3 Methodologies to fix 48% of your supply chain problems, and slash your Carbon footprint by 28% - Author: A. Yassin Ibrahim ElGabroni, MBA

Here's the truth!

Supply chains are over increasingly complex and inflexible. They're hyper-sensitive to disruptions e.g.,(pandemics, wars, and natural disasters), and they’re built on fragmented systems that can break down when faced with shortages or shifts in demand.

Context:

In order to streamline tactical and strategic operations along the value chain, more and more organizations are adopting a digital transformation strategy aligned top-down with their business strategy to:

a) reduce costs (COGS, OPEX and CAPEX).

b) Uplift top & bottom lines while ensuring a healthy balance sheet.

c) Ensure extended customer satisfaction and a high service level rate.

In their endeavor to achieve that, enterprises have been and continue to implement tools, mainly ERP systems, that can improve - among other business areas - the efficiency and accuracy of their demand forecast planning, shifting from spreadsheets to digital platforms, or a combination of both, depending on the designed operations model and methodology used.

Argument & Issue statements

Those digital tools e.g., (SAP APO) have brought quantifiable value into business and operations such as:

a) 3-7% increase in enterprise global operating profits.

b) 20-30% improvement in TTM (Time to Market) for new initiatives.

c) 28-35% improvement in forecast accuracy.

(Source: Report by Bain & Company)

However, those tools are now becoming more traditional in the grand scheme of things, where demand volatility and uncertainty are the new normal. Organizations need to continue to disrupt their digital architecture and adapt more advanced tools/methodologies or they shall get disrupted by external factors - Take COVID19 and the war in Ukraine as an example, the disruptions have put everything to the test; from resiliency to efficiency to agility.

Here are the top 4 pain points that organizations still using older generations of digital tools to manage their demand and S&OP planning are currently suffering the most:

1- Inability to cope with the shift from traditional sales to E-commerce:

More than ever, people are buying more and more online. Q1 2021 saw a?39% increase in e-commerce sales accelerated by the pandemic and it's here to stay, with an expected growth in the US valued at $6 trillion by 2024. This is a shift that the majority of enterprises were/are not ready for, with increased demand for products online, there will continue to be increased demand for fulfillment and shipping services.?(Source: own research)

2- Sudden stock availability shortages:

This is the biggest issue facing approx. 89% of all consumer goods companies, with 46-52% of the cases accounting for demand planning inefficiency and forecast accuracy. This translates to significant amounts of financial losses, for example, Walgreens - the second largest pharmacy store chain in the US - having a $1 billion in lost sales opportunity behind forecasting error, leading to the exit of 2 executives in 2014. (Source: Forbes magazine 28/08/2014 edition).

3- Limited visibility on the E2E process model:

With Data and information flowing from multiple sources, it becomes subtle to keep track and communicate effectively down the line; companies rely on a range of partners and platforms to help them fulfill and ship orders. They might work with different trading partners, manufacturing partners, retail partners, and freight and parcel carriers. On top of that, they may have an enterprise resource planning (ERP) system along with multiple sales channels, leading to friction and low-quality data entries to generate their statistical baseline forecast on a MatLoc level ( Material SKU per Location - DC/Plant).

4- Inefficiencies lead to more carbon emissions:

Supply chains have the greatest impact on sustainability goals and carbon emissions. The typical consumer goods company’s supply chain e.g., (P&G, Mondelez, Coca-Cola, etc.) creates much greater social and environmental costs than its own operations. Supply chain impacts account for more than 80% of greenhouse gas emissions and more than 90% of the impact on air, land, water, biodiversity, and geological resources. (Source: Own research + McKinney & Co. ).

Value Proposition & Solution:

We've developed a framework and identified, quantified, and qualified 3 opportunities / Methodologies to tackle the hereby mentioned inefficiencies above.

first of all, let's agree that by now we understand that older generations of digital tools & methodologies utilized in supply chain planning - which are still adopted by more than 85% of all world organizations are:

1- Unable to keep pace with the increasing complexities in supply network constant disruptions and the complex data models feeding from various source systems.

2- Lacking the "Rapid Demand Forecasting" capabilities to produce quality demand and S&OP plans, and react on "near-real-time data" to ensure decisions are made based on the most refreshed data and information available, without the need for manual touches and down-times.

3- Increase the susceptibility for supply chain inefficiencies, which by reflection augment the negative impact on environmental sustainability and enlarge the carbon emission footprint.

Below is our framework outlier developed to help tackling the aforementioned issue statements. In this article, we will focus only on " The Methodologies" lane of the framework.

Please Note: Implementation of the 3 methods at the same time is not necessary, however, each method compliments one another. The adoption of the 3 methods and the consolidation of outputs will result in formation of a formidable business and operation model.

Methodology (1) - Process mining:

Overview: Process mining serves as an X-ray capability for business process on a granular level, showing exactly what happens at every stage of every process, which enables an organization to get a zoomed-out, completely objective view of what’s happening and why it’s happening, which enables:

1) Tracking business Financial & operational KPIs correlated with the process.

2) Understanding what activities in the process are driving the loss root causes behind performance and causing Revenue leakage.

3) By application of pareto 80/20 principle, we're able to target solutions for the 10-20% of elements, causing 80-90% of the problems.

" There’s a huge scope for you to use Process mining in Supply chain Planning and Forecasting too. Process mining helps you reveal and fix the hidden inefficiencies in your processes, so you can perform at levels you never thought possible.", said Andrew London - Senior Manager at Celonis.

