Chain
• A supply chain consists of all parties involved, directly or indirectly, in
fulfilling a customer request.
• The supply chain includes
– Manufacturer
– Suppliers
– Transporters
– Warehouses
– Retailers
– Customers
• The supply chain includes all functions involved in receiving and filling a
customer request.
• These functions include
– New product development
– Marketing
– Operations
– Distribution
– Finance
– Customer service
SC Analytics
SC Analytics
SC Analytics
According to statistics, 97% of supply chain analysts believe
that big data can be extremely useful for the supply chain, but
only 17% are actually using it.
https://www.youtube.com/watch?v=he5I6ByoaB4
https://www.youtube.com/watch?v=lF1CzIfRYAw
https://www.youtube.com/watch?v=VZ5jU6BNjQs
https://www.youtube.com/watch?v=IFMp-zi_nGc
AI/ML in supply Chain
• Predictive Analytics - Using machine learning models, companies
can enjoy the benefit of predictive analytics for demand
forecasting.
https://www.youtube.com/watch?v=3kAByQnkctM
https://www.youtube.com/watch?v=xBzgdbG1Rn0
• The problems in front of you in business will often be
relatively straightforward.
• Business issues aren’t easy or black and white.
• They’re filled with a high number of constantly moving
variables that make any sort of predictability difficult to
achieve.
• Solving these problems that are seemingly impossible to
analyze is achievable, but it takes a decidedly different
approach than what many entrepreneurs and leaders are
used to.
DON’T ANALYZE AND MODEL TO DEATH, JUST GO DO IT
• Digital means that you can do this sort of real-world, experiment-
based probing much more easily.
• Problems can be analysed with simulations and what-if scenarios
without actual infrastructure
• Helps to fast forward time
• You can reduce your ideas to really small things so that you can
move through many rapid iterations.
IN THE LONG-TERM, HUMILITY YIELDS SUCCESS
• Businesses struggle against not providing or getting concrete
answers to today’s complex problems.
• Adopt a Just Do It strategy where it makes sense
• Accept “not knowing” is ok
• Being the smartest in the room isn’t as valuable as it used to be
• Execute what works for you instead of relying on copying anyone
else
• Empowers teams to take action and trust themselves and each
other, even if they don’t know precisely how they’ll get to the
finish line.
• Analytics is not about tools
• Analytics is not having abundant data
• Analytics is right usage of tools and data to make effective
decisions
• Analytics is about Mindset to have a systematic approach to make
well informed decision in a given situation of uncertainty
• DO NOT GET PASSIONATE ABOUT DATA OR TOOLS
• DO NOT THROW COMMONSENSE OUT OF THE WINDOW
Analytics Types
Descriptive Analysis Diagnostic Predictive Prescriptive Cognitive
Quickly troubleshoot a problem Robots are able to sort, pick and
within the distribution center, Run simulations to test the impact package silicon chips, car
such as an increase in no-reads of operations decisions, such as assemblies and other parts and final
Fill rates Goal setting
or an increased percentage of reducing staffing or adding more products, based on quality
products requiring human sorting equipment. assurance standards, and reject
intervention. those that fall short in some way.
Suggest new supply chain Respond more quickly to
AGVs (autonomous guided
configurations to address a changing conditions, such as how
Planning, vehicles) can avoid or navigate
problem at a particular line or to change staffing in the
Supplier lead times scheduling and around unexpected obstacles, to
facility – say, by adding more distribution center to deal with an
forecasting deliver products where they need to
sorting equipment or increased unexpected surge in unreadable
be in a distribution center.
staffing temporarily. bar codes.
Cameras and other sensors in
Take corrective action more
sortation tunnels can determine if a
quickly. For example, if
shipping label was positioned
throughput in a particular facility
Identify non-standard products, properly but was obscured by the
is slowing down, predictive
and provide insight on how to Performance and manufacturer’s packaging.
Inventory dollars analytics can suggest or even
optimize the supply chain to risk management Organizations sometimes use
initiate responses, such as
handle them more efficiently. cognitive analytics to make
shifting volume away from a
decisions on how to repackage
tunnel that is experiencing the
products to prevent such problems
slowdown.
in the future.
Sales and operations
Fraud reduction
data
Social media
engagement and
conversion
Survey results and
other forms of user
generated content
What is Supply Chain
• A supply chain consists of all parties involved, directly or indirectly, in
fulfilling a customer request.
• The supply chain includes
– Manufacturer
– Suppliers
– Transporters
– Warehouses
– Retailers
– Customers
• The supply chain includes all functions involved in receiving and filling a
customer request.
• These functions include
– New product development
– Marketing
– Operations
– Distribution
– Finance
– Customer service
Supply Chain Objectives
Supply Chain Surplus =
Customer Value − Supply Chain Cost
Supply Chain Drivers
Competitive Advantage
Operations Strategy Issues
Value Chain
Value Chain
Where is India heading?.....
Supply Chain Analytics
Supply Chain Analytics
Supply Chain Analytics
SCM Trend
Agile Network and Technology
SCM Vision
Sumantra Ghosh, VP of media sales
Partha Roy, General Manager Tech Deployment
Mukesh Tadikonda, Studio Manager
Karumuri Sekhar, Studio Techn incharge
Above team sets out from Pune to Mumbai in an aston martin. They are to meet with a Japanese team for a
$10 Million TV documentary series on the Indo-Japan Industrial collaboration and opportunities.
Unusually though Sumantra’s machine of pride the aston martin fails enroute. Its almost sure they will be
delayed for the meeting as it’s a trip with tight schedule. Sekhar and Mukesh are engineering batch mates and
they share passion to re-model cars. They believe they can check out and if it’s a small glitch it can be fixed by
them. Sumantra doesn’t want his car to be handled by people who are not experts specific to that brand and
tenchology though he appreciates Mukesh and Sekhar passion. Partha is a tech minded person but his
thoughts are on the upcoming meeting as he already sees the kind of challenge in pipeline and time pressure
his team would be subjected to once this contract is signed. Sumantra is worried that any delay in getting to
meeting would mean lost opportunity. He asks Partha to look for some alternative ways to get to Mumbai on
time. Sekhar and Mukesh are more eager to try their tech skills on repairing the car. Sumantra walks into a
café nearby to phone up the dealer and to focus on his work. Partha entrusts Mukesh the task for finding a
transport and gets back to his mental preparedness work on deployment. Sekhar walks around the vehicle
eager to show is automobile mechanical skills!!!!
The process of creating supply chain analytics typically starts with data scientists who understand a particular
aspect of the business, such as the factors that relate to cash flow, inventory, waste and service levels. These
experts look for potential correlations between different data elements to build a predictive model that
optimizes the output of the supply chain. They test out variations until they have a robust model.
Supply chain analytics models that reach a certain threshold of success are deployed into production by data
engineers with an eye toward scalability and performance. Data scientists, data engineers and business users
work together to refine the way these data analytics are presented and operationalized in practice. Supply
chain models are improved over time by correlating the performance of data analysis models in production
with the business value they deliver.
SC Analytics Future
The amount of data produced by the supply chain is expanding every year. The new
technologies and devices that are introduced to the various stages: production, transportation,
and selling, create a high number of data that can be transformed into valuable insights. The
great example that illustrates this tendency is IoT. According to the research, the data produced
by IoT is set to increase by almost 500% to 80 zettabytes by 2025. The IoT implications in the
supply chain are expected to play a significant role and become one of the dominating
technology for supply chain risk management software.
This data can be generated by the RFID tags and sensors in the warehouse to detect the precise
location of items and track their movement beyond it. The valuable information is like where
the goods are coming from, their expiration date, etc. Moreover, during transportation, the
data from GPS and that same RFID sensors can help to increase transit visibility of the shipped
items and to optimize routes for the potential delay prevention.
Furthermore, this data helps to obtain crucial information like temperature and humidity that
can affect items’ quality during their transportation. Combined with environmental data like
weather conditions and traffic data, this information can help significantly in managing supply
chain risks during transportation. But this is only one of the examples of how data analytics
affects supply chains – the possibilities are endless.
Thinking Shift
Analytical Thinking Approach
• Fact-finding and checking through thought experiments:
• Thought experiments involve hypothesizing “what-if” scenarios in the
imaginary world and allow us to see the outcome (what will happen) if
we select a certain decision
Understanding Forecasting
Analytics in Forecasting
Forecasting Techniques
Modern day forecasting Challenges
• One of the biggest challenges for business executives today is demand volatility in
relation to demand forecasting.
• Whereas data availability continues to increase, customer purchase patterns are
becoming increasingly complex
• That makes prediction harder
• There are too many factors influencing demand, ranging from weather fluctuations
to posts by social media influencers, causing customers to frequently changing
their minds.
• things that will reshape customer intentions will mostly happen quite
unexpectedly. Traditional Forecasts are only as accurate as the data, models,
resources and people that have to interpret them.
• how can we respond to these challenges?
AI and ML a solution ??
Are AI and machine learning the future for demand forecasting?
Demand forecasting is at the heart of every retailer’s supply chain, and for good
reason: the demand forecast is central to sales, profitability and the customer
shopping experience, and it has a ripple effect throughout the supply chain.
Yet despite a wealth of established demand forecasting solutions and
methodologies, retailers struggle to produce accurate, timely demand forecasts.
Danone uses many promotions and media events. More than 30% of
the total volume is sold through promotional offers such as discounts
and leaflets, so the demand forecasts were somewhat ad hoc
‘RIGHT DATA’ ANALYTICS WILL SURPASS BIG DATA ANALYTICS AS A KEY TREND
• Current reality is not knowing what sales will be, purchase new equipment despite
uncertainty about demand for products, and make investments without knowing what
profits will be.
• Managers are always trying to make better estimates of what will happen in the future
in the face of uncertainty.
• Making good estimates is the main purpose of FORECASTING
Demand Forecasting
https://www.youtube.com/watch?v=HQ7C
I7JlXJQ
Forecasting Challenges
Data Normalization
Year Sales Quantity Year Sales Quantity
2000 28 2000 28
2001 33 2001 33
2002 37 2002 37
2003 39 2003 39
2004 40 2004 40
2005 38 2005 38
2006 41 2006 41
2007 42 2007 42
2008 47 2008 47
2009 41 2009 41
2010 92 2010 42
2011 51 2011 51
2012 50 2012 50
2013 48 2013 48
2014 54 2014 54
2015 56 2015 56
2016 61 2016 61
2017 64 2017 64
2018 66 2018 66
2019 71 2019 71
Data Normalization
Year Sales Quantity Year Sales Quantity
2000 28 2000 28
2001 33 2001 33
2002 37 2002 37
2003 39 2003 39
2004 40 2004 40
2005 38 2005 38
2006 2006 38
2007 2007 40
2008 2008 42
2009 2009 44
2010 2010 46
2011 2011 48
2012 2012 50
2013 2013 52
2014 54 2014 54
2015 56 2015 56
2016 2016 59
2017 2017 62
2018 66 2018 66
2019 71 2019 71
Man Machine Collaboration
Month 2017 2018 2019 2020
Jan 0 0 0 0
Feb 0 0 0 0
Mar 0 0 0 0
Apr 0 0 0 0
May 0 0 0 0
Jun 0 0 0 0
Jul 0 0 0 0
Aug 0 0 0 ?
Sep 0 0 0 ?
Oct 0 0 0 ?
Nov 0 0 0 ?
Dec 1 1 1 ?
Panel Discussion 1
Only customer who were buying is likely to
change technology?
Sales annually for this unit is $1Million
What will be your forecast?
Man Machine Collaboration
Month 2017 2018 2019 2020
Jan 0 0 0 0
Feb 0 0 0 0
Mar 0 0 0 0
Apr 0 0 0 0
May 0 0 0 0
Jun 0 0 0 0
Jul 0 0 0 0
Aug 0 0 0 ?
Sep 0 0 0 ?
Oct 0 0 0 ?
Nov 0 0 0 ?
Dec 1 1 1 ?
Panel Discussion 2
Customer has changed technology
There are other customers in the market
What will be your forecast?
AI Plus Human Intelligence
• However, no matter how smart the forecasting solution may be,
Human logic is still needed to evaluate the relevance of the
outcomes produced by AI solutions.
Delphi method
A forecasting technique using a group process that allows experts to make
forecasts.
Market survey
A forecasting method that solicits input from customers or potential
customers regarding future purchasing plans
Quantitative Methods
Quantitative Methods
Naive approach
A forecasting technique that assumes that demand in the next period is equal
to demand in the most recent period.
Moving averages
A forecasting method that uses an average of the n most recent periods of
data to forecast the next period.
Exponential smoothing
A weighted-moving-average forecasting technique in which data points are
weighted by an exponential function
Quantitative Methods
Trend projection
A time-series forecasting method that fits a trend line to a series of historical
data points and then projects the line into the future for forecasts
Linear-regression analysis
A straight-line mathematical model to describe the functional relationships
between independent and dependent variables
Supply Chain & Logistics Analytics – Session 3
Forecasting and Accuracy Measurement Part II
Chapter 1, 2, 3 from the Recommended Textbook
What will we cover?
• What is the lowest level in the hierarchy you should forecast at? What level of product
or customer gives the best balance of forecast quality vs planning processes you want
the forecast to feed into.
• What should be the time granularity of your forecasts and the proper forecast offset
(lag)? Is it necessary to generate weekly forecasts – especially when the timing of
demand is difficult to predict – or is it better to do monthly or even quarterly rolling
forecasts
• Choice of algorithms
The average that they yield acts like an expert opinion and turns out
to be very precise.
Ensemble Methods
More Retail Limited Mini Case
More Retail Ltd. (MRL) is one of India’s top four grocery retailers,
with a revenue in the order of several billion dollars. It has a store
network of 22 hypermarkets and 624 supermarkets across India,
supported by a supply chain of 13 distribution centers, 7 fruits and
vegetables collection centers, and 6 staples processing centers.
