Digital + AI: The Key to Unlocking the Value of Transformations

by Michael Watson

Opex Analytics
The Opex Analytics Blog
10 min readSep 1, 2020

--

Introduction

Digital transformations are not new. Many companies have been at this for several years. However, it seems like the momentum is really picking up for a few reasons. First, the Covid-19 pandemic has changed quite a few business models and ways of working. Second, AI and digital capabilities have continued to rise. Finally, there is a growing body of knowledge around how to succeed with a transformation.

For example, in July of 2020, Gartner listed their top 8 technology trends for supply chain leaders looking to transform their business. This list included technologies for hyperautomation, continuous learning, AI, and a digital supply chain.

As another example, in 2019 TechRepublic published an article on lessons learned from successful digital transformations. The lessons, based on research from BCG, still apply today. These include getting support for a larger leadership team, not just the CEO, building the right skills, and building an open ecosystem. The most important lesson was the last however, don’t wait for all that — go fast and get applications built to start realizing the benefits as soon as you can.

Since so much has changed recently, especially in the area of AI, this paper is going to take a step back and talk about a key idea in the area of Digital, clearly define AI, and discuss the latest thinking in combining Digital and AI. We think that understanding these concepts deeper will allow companies and executives to craft a strategy and approach that works well for their situation.

Power of Digital

Back in 2011, Marc Andreesen wrote a seminal article for the Wall Street Journal entitled “Why Software is Eating the World.” His hypothesis was that software was taking over or upending almost every industry. He gave obvious examples from music and photography going digital to Google (a software company) becoming the world’s largest direct marketing platform.

He also mentioned the use of software to power the logistics behind Wal-Mart’s dominance in brick and mortar retail (he’ll have to update this with the changing retail landscape of course), the Military’s use of software to enable the warfighter and drones, and even Amazon first selling books on-line and then cannibalizing that business by pushing digital books.

If Andreesen were to update the article today, he will most likely continue the story with the huge surge in e-commerce (and sure to stay high as COVID-19 passes), how people now exchange money with software, the emergence of digital-first companies (i) (like Stitch Fix and Farmstead), and the use of software to re imagine the trucking market, to name just a few examples.

Looking back, he was right.

And, this trend seems likely to continue. The use of software to transform industry has been going strong for decades and shows no signs of slowing down.

As additional fuel, something else has happened since 2011 that will only accelerate and make the digital trend more powerful: the rise of Artificial Intelligence (AI).

The Power of AI

To understand the power of AI, we should define what it means for business.

The confusion comes from the fact that the word is used in two different ways.

In one context, AI means Artificial General Intelligence (or AGI). AGI is a field of study where researchers are trying to get computers to have the same kind of general intelligence as people. That is, the AGI researchers are trying to create algorithms that have common sense or can learn from one area and apply it to something new. AGI is what people talk about when they mention super-intelligence, the singularity, or machines taking over the world. We think we are far away from this. In this paper or as it relates to most business uses, we are not talking about this definition of AI.

The other definition of AI can be thought of as Practical AI (or, as it is commonly used, just AI). Practical AI is all about solving specific problems with data, algorithms, and computing power. The algorithms at the core of this are some of the same ones used in the field of AGI and a collection of others that, taken together, look like they bring human-like (or often better and faster) intelligence to the problem. Collectively these algorithms help find patterns, read documents, recognize images and speech, make predictions, and help find the best solutions for given objectives (ii). And these algorithms are not standing still. Since 2011, the research community has produced a steady stream of new algorithms, new approaches, and new open source options. This means that we can solve problems faster and more accurately, as well as solve problems that weren’t solvable a few years ago.

There is a strong desire to continue to use practical AI because there is business value.

We’ve seen examples of companies extend forecasting to predict next week’s order from a large customer using additional external data like pageviews, better predicting the transportation price on a specific lane, reading PO and invoice data to spot erroneous charges, predicting stock-outs, or improving employee safety with AI.

Each of these applications and many more would be classified under the umbrella of AI. But, having just a handful of these applications isn’t really embracing the AI movement. To truly have an impact from AI, you need to combine AI with Digital.

Combining Digital and AI for Better Decision Making

Think of Digital as the muscle — it stores data, moves it around, cleans and organizes it, gives it context, makes it available to users, and makes transactions happen. Then, think of AI as the brains — it takes the data and finds the patterns, makes the predictions, or sorts through it to find the best answer given the goal or goals the user is trying to optimize.

Either can work by itself: You can use Digital to create a product that wasn’t there before (like digital transfer of money), offer your inventory across different channels (in store, on-line), or automate routine manual processes. You can use AI to gain insights about your customers, develop a better inventory strategy, or determine risky parts of your supply chain.

But, when you combine the power of AI within a Digital strategy, you can start to dramatically improve decision making. You can make more decisions, make them better and faster, and automate decisions — that is, you let the digital infrastructure get the problem to the AI algorithms, you let the AI engine do the thinking, and then the digital infrastructure pushes the decisions out for execution.