Expected Value: Through industry benchmark, when adopting this methodology, you can expect to improve your business by:

a) Up to 58% increase in process efficiency and conformance rate to comply with the ideal process flow in what's known as " The Happy Path", which increases resilience, speed and eliminates friction.

b) Reduce Business Waste from supply chain by up to 42%, which contributes to increased sustainability and reduced carbon footprint.

c) As a bonus, Increase Total Revenue by 0.8-1% per process improved.

Proposed Technologies to consider: Celonis - UiPath - IBM PM - Pega (The proposal is based on Analyses done by Gartner + Customers / End users reviews + Own review)

Methodology (2) - Consensus, Concurrent & Touchless Demand and S&OP planning:

Overview: The Demand Consensus Planning final baseline forecast, in this method, is generated based on four main process components / building blocks and workbooks ( Statistical Forecast - S&OP Sales forecast- S&OP Marketing Forecast - S&OP Finance operating plans). Weights are distributed based on the analyses generated by the involved Machine Learning algorithms and the historical forecast accuracies per building block.

The touchless component of demand forecasting is more technically complex; where a series of end-to-end automation processes fuel real-time algorithmic planning in the following order:

1- Cognitive data layer:?Automated data streams are fed to the used technology in real time

2- Pre-processing:?The data is cleansed and qualitative data is encoded.

3- Forecasting:?Using a portfolio of algorithms and methods, the system uses statistics and machine learning to make predictions.

4- Ranking algorithm:?The best metric is determined and top ranks are returned to the end user for advanced parameter tuning.

5- Post-processing:?The action approved by the end user is fed back into the system to fine-tune automation rules and improve forecast accuracy

(Touchless Demand E2E process Source: Report by the "Cognitive Automation Community")

The last component is the concurrent planning, which allows information and data to flow simultaneously in many directions. This method allows for infusing the entire supply chain network with up-to-date data to increase rapid and finely-tuned preemptive decision making where any changed part / parameter of the supply chain planning area influences the rest. For example, arising risk from a vendor that's not going to be able to supply Raw material part required for a given product to be manufactured on-time, results in streamlined effect in all planning areas ( Production, Demand, Replenishment, Inventory, etc.) for planners to see and react upon.

Expected Value: Organizations adopting this methodology can expect to improve business metrics as follows:

a) 15-40% Reduced Inventory Levels.

b) 4-6% Improved Service Level.

c) 15-40% Reduced Short-term forecast error

d) 5-20% Reduced mid-term forecast error.

e) Reduce emissions by 32-35% over a period of 10 years ( approx. 3.3% YoY)

Proposed Technologies to consider: Kinaxis - SAP IBP - Blue Yonder - Custom made solutions by Digital consulting firms ( IBM, Accenture, TATA, Infosys, Cognizant, Capgemini) - (The proposal is based on Analyses done by Gartner + Customers / End users reviews + Own review).

Methodology (3) - Digital Twins & Neural Network:

Overview: Artificial Neural Network (ANN) algorithms have been found to be useful techniques for demand forecasting due to their ability to accommodate non-linear data, to capture subtle functional relationships among empirical data, even where the underlying relationships are unknown or hard to describe. The model uses outputs from ML algorithms like Naive and ARIMA to train a separate model; which improves the ability to predict seasonality in demand volumes and trends on a granular sub-aggregate level.

While the Digital Twins enables building an identical digital model of an organizations supply chain network, processes involved per business area, and allows an organization to simulate an event scenario e.g., the next pandemic or world conflict say in Taiwan economic area and the implications on supply chains. Manufacturers in the supply chain can use digital twins as the basis for alignment and local / global decision-making in the supply chain. In addition, by using data and predictability outcomes in digital twins and cutting the requirement for physical prototyping, teams and organizations will find more flexible testing scenarios, which will help save time and costs. Digital twins modeling and simulations can be additionally combined with IoT data to drive formidable insights and extended accuracy.

"Digital twins can play a big role in helping companies model the potential impacts of disruptions to find and fix vulnerabilities that could harm the business. They also can be a powerful tool for optimizing supply chain networks and processes, as well as inventory." , said Jaime Rodriguez - Managing Director at Accenture.

Expected Value: Organizations adopting this methodology can expect to improve business metrics as follows:

a) Decrease overall GHG emissions produced by a manufacturing company by 5-7% over a period of 10 years ( approx. 0.6% YoY)

b) Optimize Supply Network & Planning efficiency by 50-60%

c) Improve Demand tendency to bias by 15-30% depending on the complexity and the trainability of the Neural Network.

d) Improve Service Level by up to 6%.

e) Increase Stock availability and reduce stock-outs by up to 12%

Proposed Technologies to consider: Unable to give much proposals in this area, however, utilizing Celonis together in SYNC & Integration with BusinessOptix shows promise and value has been proven through Big Data Samples to demonstrate POV/POC (Own review).

Mentioned in This Article: 埃森哲 , 麦肯锡 , 波士顿谘询公司 , 贝恩公司 , 高知特 Cognizant , 凯捷咨询 , 印孚瑟斯 , IBM , 宝洁 , 可口可乐公司 , 亿滋食品 , Tata Group , Kinaxis , Celonis , SAP , BusinessOptix , BlueYonder , UiPath , Pegasystems

Related Hashtags: #digitalbusinesstransformation , #digitaltwins , #processmining , #demandplanning ,#supplychain , #digitalstrategy , #erp , #artificialintelligence #sustainability , #carbonemissions