With such a large network, it’s critical for MRL to deliver the right
product quality at the right economic value, while meeting
customer demand and keeping operational costs to a minimum.
With such a large network, it’s critical for MRL to deliver the right product quality at
the right economic value, while meeting customer demand and keeping operational
costs to a minimum. MRL collaborated with Ganit as its AI analytics partner to
forecast demand with greater accuracy and build an automated ordering system to
overcome the bottlenecks and deficiencies of manual judgment by store managers.
MRL used Amazon Forecast to increase their forecasting accuracy from 24% to
76%, leading to a reduction in wastage by up to 30% in the fresh produce category,
improving in-stock rates from 80% to 90%, and increasing gross profit by 25%.
More Retail Case Discussion
An ML-based approach unlocked the true power of data for MRL. With Forecast,
two national models were created for different store formats, as opposed to over
1,000 traditional models that was existing.
Forecast also learns across time series. ML algorithms within Forecast enable
cross-learning between store-SKU combinations, which helps improve forecast
accuracies.
Additionally, Forecast allows to add related time series and item metadata, such as
customers who send demand signals based on the mix of items in their basket.
Forecast considers all the incoming demand information and arrives at a single
model. Unlike conventional models, where the addition of variables leads to
overfitting, Forecast enriches the model, providing accurate forecasts based on
business context. MRL gained the ability to categorize products based on factors
like shelf life, promotions, price, type of stores, affluent cluster, competitive store,
and stores throughput.
More Retail Case Discussion
Air Travel Forecast Mini Case
Passenger traffic at the world’s airports is projected to reach 6.5
billion in 2022, down by -28.9% on 2019 levels, according to
Airports Council International (ACI) World.
What factors can play key role in forecasting of Global Air Travel in
2022 – 2023? Keeping analytics in view, how can this situation be
forecasted?
Backup Slides
Supply Chain & Logistics Analytics – Session 4
Inventory Management Concepts
Chapter 11 and 12 from the Recommended Textbook
Session 4 Topics
Types of Inventory
Cycle Stock
Safety Stock
Product Classification
Inventory in Supply Chain
Amazon Inventory Case
When Jeff Bezos opened his revolutionary business in 1995,
Amazon.com was intended to be a “virtual” retailer—no
inventory, no warehouses, no overhead—just a bunch of
computers taking orders for books and authorizing others to fill
them.
I = DT
Where,
I is Inventory
D is Demand
T is Lead time
Inventory Types
• Raw materials
• Work-in-process (WIP)
• Finished goods
• Maintenance, repair and operating supplies (MRO
goods)
• Transit inventory/Pipeline
• Safety stock / Buffer inventory
• Anticipation Inventory
• Cycle inventories / Lot-size inventories
A total-cost optimization approach to supply chain planning ITC Limited is a multi-billion-
dollar conglomerate based in India with a diversified presence in fast-moving consumer
goods, hotels, paper and packaging, agriculture and information technology.
The company’s packaged foods business is experiencing fast growth, with biscuits (cookies)
representing its largest single category.
The biscuit segment experiences high demand volatility and materials cost fluctuations.
ITC recognized that its spreadsheet-based tools did not account for its continuously changing
cost structures.
The company needed a more powerful tool that would allow it to take a total-cost
optimization approach toward its strategic, as well as tactical, supply chain planning.
Challenges
• ITC’s spreadsheet tools did not account for its continuously changing cost structures.
Therefore, they could not see the macro effects of the supply chain decisions they were making.
• Cost optimization is complex at ITC due to frequent manufacturing changeovers required to
produce its more than 120 distinct biscuit SKUs across its network of 17 factories and the
dramatic price fluctuations in the agricultural sector that affect their materials costs.
• ITC’s spreadsheet-based planning did not allow them to make effective medium and long-
range strategic plans
AI in Inventory Management
AI-powered inventory management can provide a solution to human error by letting the
computer do the math.
Determining the right amount of stock, in the right place, at the right time at the right
costs as well as the right price.
That is what inventory management is about in business terms and could be determined
automatically by the AI application.
The outputs of the model reflect the most optimal decisions that can be taken.
Demand Rate
Safety Stock and Inventory Level
Service level increase towards 100% leads to disproportionate increase in safety stock
Inventory Related Costs
Holding Cost
• Cost of Capital
• Obsolescence / Spoilage
• Handling Cost
• Occupancy Cost (Basic storage cost)
• Miscellaneous.. Insurance, Damages, Security overheads….
Ordering Cost
• Buyer Time Cost
• Transport Cost
• Receiving Cost
• Other costs… Insurance, Taxes, damages in transit…
Safety Stock > Service > Cost
“More is the market leader in the ‘Fresh’ category in food and grocery in India. To run a
viable business, More needs to simultaneously manage in-stock availability of fresh
produce, while minimizing wastage. To balance these competing priorities, More partnered
with AWS and Ganit, a data science consulting company, to build and deploy a demand
forecasting and automated ordering system built around Amazon Forecast. We needed to
build a very granular forecast at store-item-day level, therefore we prioritized the
development effort based on ABC-XYZ framework.
Inventory Case Discussion
The store-item combinations were plotted on a 3x3 matrix: ABC axis of sales saliency (A –
high, B-Medium, C-low) and XYZ axis of forecastability (X-easier to forecast, Z-difficult to
forecast) based on historical pattern. As expected, forecast accuracy of items in ABC-XY
buckets was much superior to the Z bucket. However, for combinations in the Z bucket,
Amazon DeepAR+ significantly outperformed traditional methods like exponential
smoothing yielding an incremental 10% forecast accuracy. This was possible because of
Amazon Forecast’s ability to learn other SKUs (XY) patterns and apply to highly volatile
items in the Z bucket.
Using Amazon Forecast, we have been able to increase our forecasting accuracy from 27%
to 76% reducing wastage by 20% for the fresh produce category. Amazon Forecast provides
a distribution of forecasts which helped us optimize our under and over forecasting costs
leading to stock-outs at 3% and improved gross margins. This makes it easier for our store
managers to place more accurate purchases orders by looking at the daily forecasts. We
are now expanding the model to other categories, iterating with additional related
datasets, and adding newer data to Amazon Forecast to continuously improve the model
accuracy.”
Product Classification
Cross Mapping of Products
Inventory and Planning Decisions
Inventory Strategy Approach with Analytics
How much, where, and in what form should inventory for a product be held in the supply
chain?
How can I evaluate the impact of network strategy and variability on inventory?
How can I explore trade-offs between service level objectives and inventory holding costs?
Where should I set push vs. pull boundaries (also known as decoupling points) in my
supply chain (that is, make-to-stock vs. make-to-order)?
What inventory should I hold in order to maximize fill rates under a top-down budgetary
constraint for inventory investment?
The AI tool has been trained to recognize, identify and count the different Coca
Cola products in the coolers.
The tool could combine this data with information received from demand
forecasting and automatically calculate an order to restock.
24/7 access to data: always insights into the data of the inventory gives practical
benefits to get a competitive advantage.
