Turning Data into Insights
Read how this Chief Data & Analytics Officer learned to turn data into actionable insights.
Published: March 10, 2023
Put theory into practice
Das Dasgupta is a Chief Data & Analytics Officer working out of Los Angeles, California. His career has seen him move from various roles and industries. His foundations were laid in supply chain management as an early member of i2 Technologies. At EY and McKinsey, he worked in data planning. Following those roles, he then became the head of North America’s Customer Experience at Amazon. Das was recognized as one of the top 100 innovators in Data Science and Analytics, where in 2019 he was rewarded with the Viacom Automation Award for visionary leadership. Das is currently leading advanced analytics, data operations, and digital advertising portfolios for Saatchi & Saatchi.
Das came from an academic family and received a BA, MA and PhD. After academia, he joined i2 Technologies and he built on this experience for a career in the technology industry. The work taught him to understand supply chain challenges and find business-oriented solutions. At this point of his development, Das realized that he had never been on-the-ground observing the supply chain at its base; his relationship with it was solely theoretical.
This changed when Das joined Amazon and asked to run North American Customer Experiences (ACE). The ACEs team comprises three groups: Think, Teach and Deliver. The Think team consists of data scientists and advanced analytics technologists, the Teach team consists of Lean Six Sigma Master Black Belts and the Deliver team is a dotted relationship with about 120 general managers of Amazon’s fulfillment centers (FCs). As Das says:
“On the first day I showed up there my boss said, now it’s time to chuck your ego out of the door and go to the floor.”
From the theoretical to the practical: Das put on gloves, warehouse overalls and was shown the ropes of picking, packing, and shipping. For three months, warehouse workers showed him how the supply chain works at the base level. Impressed by the rapid hand-eye coordination of the workers packing, Das was told by a colleague:
“You scan the item. Scan the tote. Scan the shelf. Scan your trolley. These four scans, if you don’t do that, data is not going to be made. We are making a digital twin here.”
Amazon, being a digital native, realizes the importance of gathering data and converting it into knowledge in order to continue a seamless operation and please a vast consumer population, all day, every day. No data is wasted.
From the granular to the global
After 18 years of supply chain management and planning, helping companies transform processes and rollout S&OP processes worldwide, Das realized he knew little about supply chain management. At this point, after observing fulfillment in Amazon’s warehouses – from the bottom up, the practical, from product to customer – Das believes his learning began.
Amazon has its own homegrown solutions, the IT systems are not borrowed from anyone, the structure is unique to them. For example, instead of SKU they use an individual Amazon Stock Identification Number (ASIN). In the sprawling universe of Amazon, this technology allows them to tell where a product is at any time – from the warehouse floor to the customer’s door. As Das points out:
“At Amazon no data is wasted. It’s the standard all companies must strive towards. Data has to be created, managed and converted into real knowledge so the company can learn and run the supply chain to its full potential.”
Use the data as the dots to map your digital transformation. Digital native companies have a head start but with new and revolutionary software like that of o9’s Enterprise Knowledge Model, traditional companies can now attach themselves to a digital twin and modernize their business and supply chain.
The simple advice is to always be humble, absorb information and learn as much about your organization – from conference table to warehouse line. Make sure the data comes from the place that created the data. Avoid an approximation and cut all guesswork.
Once you have gathered data, carry the objectivity of it into your company’s ethos. A company needs to be all on the same page with data-driven insights forming the strong spine of the organization, this leads to less disagreements, less guesswork and more transparency. Classically, this has been a problem for companies that are not digitally native. The aim is for a company to create one version of the truth and this can be done through data.
Data is key to communication across your whole supply chain. Software, like o9’s Enterprise Knowledge Graph, can give non-digital native companies the means to hone in on a specific FC and scrutinize the process, find a problem and solve it.
Data is data and it carries an objective truth. Companies willing to transform need to have all their team on the same standpoint: trust the data, from the bottom to the top, the granular to the global. Whatever the question – data is the answer!
Pinpoint the problem, find the solution
Back in 2015, Amazon was seeing a very large number of returns for its heavy and bulky equipment, particularly large-screen televisions. There were about 5.3% of those returns marked as defective or damaged. Sometimes an irate customer would figure out Jeff Bezos’s email address and send a complaint detailing a bad customer experience. Jeff would then send this email down to his direct reports who flip it to their direct reports.
Das gives some advice: “The easiest way to get fired by Amazon is to say – oh, I’m taking care of this, no worries.” In order to get the solution, action is required. Efficient problem-solving is the spine of Amazon’s business model. The problem-solving protocol means the next few days or weeks or however long it takes are spent uncovering the root causes with data. Write-ups, reports are written and need the numbers and data to substantiate any claim. Full comprehensive reports were required, as Das says:
“There were no PowerPoints allowed because PowerPoint is considered to be “fluff.”
At Amazon, this “fluff” is avoided. Therefore, Das went to warehouses and FCs that dealt with bulky products, took pictures, asked questions and did a lot of diligence. They found numerous root causes:
- Every time television was traveling more than 500 miles, the probability of it getting damaged was exponentially increasing.
- Many times inbound operations did not notice that a television was being delivered and it was defective.
- If a television falls from a height of 12 or more feet then it has a high probability of being defective. They did conventions with Samsung, Sony and Vizio etc. They dropped televisions from a height. Then, they took 3D images of it in order to see how it breaks and how far it needs to fall to break.
- The television should be standing erect and not slanted.
- We found out how the forklift operators should be actually making sure that they’re holding the televisions by the sides and not on their faces.
This sort of granular analysis is an example of why Amazon is at the top of the supply chain game. After finding out these causes, they then collected data and rolled out measures in every FC. Every operator shipping televisions took a picture of the truck departing the dock. Das and his team could then see exactly how the shipping was carried out. In three months, the results reported defective products to drop down from 5.3% to 1.8%. This roughly accounted for $80 million of savings. As Das says:
“That is where the rubber hits the road and data science works with the supply chain team, data, digital twins all come together in reality. This is an example of how Amazon turns data into decisions.”
Amazon started out as a digital platform and continually converts anything physical into digital. A new Amazon hire – an Amazonian – is not only taken through that process of understanding how it works in the fulfillment world, but also every single person is given some training in digital literacy. Operators in lines, area managers and senior managers are on the floor wearing gloves and overalls are often from Harvard or MIT with an MBA, Master’s or PhD in computer science. They are just going through the process and learning. They run the line and then eventually they may take a role back into Seattle at the HQ.
Problem-solving is a fundamental DNA at Amazon. A problem solver at Amazon is someone who could understand data, see what it says on the screen, be able to map it in a picture, into what’s going on on the floor, then go look for it. This is taking us from math to the atomic world and vice versa. The atomic world of every single digital twin traveling out there, back into mathematical processes where we are hitting the books again and doing our calculations and coming up with additional ideas that we should try and then make it really happen. So without data literacy no one in Amazon is going to be able to survive because that is you know, key to your understanding of how end-to-end processes and supply chain works.