Competing in the Age of AI by Iansiti and Lakhani coupled with our own experience shows the value of combining Digital with AI. In fact, in their book, Iansiti and Lakhani make it clear that businesses can’t have a true digital transformation without AI. Likewise, you can’t get the full value of AI without a digital strategy. In their framework, they have three basic building blocks (they have some additional points, but we’ll simplify to three here):

  1. A Data layer with data pipes and openness of data so that all groups can access it.
  2. An Algorithm layer to build the brains of the applications — to make a prediction, find a pattern, or make a recommendation.
  3. A Software layer to deploy the application to business users or to automate decisions within existing systems.

As you move toward these building blocks, you will start to realize the benefits of a digital transformation and becoming an organization that embraces AI.

If you apply these ideas to the core of your business or supply chain, you can start to remove human decision making as a bottleneck. This, in turn, can allow you to massively scale the business in a way that outpaces traditional operating models.

For example, the following graph, from an HBR article on the book, highlights the value of combining Digital + AI to automate processes. Once the process is automated, you can easily add the next user (customer, vendor) without increasing cost. This contrasts with traditional operating models where you need to add more costs and more people for new users.

Another benefit of these building blocks is that they allow you to create not just a few apps, but to build 100s or 1000s — to create an app factory. This part of the business scales too and allows you to solve new problems, improve existing decision making, and to quickly upgrade as new ideas surface. And, to prevent chaos in building the app factory, these apps pull from a common data source, use standard components and interfaces, and have processes in place to check and maintain them.

By creating this foundation, you are also increasing your agility — you can respond quickly to changes. This is especially important as the economy pushes through and to the other side of the COVID-19 pandemic.

Examples of Companies Embracing Digital + AI

Ant Financial (a spin-off of China’s Alibaba) is a compelling example for how Digital + AI can break bottlenecks and allow the business to scale (iii). They were built with a Digital + AI mindset from the start. Their market involves serving many smaller customers (unlike banks in the US and Europe). One example is the loan processing step which they call 3–1–0: three minutes to fill out an application, 1 second to approve or reject it, and 0 human intervention. This thinking is applied across their whole portfolio. The technology investment is large, but adding an additional customer has almost zero marginal cost: they broke a human bottleneck and created a business ready to scale. They serve 700 million customers with 10,000 employees. In contrast, Bank of America serves 67 million customers with 209,000 employees (iv).

Barilla Pasta is pushing to digitize their food supply chain to give consumers more information, but also to monitor food safety and reduce fraud and waste. In addition, Barilla’s VP of Supply Chain Design, Planning and Customer Service has spoken at several events on the importance of creating a data foundation (a digital twin of the supply chain) to allow many more decisions to be made with thorough analysis at a faster pace. A related example, Belcorp (a $1B+ Latin American beauty and cosmetics company), describes that by having this foundation in place, they can build many additional apps to make better sourcing, transportation, omni-channel, and planning decisions — in the terms from above, Belcorp is creating an app factory to make better decisions in a seamless manner.

Leading e-commerce sites are using AI to help uncover the root cause of order failures. These order failures could include missing the delivery promise date, sending the wrong items, needing to use expedited shipping, running out of inventory, or even inventory shrinkage. The digital aspect involves pulling together the life of an order — this is not trivial as an order will pass through many systems and processes from the point of order until final delivery. Then, the AI algorithms help make sense of the data and automate the root cause labeling and analysis of the data. This helps find and break bottlenecks in the system. And, as mentioned above, once the data foundation is in place, many other apps can be added to continue to make things better. This can include decisions on when to expedite orders, how to prioritize limited capacity, and how to help with short-term labor planning.

The above examples are just a small slice of what is possible. And, the possibilities continue to grow as companies find more ways to leverage new sources of data and new algorithms to improve decision making. Also, you can see how these business applications differ from Artificial General Intelligence — these algorithms solve a practical business problem with what looks like human-like decision making. But, they only solve the problem that they are suited to solve — no one thinks that these algorithms will start thinking on their own.

Committing to the Digital + AI mindset is a big investment with potentially large returns. But it is not an all-or-nothing payout: you can reap returns with smaller projects as you build out your Digital + AI foundation.

Conclusion

For companies going through a transformation, there is power in combining digital and AI. These areas have been evolving rapidly, but many of the key pieces have matured to the point where companies should no longer wait to get started. The Digital gives you the muscle and the AI gives you the brain. One without the other won’t likely get you the transformation change that you seek.

End Notes

i. These are firms that started out as digital — that is they were designed from the ground up to be digital, to use algorithms at the core of the business. They were built to digitally scale quickly.

ii. For a more detailed discussion on this see our blog post, Why Business Leaders Should Think of AI as an Umbrella Term. The blog also talks about how AI in the business community has replaced the term “Analytics” to refer to these algorithms. And, we think this is a good thing. Analytics lets you get away with just doing reports. But, AI forces you to think about how algorithms can help you make predictions, read documents, or decide what to to — a step up from reporting.

iii. The Ant Financial example is from the book Competing in the Age of AI.

iv. From Competing in the Age of AI, page 26

--

--