Coca Cola Case Discussion
Several key benefits for the use of AI in inventory management are:
Easy integration with current systems: most companies already make use of ERP
and CRM systems, which integrate easily with an AI application in inventory
management
“At Meesho, we have a lot of products with short life spans and it is important for us to
react to key metrics associated with product performance and manage our inventory
optimally. By using Amazon Forecast, we were able to predict the demand forecasting of
products at weekly/daily with an increase in forecasting accuracy of 20% in comparison to
our existing solution. Amazon Forecast provides easy to use APIs that helped us to easily
build an automated system in half the time it would have taken for an in-house model. We
have achieved promising results so far with Amazon Forecast on our current inventory and
we plan to continue leveraging it to improve forecasting accuracy on our ever increasing
assortment of products.”
Demand Solutions in 3M
• Multi Division, Multi SKU set up
• Varied customer segments
• Healthcare, Industrial, Safety, Consumer, Automotive
• Diverse Markets – Mature Customers, Smaller customes
• High level of new product introduction and new market penetration
• Marketing and sales depended on each other for market inputs
• Last minute promotions, price offs to counter competition move
• Distributors data sharing was not regular, not standard formats
• Demand solutions implemented by Supply Chain
• Forecasting and inventory management tool
• Required sales and marketing to plug in data
• S&M depended on distributor input
• New generation sales team wanted more automation with distributor plugging in data
directly into the system
Cycle Stock Calculation
How can I evaluate the impact of network strategy and variability on inventory?
How can I explore trade-offs between service level objectives and inventory holding costs?
Where should I set push vs. pull boundaries (also known as decoupling points) in my
supply chain (that is, make-to-stock vs. make-to-order)?
What inventory should I hold in order to maximize fill rates under a top-down budgetary
constraint for inventory investment?
Total Cost
Different stages exploit economies of scale to lower total cost.
The costs considered include material cost, fixed ordering cost,
and holding cost.
Ordering Cost
• Buyer Time Cost
• Transport Cost
• Receiving Cost
• Other costs… Insurance, Taxes, damages in transit…
Economic Order Quantity (EOQ)
Economic Order Quantity (EOQ)
Origin UK
Destination India
Lot Size
Costs influenced by Lot Size are
1. Price
2. Holding Cost
3. Ordering Cost
There are numerous formulas for safety stock based on lead time, variability and service
level target. These approaches fail to consider the interdependencies across nodes in the
supply chain. The closer to the final customer, typically the more intermittent and variable
the demand signals become. As we look into the upstream supply chain, pooling demand
signals results in smoother demand that is easier to forecast. The trade-off between
inventory form and function is optimized and the right policy to stock each node is
determined. This solution must include a holistic consideration of in-transit, cycle stock,
WIP, quality holds, and safety stock.
Backup Slides
Lot Size
Costs influenced by Lot Size are
1. Price
2. Holding Cost
3. Ordering Cost
• A store orders 80 printers from the manufacturer each time they place an
order. The lot or batch size in this case is 80 printers.
• Daily sales is 4 printers
• It takes an average of 20 days before the store sells the entire lot
• The computer store holds an inventory of printers because it purchases a lot
size larger than the store’s daily sales.
Supply Chain & Logistics Analytics – Session 6
Analytics in Inventory Management MOQ, EOQ,
Inventory Optimization Part II
Chapter 11 and 12 from the Recommended Textbook
Session 6 Topics
“At Meesho, we have a lot of products with short life spans and it is important for us to
react to key metrics associated with product performance and manage our inventory
optimally. By using Amazon Forecast, we were able to predict the demand forecasting of
products at weekly/daily with an increase in forecasting accuracy of 20% in comparison to
our existing solution. Amazon Forecast provides easy to use APIs that helped us to easily
build an automated system in half the time it would have taken for an in-house model. We
have achieved promising results so far with Amazon Forecast on our current inventory and
we plan to continue leveraging it to improve forecasting accuracy on our ever increasing
assortment of products.”
Demand Solutions in 3M
• Multi Division, Multi SKU set up
• Varied customer segments
• Healthcare, Industrial, Safety, Consumer, Automotive
• Diverse Markets – Mature Customers, Smaller customes
• High level of new product introduction and new market penetration
• Marketing and sales depended on each other for market inputs
• Last minute promotions, price offs to counter competition move
• Distributors data sharing was not regular, not standard formats
• Demand solutions implemented by Supply Chain
• Forecasting and inventory management tool
• Required sales and marketing to plug in data
• S&M depended on distributor input
• New generation sales team wanted more automation with distributor plugging in data
directly into the system
GMROI Model
GMROI Model
Comifar Case Study
Exploding size and complexity leads to a moment of truth
With 21 distribution centres and three hubs over the whole of Italy, Comifar
Distribuzione distributes 20K orders per day from over 150K SKUs of
prescription drugs, over-the-counter medications, nutritional supplements,
and much more. With 19% of the market, they are the biggest pharmaceutical
distributor in the country.
Ordering Cost
• Buyer Time Cost
• Transport Cost
• Receiving Cost
• Other costs… Insurance, Taxes, damages in transit…
Economic Order Quantity (EOQ)
EOQ for Multiple Products
• In current business world there is high chances of multi
product shipments
• That makes the EOQ calculation complex
• Consolidation from multi suppliers
• Milk Run across suppliers for more frequent supplies
• A store orders 80 printers from the manufacturer each time they place an
order. The lot or batch size in this case is 80 printers.
• Daily sales is 4 printers
• It takes an average of 20 days before the store sells the entire lot
• The computer store holds an inventory of printers because it purchases a lot
size larger than the store’s daily sales.
Inventory Optimization Case Study
Food and beverage products have a short shelf life, there’s a big upside to getting inventory levels
just right. So when our client, a Fortune 500 food manufacturer, noticed that their days of on-hand
inventory for some SKUs was encroaching upon the shelf life of those products, they knew
inventory levels were too high, and they needed to improve their inventory planning process.
Getting hypothetical
To help the company fully understand what was driving its recommended inventory
targets, Chainalytics developed a custom tool with “what if” capabilities which allows
users to change inputs like demand, lead time, and service levels (i.e., fill rates) to see
the potential impact on cost, and working capital.
The what-if tool has an MS Excel front-end and uses the R programming language for
data processing and computation. This combination of open source and commercially
available tools provides a familiar user experience for the planner and is easy to
maintain internally by the company’s IT department. Working with the manufacturer’s
planning team, Chainalytics supported an initial showcase/pilot of the tool to later roll-
out to the rest of the planning group.
Inventory Optimization Case Study
Problem solved
This inventory target setting engagement delivered many benefits for the
manufacturer, including:
Product Segmentation: The client company is now able to set custom break-points for
velocity and understand variabilities as well as segment based on unit standard costs.
With full visibility into how SKU demand behaves based on velocity, variability and cost,
the manufacturer can set its desired case fill rate setting in each individual product
segment.
Inventory Target Setting: The most important output of the tool developed by
Chainalytics is the calculation of Target Stock. Intelligence into ideal levels of inventory
to meet demand is broken down into “cycle stock” — which is the required level to
fulfill demand—and a “safety stock”—which buffers against volatility in demand or
supply. Together, these numbers ensure the manufacturer will be able to satisfy
demand without running out of stock or carrying excess inventory.
