Summary: Can all AI strategies be defined by a few common needs or are the different AI strategy models sufficiently unique that they need to be considered as separate approaches.
There’s an article in the current Harvard Business Review titled “Competing in the Age of AI”. Since we’ve spent a good deal of energy thinking and writing about various AI strategies we were anxious to see what these leading business academics would add.
In this article the authors try to develop a series of fairly simplistic rules that apply to all companies seeking to implement an AI strategy. This struck us as a little wrong-footed since we’ve identified six different AI strategies that involve very different business models. We’ll come back to their rules later.
What caught our attention however were the specific businesses used as examples in the article. In addition to the well-known AI platform models of Google, Facebook, Alibaba, Tencent, and Wayfair, they proposed that their rules would apply equally to other models including Ant Financial, Zebra Medical Vision, Indigo Ag, and Ocado.
We know the platform players well. Remember that the core concept in a platform strategy is a two sided market (the platform) in which the business does not own either the supply or demand side resources, but binds them together and profits from the network effect of more and more users matching supply and demand through the platform.
The question we needed to research was do these other referenced companies represent a different set of strategies we might have missed. Here’s what we found.
Pure Play AI Products Companies – Vertical Strategy
Two of the example companies, Zebra Medical Vision and Indigo Ag are pure play AI product companies.
Zebra Medical Vision is focusing on medical applications of computer vision to provide automated interpretation of radiological images to speed up image diagnosis and presumably make it more accurate. Their product is an all-in-one platform for radiologists based entirely on the AI techniques of computer vision and image classification with DNNs. This fits our definition of a vertical strategy very well by providing a full process solution based on deep knowledge of a single industry with significant ownership of the underlying data on which the models are based.
Indigo Ag is similarly a pure play AI company using the deep learning techniques of computational biology to identify microbiome biologics that can be applied directly to seeds or soil to enhance output.
They combine this in a soil-to-sale platform that includes crop satellite imagery and a marketplace (platform) for matching growers and buyers (although the core of the company is not a platform strategy).
These business models started with a focus on a set of AI techniques and built unique business verticals not resembling any current traditional competitors. There are still vertical market opportunities available but they are not common. Our observation is that most of these companies become M&A targets for established industry players.
AI Disruptors
If we missed anything in our previous discussions of AI strategy it may be this category of AI Disruptors. While essentially all large well established companies are constrained to use an “Applied AI” strategy in which AI is grafted onto the existing business model to optimize various processes, these AI Disruptors set out to recreate the capabilities of an existing business without reference to old-model processes.
Ant Financial spun out from Alibaba just five years ago Ant already services one billion customers and employees one-tenth the number of workers compared to traditional financial institutions. Per the authors:
“Unlike traditional banks, investment institutions, and insurance companies, Ant Financial is built on a digital core. There are no workers in its “critical path” of operating activities. AI runs the show. There is no manager approving loans, no employee providing financial advice, no representative authorizing consumer medical expenses. And without the operating constraints that limit traditional firms, Ant Financial can compete in unprecedented ways and achieve unbridled growth and impact across a variety of industries.”
Ocado is an on-line supermarket founded in 2000 with the goal of transforming the supermarket industry. Although they originally had a brick-and-mortar arm that’s subsequently been spun off, their focus has been on AI-enabled robotic pick-and-pack systems which is their core competence. They claim to be the world’s largest online-only grocery platform.
While they are disrupting the traditional supermarket business model they are also successfully marketing their warehouse fulfillment equipment to traditional retailers looking to add AI enhanced automation.
Do General Rules Apply to All AI Strategies
The HBR authors assert that what all these models including the giant platform players have in common are these four requirements.
“Four components are essential to every [AI] factory.
The first is the data pipeline, the semiautomated process that gathers, cleans, integrates, and safeguards data in a systematic, sustainable, and scalable way.
The second is algorithms, which generate predictions about future states or actions of the business.
The third is an experimentation platform, on which hypotheses regarding new algorithms are tested to ensure that their suggestions are having the intended effect.
The fourth is infrastructure, the systems that embed this process in software and connect it to internal and external users.”
At its most basic level this is certainly true of ‘Applied AI’ strategies for existing businesses. But it strikes us that the design, methods of implementation, and even the management of these required elements varies greatly by strategy type. Mixing the requirements of existing businesses seeking to graft on AI with those of platform, vertical, or AI pure play companies is not very explanatory.
For example, when Walmart wants to add AI enabled automation to its warehouse it considerations are quite different from that of Ocado. Walmart has the advantage of a huge trove of consumer purchase history. Ocado no doubt initially suffered from the cold start data problem.
When Zebra or Indigo were designing their platforms they were much more concerned with the effort of creating algorithms than other verticals where the much simpler application of ML models likely sufficed.
When Ant turned its AI loose on loan approvals it desire for the process to be largely autonomous led to much different monitoring and control processes than the average financial institution which may use ML models to make recommendations reviewed by managers.
So while these four elements of the so-called AI Factory are probably common, the considerations in design and implementation are decidedly not common.
Other articles on AI Strategies
Six AI Strategies – But Only One Winner
It’s Official – Our DNN Models are Now Commodity Software
AI/ML Lessons for Creating a Platform Strategy – Part 2
AI/ML Lessons for Creating a Platform Strategy – Part 1
A Radical AI Strategy – Platformication
Now that We’ve Got AI What do We do with It?
Capturing the Value of ML/AI – the Challenge of Offensive versus Defensive Data Strategies
The Case for Just Getting Your Feet Wet with AI
The Fourth Way to Practice Data Science – Purpose Built Analytic Modules
From Strategy to Implementation – Planning an AI-First Company
Comparing the Four Major AI Strategies
Comparing AI Strategies – Systems of Intelligence
Comparing AI Strategies – Vertical versus Horizontal.
What Makes a Successful AI Company – Data Dominance
AI Strategies – Incremental and Fundamental Improvements
Other articles by Bill Vorhies
About the author: Bill is Contributing Editor for Data Science Central. Bill is also President & Chief Data Scientist at Data-Magnum and has practiced as a data scientist since 2001. His articles have been read more than 2.1 million times.
He can be reached at:
