Looking at the performance of IBM shares over the past five years, it is clear that a change in strategy is needed. IBM’s share price is down approximately 9% since 2011 compared to a 54% gain in the S&P 500. The goal of this article is to develop a strategy for IBM to leverage the power of IBM Watson artificial intelligence to stage a comeback.
The proliferation of cloud, social and mobile technologies have led to the most successful and innovative companies becoming increasingly concerned with the ability to successfully build a digital platform. Apple, Google, Facebook and Amazon each created platforms that co-create value by connecting to other business who can build products and services on their platforms. For a company to succeed in building a platform, they ought to focus only on their comparative advantage and try to capture as much value possible.
In the field of artificial intelligence, the next big frontier is in applying cognitive intelligence to big data. APIs in particular are so intriguing because of the possibility of making machine learning easy to use and highly accessible. This can be achieved by removing the most arduous aspects of designing and utilizing machine learning algorithms. This, in turn, opens up the door for developers to focus on managing their own data, enhancing the user experience and interface, and gaining predictive insights from their data.
The key benefit derived for the creator’s of API’s is getting a deeper understanding into their customers wants and needs. The creators can then leverage that data to build better products and services they can sell to the users of their APIs. The approach can be outlined as follows:
- Commoditize differentiated services by making APIs open and accessible
- Use platform to gain insights into customer wants and needs
- Create superior products and services that are customized for customers
- Sell those products and services to customers
Case Study: Amazon Web Services as a Cloud Services Platform
Initially, Amazon built AWS for their own internal use and it was only a footnote that the company could eventually sell services on the AWS on the platform and sell it as a commodity. In 2006, the company decided to launch AWS and has had incredible success in building the cloud services platform. Multi-billion dollar companies, including Netflix and Airbnb, came to the conclusion that it is more economical to pay Amazon than to build their own data centers.
The vast majority of existing companies will increasing rely on public clouds and Amazon remains well ahead of the competition because of the simplicity and economies of scale of AWS. For startups, the case is even more clear because it is significantly less expensive to utilize Amazon’s cloud platform than to build out an infrastructure. This used to be a significant upfront, fixed-cost faced by all startups. Companies who may have previously been hamstrung without access to data servers or dependent on venture capital funding can now have access to one of the world’s most powerful cloud computing platforms for relatively low cost. The startups also benefit from switching the fixed costs to variable costs as companies only need to pay for the AWS resources as they use them.
The Business of Artificial Intelligence
The business opportunity of AI remains enormous. Not only will machines process data faster, but they will also be able to expand the scope of processable data. Traditionally, companies depend on mathematicians and engineers to build data models but even the most advanced quants are only capable building, testing and analyzing one or two models per week. Today, through machine learning tools, companies can create thousands of models per week using the massive processing power and memory available.
The machine learning techniques of today typically uses brute force solutions such as a series of mini-max decision rules. True AI, however, remains an incredibly hard problem because cognitive intelligence goes far beyond simple problem solving methods and encompasses far more variability. After numerous false starts, there once again seems to be a sense that computer science is on the verge of cracking artificial intelligence. Multiple techniques have been developed including deep learning and smart algorithms that are constantly improving to get to the point where a computer can teach itself how to think.
IBM Watson: Artificial Intelligence as Digital Platform
“My own view is that every single professional on the planet can be as good as the best professional in their field with the help of a cognitive assistant”
-The future of IBM Watson
In IBM’s most recent letter to the shareholders, CEO Ginni Rometty sends a clear message — the future of IBM is in Watson. The letter goes into great detail to describe how powerful Watson is in the field of cognitive business, describes as the convergence of digitization and intelligence. IBM’s Watson represents a break through in artificial intelligence because of the computing system’s ability to understand natural language. Rometty frames the importance of this milestone in this way:
“That potential lies in the 80 percent of the world’s data that is unstructured: everything we encode in language—from textbooks and formulas to literature and conversation— plus all digital video, audio and images. This unstructured data has been essentially invisible to computers. They can capture, store and process it, but they cannot understand what it means. But with cognitive technology, we can now probe this “dark data.
Cognitive systems can ingest it all, and they can understand its meaning, through sensing and interaction. They can reason about it, generating hypotheses, arguments and recommendations. And unlike any computing system we have known, they are not programmed. Rather, they learn—from training by experts and from their own experience. In fact, they never stop learning.
Cognitive includes—but is broader than— artificial intelligence, machine learning and natural language processing. And its embodiment is Watson.”
The question then becomes:
The answer lies in the source of IBM’s revenues:
As much as IBM may outwardly want to project an image as a computing company, 60% of their total revenue comes from their Global Technology Services and Global Business Services segments. These segments primarily sell IT management and consulting services to enterprises. This means IBM must compete with enterprise consulting companies such as Ingam Micro, Accenture and TCS for highly lucrative enterprise contracts. Most of the largest companies in the world fall into the following sectors: financial services, telecommunications, oil and gas, and healthcare. It is no wonder then in an interview in 2011, Manoj Saxena, the GM of Watson Solutions said the following:
“Watson’s capabilities will be particularly powerful for information-intensive industries such as healthcare, government, telecom and financial services — where the volume and variety of information is constantly changing.”
IBM Watson allows IBM to follow the framework for building a successful platform:
- IBM builds a highly differentiated product (Watson) that they can use to get into the door of these large enterprises by leveraging its unique computing power — especially the ability to understand natural language in data intensive industries.
- IBM gains more and more insight into their enterprise customers wants and needs
- IBM builds superior products and services for these enterprise customers
- IBM monetizes these products to customers on platform.
The Future Growth of IBM
Rometty also said the following in her most recent letter to shareholders:
“The word “platform” is important. Much more than simply a faster and cheaper way to access IT, a cloud platform is a new model of innovation, manufacturing and distribution. Cloud platforms provide an open environment for collaboration and rapid scaling. They expose growing libraries of APIs from which partners and third parties across a broad ecosystem can create new, innovative solutions. And cloud offers access to multiple data sets and relevant expertise — not only about technology, but also from business and societal domains”
The future potential of Watson and IBM may be fully realized by the more than 2,500 developers and startups who have reached out to the group since launching in 2013. Like Amazon did with AWS, IBM can use Watson as a platform to successfully build out a new business. Where AWS saved startups from the upfront, fixed costs associated with data services, IBM can significantly lower the costs in startups building their own machine learning. IBM has made Watson’s machine learning service available via the IBM Watson API whose utility is in preparing, processing and analysis of data.
IBM Watson is differentiated in its ability to communicate via listening, watching, talking and even understanding large amounts of big data, even in natural language. Developers are able to leverage this awesome performance of Watson by embedding these capabilities into their own products and services. IBM can then apply the same strategy of providing the API, capturing data about their customers and then building customized products and services they can sell to those customers.