Inventory Optimization Case Study
Problem solved
What-if Reporting: Chainalytics’ solution gives the manufacturer the ability to compare
scenarios by location, segments, SKUs, etc. and to create waterfall/bridge reporting for
sensitivities.
The company can now easily calculate how to achieve desired output by making
changes to multiple inputs. Variances in that output can easily be explained by
identifying which inputs were altered.
Ongoing Analysis: The tool delivered by Chainalytics can be used at regular intervals to
account for changes in the future. This allows the company to recalculate inventory
targets several times throughout the year based on the planning cycles to keep its
operations optimised for changes to real-world scenarios as they occur.
Working with hundreds of distributors across India and China to deliver products to
retailers requires accurate and reliable insights into the drivers of inventory. Our client
is now equipped to plan inventory for efficient distribution without waste.
Indicative Methods
Forecastability Matrix
Supply Chain & Logistics Analytics – Session 7
Analytics in Sourcing, Transportation, Warehousing
Chapter 11 and 12 from the Recommended Textbook
Session 7 Topics
Sourcing
Warehousing
Transportation
Analytics in Sourcing, WH and Transportation, M & S
Total Cost of Ownership
Strategic Sourcing
Good procurement procedures help organizations get goods and services they
need
Great procurement procedures aim for the lowest cost possible.
Procurement focuses on the immediate rewards instead of what's good for
the business over the long term.
The key to strategic sourcing lies within current procurement data that can
unlock even more benefits for your business.
Strategic Sourcing Analytics
1. Start with a big picture analysis. Look at all purchasing data, then break it
down into different departments and locations. Analyze different spending
categories, as well as patterns in buying. The more granular the data, the
more information you have to analyze.
3. Analyze the supplier market. What suppliers can provide your business with
the Maintenance, Repair and Operations (MRO) products needed to keep
buildings running and in repair? Develop portfolios about potential suppliers,
including their strengths and weaknesses.
Strategic Sourcing Analytics
4. Define your supplier criteria. Know what the business needs from suppliers
so it's easier to determine who can meet the criteria and who can't.
Determine what qualities are most important in your suppliers and engage
them accordingly.
5. Negotiate with a list of select suppliers. Reach out to a carefully chosen list
and start the negotiation process for products under contract.
6. Incorporate the new suppliers into your procurement program. Make it easy
for contractors and employees to order from the new suppliers. This helps
eliminate maverick spending and increased costs.
7. Track the performance of all suppliers. Make sure suppliers are adhering to
terms and conditions and providing the optimum level of service. Continually
analyze their performance, as well. If they aren’t up to par, investigate
potential new suppliers who may be better suited to serve the business.
Strategic Sourcing Metrics
1. SPEND ANALYTICS - This one is crucial. Procurement teams
need to know how much is being spent, when, and where.
Without the data, procurement is often hit or miss.
Optimisation Data analytics reveals insights into the effectiveness, form and
productivity of the system and can be used for the purposes
of maintenance, operations and asset management decisions.
Visualisations Visualized data insights are tailored to specific roles and focus
areas to assist decision-making at all levels.
• Maintain a strong customer focus and ensure that goods arrive with
customers on time – wherever they are.
• Master new data types and harness data from third-party carriers to
optimise routes and build smarter transportation strategies.
• Cut costs and improve efficiency in any way possible to protect profit
margins – without sacrificing delivery speed..
Total Cost of Ownership Model
Total Cost of Ownership
Total Cost of Ownership
UK to UAE
ITALY to UAE
Should Cost Modelling
Trade off Concept
Trade off Concept
Trade Off
• Benetton requires retailers to commit to about 65 percent of their orders about seven
months before the start of the sales season.
• Benetton subcontracts production of this portion without uncertainty to low-cost
sources that have long lead times of several months.
• For the other 35 percent, Benetton allows retailers to place orders much closer to or
even after the start of the selling season.
• All uncertainty is concentrated in this portion of the order. Benetton produces this
portion of the order in a plant it owns that is very flexible.
• Production at the Benetton plant is more expensive than production at the
subcontractor’s. VOLUME BASED TAILORED SOURCING
ITC has a highly-integrated supply chain with ~100 factories (MF) and 60 major FG warehouses
(WH) around India. Packaged product meant for retail flows from factories to large regional
warehouses, that hold roughly 2-4 weeks of stock. These warehouses serve two functions,
inventory storage and order fulfillment to wholesale distributors (WD). Each warehouse touches
between 80-120 WDs. Like other warehouses in India, ITC’s warehouses run a largely manual, paper
based operation. And while it could have been more operationally efficient for ITC to operate fewer
warehouses, and assign more WDs to each, the indirect taxation structure in India (until before
2017) makes it expensive for them to do so.
The company’s CPG business is among the fastest growing in India. The company is constantly
adding new product categories and expanding existing product lines. These developments put a
huge operational and financial strain on the organization’s logistics.
ITC Case Discussion
Rising real-estate costs: Manual-picking and paper-accounting based warehouses means products are
stored in short-stacks (6ft) at ground level. Unlike warehouses in Europe which are 80-160ft tall, ITC’s
facilities are 10ft tall. Thus, floor space productivity is low.
Low product margins: ITC sells perishables that with low gross margins. Thus, warehousing and
logistics have tight cost bounds.
Large SKU assortment: With over 3000 SKUs, a long tail and regional/seasonal variants the SKU
assortment multiply the operational complexity of warehouse operations.
Poor inventory management: Warehouses store/fulfill orders based on paper based accounting
systems, often leading to pilferage, delays and errors.
Lack of skilled productive labor and escalating labor costs: Minimum wage is growing fast and skilled
labor is hard to find.
Lack of consolidation: With 60 warehouses around India, ITC’s customer base is fragmented, order
SKU/volume variation is high and downstream logistics is sub-optimal.
ITC’s direct costs are near/better than the industry benchmarks given the high levels of automation in
its manufacturing facilities and general focus on process excellence/optimization. Logistics –
warehousing and transportation – constitute the largest indirect cost. Management realizes it is
imperative to revolutionize warehousing if they had to keep ITC competitive going forward.
ITC Case Discussion
Phase 1
WMS integration with ERP systems (WH, MF, WD): WMS system speaks with a
fully integrated ERP system to pull master data, track inventory in transit and
identify inventory levels downstream.
Migration to 40ft racked warehouses: Migration from traditional ground level warehouses
to 6-8 level (3ft tall window) pallet storage rack warehouses. Effective inventory stored
per unit area grows 4x.
Basic automation with pallet trucks: Pallet trucks (PT) to convey pallets around the
warehouse, and Height-reach trucks (HRT) to place/pick pallets in to rack-storage
locations.
WMS driven, hand-held terminal (HHT), Barcode assisted operations: WMS generated
optimum storage tasks (instructions to PTs & HRTs to move pallets between locations),
optimum case-picking routes for manual order picking, HHTs for pick validation/integrity.
Enabler: In 2017, the hallmark GST-bill goes into effect altering the indirect taxation
structure [5][6]. Organizations can now fully offset tax paid on cross-border sales against
tax liabilities in manufacturing region. ITC can now consolidate warehouses to improve
efficiency. Thus, the CEO has set up a task force to establish phase 3.
ITC Case Discussion
Phase 3
Consolidation – mega warehouses cum factories: ITC has formulated a NPV-positive automation
model and plan to build mega warehouses, co-hosted with factories, in a phased manner to meet
2025’s logistics requirements.
80-100ft tall highly automated warehouse schema: Each warehouse is expected to store and fulfill
about 8-10x the current product volume handled by a large warehouse.
Migrate to a more advanced WMS cum Equipment Control System (ECS): The organization is investing
in a plethora of IT Hardware and advanced WMS/ECS software [9] to instruct automated storage
systems, conveyors, sorters, palletizers and other equipment.
“Our long-term goal is to reach the retail points directly from the manufacturing facility, so that we
are able to improve our competitiveness in the market. However, this will take time, given the
existing complexities and challenges that are inherent in the transportation norms and regulations
today.
ITC has one of the most extensive distribution networks in India. Its products are available at 4.3
million of the estimated eight million retail stores in India. Of this, about 2 million are under ITC’s
direct distribution network. ITC will replicate the new factory-to-retail point distribution in phases,
and eventually bring all the 2 million retail points under the new structure.
Traditionally, ITC, like all packaged goods companies in India, stocks products at its distribution
centres. Products first go to its exclusive wholesale dealers, then distributors and direct retailers. In
some areas, mostly in rural India, the big retailers supply products to smaller retail outlets. The
entire process takes anything from one week to three weeks, depending on the distance between
the retail outlet and the factory.
ITC Case Discussion
ITC plans to remove as many stocking points as possible as part of its new model. At each stocking
point, products get degraded due to handling. The company will require less working capital if it
eliminates some stocking points. “This will ensure better profitability and we’ll get closer to the
consumers," Sumant said.
“Our ultimate goal is to ensure a supply chain that is as close to the market as possible with
minimum stocking points. This will ensure greater freshness of products, greater responsiveness
and lower costs.“
According to a McKinsey and Co. report on the future of retail supply chains, companies can reduce
costs by about 20% at the distribution centre level, while optimal deployment of inventory can
reduce working capital by about 10%.
“This is the right move, but it may not be easy for any FMCG company to replicate the factory-to-
retail points direct supply given the scale and nature of Indian retailing. Every FMCG company is
trying to come as close as possible to its consumers," said Abneesh Roy, associate director
(institutional equities research), Edelweiss Securities Ltd.
ITC Case Discussion
The new model will help the company get more data about consumers and, in the long run, that
and not cost could be the biggest benefit for ITC, Roy explained.
Indeed, the new model will help the company respond better to market demand, Sumant admitted.
The new distribution model is part of ITC’s plan to reach ₹ 1 trillion in revenue from its cigarettes
and packaged goods business by 2030. The company has also lined up investment of ₹ 25,000 crore
in 65 projects, including 25 factories for packaged goods, covering an area of 28 million square feet,
according to its annual report.
It added that the focus is on building flexibility and agility across the supply chain to ensure delivery
of volume and variety in a timely and cost-effective manner.
ITC is also in the process of automating all its 62 warehouses and its 1,550 wholesale dealers. “The
distribution network includes over 1,550 wholesale dealers and a force of 25,000-plus sales people
who are IT-enabled and empowered with hand-held devices, which ensure that every member of
the sales team has access to real-time actionable information on business imperatives and
performance. The supply chain mechanism is determined by analytics based on real-time feeds and
insights for over 1,000 stock keeping units (SKUs). This enables informed decision-making and helps
determine inputs, while ensuring an efficient distribution based on store-level demand," said
Sumant.
Supply Chain & Logistics Analytics – Session 8
Distribution Network
Chapter 4 and 5 from the Recommended Textbook
Term End Exam
2 DQ model 15 Marks each.
Caselet Based or Option to choose your own situation
Sub Questions with marks breakdown
Theory plus Application
Be guided by marks split
Be objective based on the marks split
Apply in sufficient level of depth
Answer specifically, separately, based on situation
Add examples where needed absolutely needed, but focus on
situation context
Focus Topics
AI, ML application in Supply Chain
Types of Analytics, Data Sources and Application of Analytics
to gain competitive advantage
Analytics in forecasting, Accuracy Measure, choice of models
Product Classification, EOQ, MOQ, (Forecastability, Volume)
Safety Stock, Service, Cost related analytics
Total Cost of Ownership
Analytics in Drivers, cost, responsiveness
Analytics in determining Trade off between cost,
responsiveness
Network Optimization Models – COG, Factor Rating, Cost
Volume
Session 8 Topics
With six national distribution centers (NDCs) carrying the entire portfolio of products,
the organization was carrying both excess and incorrect inventory. This forced the
fashion retailer to frequently initiate costly transfers between stores and NDCs to
correct stock-outs and satisfy customer demand. The already-stressed NDC network
would not have been able to cater to its growth plans. Further, the current network
was capable of supplying just 75 percent of demand within 48 hours – a metric they
sought to improve significantly.
In the fashion retail business, quick turnover of products is vital to keep shoppers
returning to retail locations, making efficient distribution paramount to continued
success. Supply chain network required was not only satisfy this demand, but would
also be scalable to serve a new B2B operation intended to distribute apparel to
independently-owned, smaller footprint stores.
A Fast, Scalable Distribution Strategy for Leading Fashion Retailer in India
Supply chain network design team analyzed this complex set of needs and designed an
optimized distribution network that includes a combination of consolidation centers
which acquire products directly from suppliers and store fulfillment centers.
The new design allows the organization to optimize transport times by giving store
fulfillment centers the option to either ship directly to stores or through any of its
cross-docks.
Vast network of cross-docks was a key design feature because it allows the company’s
fleet to fulfill products using milk runs.
Instead of waiting for different trucks to replenish specific SKUs, a single truck can be
loaded with several different items at different cross-docks and deliver a large variety
of SKUs directly to each location in a single trip.
Coupled with optimized shipment sizes and frequencies, this model both shortens the
time necessary to fulfill demand and reduces inventory.
The supply chain solution gives the fashion retailer the capability to reach 98.4
percent of its customers within 48 hours. Although the company was willing to
absorb increased costs to accomplish its growth goal, the new distribution
network would actually reduce its supply chain cost by 10%. This cost savings
is achieved through optimizing fleet mix and size, warehouse size and staffing
levels, and increased backhaul opportunities.
This new supply chain network model also gives the fashion retailer the
necessary scalability to keep its stores serviced appropriately across multiple
formats. The fulfillment speeds and supply chain cost reductions are expected
to give each location the ability to operate at their target margins while
growing the business through a quickly growing B2B network of retail
customers.
DHL Pharma Network Expansion
DHL Supply Chain, the Americas leader in contract logistics and part of
Deutsche Post DHL Group, is expanding its pharmaceutical and medical device
distribution network by 40 percent this year with an investment of $150
million. With the ultimate goal of bringing critical healthcare products closer
to trade partners and patients, DHL Supply Chain plans for nine new sites by
the end of 2019. The investment includes costs to invest in new buildings and
technology, as well as the fit-out and start-up of new or expanded operations.
These sites, strategically located within the United States, are fully licensed with
temperature controlled space that supports pharmaceutical storage requirements.
The sites also allow for packaging and managed transportation for integrated
solutions.
As the leading logistics provider for the life sciences and healthcare industry, DHL
provides forward-thinking, intelligent healthcare logistics through a holistic range of
patient-centric solutions. From clinical trials to point of care, and every step in
between, DHL's more than 9,000 specialists leverage the power of digitalization to
connect the entire value chain - pharmaceutical, medical device, clinical trial and
research organizations; wholesalers and distributors; hospitals and healthcare
providers.
Yasso Supply Chain Optimization
Founded by two childhood friends turned competitive athletes, Yasso believes
consumers should never have to sacrifice taste for nutrition or settle for the
high-calorie choices found in traditional premium brands.
As a result, case fill metrics were low and conversations with key accounts were
focused more on service issues than on promoting the brand and introducing new
products. Further complicating the situation, Yasso’s supply chain manager found
their existing planning process and tools failed to provide adequate visibility into
the current low inventory situations or a timeline as to when the company could
recover from it.
After a thorough assessment, fair few issues were identified. For starters, Yasso’s
on-hand inventory system accuracy was suspect, with confidence only in the
numbers at the beginning of the month after a cycle count. Second, the existing
tool’s projections of future inventory projections were based on two flawed
assumptions — that all orders would be filled and that product supply after the
first frozen month was not limited. Finally, the information coming from co-
packers about the status of production was largely incomplete, resulting in
occasional surprises when trucks would arrive at Yasso’s warehouse.
Yasso Supply Chain Optimization
To address these issues, Yasso had to improve the planning process through
better data and a more realistic view of supplier’s capabilities and limitations.
Yasso took a “boots on the ground” approach to visit each co-manufacturer to
better understand their resources, equipment, constraints, and concerns.
With the new planning process in place, Yasso has experienced significant
benefits. By utilizing accurate supplier data along with the Planscape
technology, Yasso has realized significant customer service improvements,
moving from worst to first in their current case fill percentage of 99.3%.
Lead Time Reduction
The retailer had experienced 3-years of rapid growth, which resulted in the
relocation of their distribution centers to accommodate higher volume levels.
As a result, the operations team relied upon an automated conveyor system
designed to expedite sorting for outbound distribution and product categories.
However, their functional requirements to move throughout the system were
not well defined. This resulted in a high rate of required manual sorting
impacting secondary labor costs.
To better understand the problem, the root cause of cost increases was looked
in. Upon assessing the current state of systems and processes, the team
realized a large part of the problem stemmed from a weak vendor compliance
program, with minimal emphasis on packaging and item categorization. To
alleviate the issue, the team worked closely with the retailer’s internal
stakeholders to identify and define categories that required classification as
“non-conveyable” for their distribution centers. These products required
additional handling due to size, fragility, or packaging material types.
Packaging and Handling Case Study
To streamline inbound operations and decrease deficiencies requiring secondary
handling, company created standards, identified non-compliant products and
packaging, and then trained frontline staff to handle products that fit this non-
conveyable category. This, in turn, allowed the organization to emphasize holding
suppliers accountable for following the new packaging standards. Specific departments
were given appendixes for their products that allowed them to identify strength
requirements, utilize testing guidelines, and reallocate conveyable vs. non-conveyable
inbound items to the appropriate receiving area at the distribution center.
Acquisition is one of the greatest challenges for supply chain. Best strategies
for integrating the acquired firm’s largest overlapping business unit into its
existing supply chain infrastructure, as well as a go-forward network strategy
for integrating the remaining businesses.
Team factored in not only key costs (inbound and outbound transportation,
manufacturing, warehouse fixed and variable costs, inventory carrying costs,
etc.) but key constraints like demand and supply, facility capacity or
expandability, and lead times to customers to address:
Network Design and Cost Savings
North American finishing optimization: Given long raw material shipping lead
times from China and India, the client needed to settle on the best mix of U.S.
and Asian locations to hold raw materials and produce finished goods,
including the best placement of the acquired firm’s products within its DC
locations–and in what volumes, and determine the optimal manufacturing,
distribution and 3PL distribution footprint.
The “bought-in” strategy was confirmed as the lowest cost sourcing approach,
with no cost-based financial benefit to increasing the total amount of product
postponed in North America beyond the client meeting service requirements.
Sub-assembly is to be sourced from India as the current lowest-cost Asia
source, while finishing in China or North America had higher costs.
Network Design and Cost Savings
Planning transformation: To ensure it could reconfigure its network when and
where it is needed in the future, company implemented more efficient
inventory planning and deployment processes, including integrated demand
and supply planning (IDSP).
Results
Company achieved its goals of reducing costs, improving service, and driving
growth and increased profitability for the newly acquired business through a
series of immediate, near, and longer-term network optimization initiatives.
Network Design and Cost Savings
As a result, and with its optimized supply chain network, the combined entity
is on track to realize a total savings opportunity of 6.1% of the total cost based
on projected sales volume growth targets over the coming three years,
including:
3.3% of total cost by prioritizing where to source and store finished products,
driven by increasing Asia production and allocating more distribution volume
to North American DCs with optimal port access
• Factor-rating method
• Locational cost–volume analysis
• Center-of-gravity method
• Load-Distance Model
• Facility Allocation
Factors Affecting Location Decisions
A three-tier model for assessing the competitiveness of a location
Transport Network Optimization
Truck utilization was enhanced from the existing 65% to 88%, an improvement of
35%, and reduced the unit cost per unit weight shipped, which is an important
metric.
One initial observation was that the truck volume utilization on most routes was low
and there were instances of routes where higher capacity trucks were used, but a
lower capacity truck on those routes would have been more efficient. They also had
a higher proportion of trucks at the main hub, while the number of loads originating
from the main hub was not exceptionally high. There was an opportunity to
reposition some of the trucks from the main hub to other hub locations.
Transport Network Optimization
Blue Yonder Transportation Modeler software used to build the model.
The different components modeled include the hubs and branch network and fleet
details such as truck types, capacity, business rules, and input costs.
3PL stipulated some business constraints while building the model. They did not
want the ratio of dedicated fleet to private fleet to exceed 15%. A vehicle allocated
to a hub should always return to the hub. The hub and branches on an existing route
would remain the same and shipments destined to a specific branch would be part
of an existing route. The routes serviced by for-hire trucks had to be serviced by for-
hire trucks and could not shift to fleet.
Three key unconstrained scenarios with different weight tonnages were developed –
current, growth projection for the next two years, and growth projection for the next
five years.
All scenarios focused on the optimal selection of trucks of different capacities on
routes and optimal allocation of trucks to hubs.
Transport Network Optimization
The product shipped tended to “cube out.” Hence the cubic volume was the critical
measure to optimize. The modeling recommended improved truck cubic utilization.
Truck utilization was enhanced from the existing 65% to 88%, an improvement of
35%. The model recommended the optimal equipment usage on routes while
ensuring allocation of larger-size trucks as the weight tonnage increased. This
specific optimization reduced the unit cost per unit weight shipped, which is an
important metric. The truck profile showed a decrease in the required number of 10-
ton trucks and an increase in 16-ton trucks with increased weight tonnage.
For future road map of their business, 3PL could chart the future roadmap that
included investing in higher capacity 16-ton trucks, reallocation of trucks across
hubs, and retiring trucks near end of life. The benefits realized included improving
truck utilization by selecting the right weight capacity truck on each route, thereby
lowering the unit cost per unit weight
Location Evaluation Methods
• Factor-rating method is location method that instills
objectivity into the process of identifying hard-to-evaluate
costs.
• Locational cost–volume analysis is a method for making an
economic comparison of location alternatives.
• Center-of-gravity method is a mathematical technique used
for finding the best location for a single distribution point that
services several stores or areas.
• Transportation model is a technique for solving a class of
linear programming problems.
Factor Rating Method
• There are many factors, both qualitative and quantitative, to consider in
choosing a location. Some of these factors are more important than
others, hence weights can make the decision process more objective.
• The factor-rating method is popular because a wide variety of factors,
from education to recreation to labor skills, can be objectively included.
1. Develop a list of relevant factors called key success factors
2. Assign a weight to each factor to reflect its relative importance in the
company’s objectives.
3. Develop a scale for each factor (for example, 1 to 10 or 1 to 100
points).
4. Have management score each location for each factor, using the
scale in Step 3.
5. Multiply the score by the weights for each factor and total the score
for each location.
6. Make a recommendation based on the maximum point score,
considering the results of other quantitative approaches as well.
Factor Rating Method Example
Locational Cost-Volume Analysis
• Locational cost–volume analysis is a technique for making an economic
comparison of location alternatives.
• By identifying fixed and variable costs and graphing them for each
location, we can determine which one provides the lowest cost. Locational
cost–volume analysis can be done mathematically or graphically.
• The graphic approach has the advantage of providing the range of volume
over which each location is preferable.
• The three steps to locational cost–volume analysis are as follows:
1. Determine the fixed and variable cost for each location.
2. Plot the costs for each location, with costs on the vertical axis of the
graph and annual volume on the horizontal axis.
3. Select the location that has the lowest total cost for the expected
production volume.
Locational Cost-Volume Analysis
Centre of Gravity Method
• The center-of-gravity method is a mathematical technique used for
finding the location of a distribution center that will minimize
distribution costs.
• The method takes into account the location of markets, the volume
of goods shipped to those markets, and shipping costs in finding the
best location for a distribution center.
• The first step in the center-of-gravity method is to place the
locations on a coordinate system.
• The center-of-gravity method assumes that cost is directly
proportional to both distance and volume shipped.
• The ideal location is that which minimizes the weighted distance
between sources and destinations, where the distance is weighted
by the number of containers shipped.
Centre of Gravity Method
Location Decision Case Study
Prior to the analysis, the organization’s complex retail and e-commerce supply
chain network had over 1,000 stores spanning 49 states and Puerto Rico, and it
needed a significant network redesign in order to effectively compete in the
evolving retail environment.
The company distributed products through several area logistics hubs (ALHs) and
11 store management facilities (SMFs). The question was, “Does this distribution
model fit the emerging retail distribution requirements?”
Some DCs had inadequate capabilities for picking, packing and shipping to meet
evolving store requirements and required large capital investments to reach
required threshold capabilities.
ALHs shipped to both the smaller store management facilities (SMFs) and to
some stores after cross-docking inbound products.
Distribution Network Case Study
Supply chain design analysis
The retailer took large-scale supply chain design. Analyzed over 200 scenarios
and sensitivity tests to model the best approaches for meeting store service
requirements, realizing savings, and determining the best consolidation
opportunities.
The supply chain design initiative proved so successful that the client asked
Chainalytics to provide managed analytics services to dynamically update its
supply chain models with data and to design new models for emerging product
lines which promote a “store within a store” presence at many of its retail
locations.
Distribution Network Case Study
Additional analytics were performed to determine the most effective
reconfiguration of a distribution of a specialty product line of balance lower
transportation costs against the increased inventory in the pipeline. Over 30
scenarios were evaluated to assist in identifying the most effective mode for
distribution.
Traditionally recognized by the iconic Morton Salt Girl, the average consumer may not
be aware that Morton Salt produces two broad sets of products relying upon
customized distribution channels. The first, referred to as “bulk” product, is primarily
used for bulk de-icing salts that keep roads safe in the winter as well as bulk production
for industrial customers. The bulk network consists of extraction mines, solar plants
and storage stockpiles located across North America. The second set consists of
“packaged” products (i.e., food, package ice melt, water softening salt and pool salt)
and moves through a network consisting of mines, plants, co-packers, and distribution
centers (DC). The average daily consumer of Morton Salt’s products typically purchases
from this package product category.
Supply Chain Optimization Case Study
In the years, Morton Salt leadership realized the organization’s supply chain
network existed as a patch-work of too many distribution centers, utilizing
approximately 40 DCs in 2011. By late 2013, the manufacturer had successfully
reduced this number to 26 DCs. Morton Salt’s leadership team felt there was
an opportunity for further rationalization, but that a project of this scope
would require greater analytics to support additional decisions involving
overall network design.
With the model demonstrating realized savings that were in line with the pre-
established expectations, Morton Salt executed the optimized baseline in
2014.
Supply Chain Optimization Case Study
Upon establishment of the new network, the company embraced a proactive
approach and worked to develop an “optimized future network.”
Advancing the design model, the team identified five warehouses to close or
exit due to geographic irrelevance while pinpointing three new warehouse
locations that would better serve Morton Salt’s client base.
The future state network also incorporated three plant closures into the
analysis.
Morton Salt implemented the initial changes to their supply chain network in
2015.
Through ongoing collaboration and continuous evaluation, the company was
well on its way of realizing the long-term benefits of consistent supply chain
network design.
Service Location Strategy
• Focus in industrial-sector location analysis is on minimizing cost
• Focus in the service sector is on maximizing revenue.
• Manufacturing firms find that costs tend to vary substantially among locations
• Service firms find that location often has more impact on revenue than cost.
• Therefore, the location focus for service firms should be on determining the
volume of customers and revenue.
• There are eight major determinants of volume and revenue for the service firm:
1. Purchasing power of the customer-drawing area
2. Service and image compatibility with demographics of the customer-drawing area
3. Competition in the area
4. Quality of the competition
5. Uniqueness of the firm’s and competitors’ locations
6. Physical qualities of facilities and neighboring businesses
7. Operating policies of the firm
8. Quality of management
Clustering
Clustering The location of competing companies near each other,
often because of a critical mass of information, talent, venture
capital, or natural resources.
Location Strategy Goods Vs Service