IBM’s Deja Vu in Disruption
Is IBM making it's biggest mistake since 1970 by not fully embracing open innovation?
IBM, a computing and business services powerhouse with a lifespan of over 100 years, has almost 400,000 employees in 175 countries. IBM currently has six business units including IBM Cloud, IBM Watson, IBM Services, IBM Systems, IBM Security, and IBM Research . One of IBM’s guiding values is “Innovation that matters, for our company and for our world” . Though IBM indeed has a history of fueling and funding innovation through research and development, they have also been on the opposite end of industry-wide innovation that they produced – losing a valuable first-mover advantage.
IBM misses its boat
In 1970, IBM published research that would soon change the computing and software industry for years to come. The paper introduced a new concept of “relational database”, a new type of database for storing and querying data using relational algebra and relational calculus . Despite the overwhelmingly positive reception of this theory in the academic computer-science community, it took years for IBM to bring a product reflecting this innovation, System R/SQL, to market. In between that time, Larry Ellison read the aforementioned IBM paper and subsequently founded a company that developed and licensed their own relational database technology . That company, Oracle, beat IBM to market and eventually grew to become not only one of IBM’s biggest competitors (in its software business lines), but also one of the largest tech companies in the world.
Fast forward to 2018, and IBM continues to push the envelope in revolutionizing the way we think of business. IBM Watson is making strides in AI/machine learning and IBM Blockchain is mainstreaming the adoption of enterprise blockchain solutions. However, as history repeats itself, IBM Research continues to publish pieces on disruptive technologies and practices, without putting its findings into practice. One example of this is IBM’s praise of open innovation but reluctance to apply it to its own technology.
Open Innovation and IBM
Open Innovation is essentially the practice of organizations utilizing contributions of entities outside of their organizations to improve upon and/or build products, services, and technology. Open innovation is important to IBM’s management of process improvement and product development because industry scale open innovation threatens the livelihood and existence of IBM and every other technology company. Technology that is developed by interested participants free of corporate dichotomy is arguably comparable, if not superior, to technology developed in closed systems and private enterprises. By restricting contributors to employees tied to incentives of traditional employee-manager relationships, a firm is unable to compete with technologies produced by developers who willingly participate without direct compensation, purely via satisfaction, enthusiasm, and indirect incentives for the technology being built . “No matter who you are, most of the smartest people work for someone else” . Examples of open innovation in open source networks include the Linux operating system and Bitcoin.
IBM and several other technology firms have typically utilized open innovation in encouraging employees to work on products/services outside of their day-to-day responsibilities, successively patenting and incorporating these innovations in incumbent products or new services . As such, IBM’s management is slowly incorporating open innovation to its business models and employing open innovation on small segmented products, but not core business products – presumably due to fear of losing competitive advantages due to intellectual property sharing. However, if IBM does not incorporate open source contributions on its core/patented technologies and processes, it is at risk of being disrupted by truly open source systems.
Two examples of IBM business units that are at risk of being rendered obsolete are:
- IBM Watson
- Though IBM Watson pulls information from open sources, it is primarily of closed-source technology and infrastructure. If open sourced, it can be argued that the efficiency and progress of Watson would be multiplied. Because outside developers can not contribute to the development of Watson, Watson will likely be outpaced by an organic open sourced machine learning technology collectively built by passionate, project-motivated individuals.
- IBM Blockchain
- IBM blockchains are private and tailored to enterprise clients. However, IBM competes with public blockchains that cost nothing for a business to use and implement.
I recommend that IBM aggressively pursue opportunities without enforcing patents, intellectual property, and for-profit infrastructure. Instead, they can indirectly recruit the world’s best developers to work on internally incubated projects and use the resulting technologies as base layers to products and services for existing clientele. Just as the internet is built on open source protocols (TCP/IP, email etc), so too will truly transformative technologies.
Two questions for my classmates:
- Are the benefits and risks of open innovation mentioned above enough to warrant immediate action by IBM?
- How can IBM fully embrace open innovation while not innovating themselves out of business? What ancillary services and business models can IBM take advantage of if they can no longer monetize proprietary technology?
 IBM. “IBM Annual Report 2017” 2017. [Online] https://www.ibm.com/annualreport/2017/assets/downloads/IBM_Annual_Report_2017.pdf [Accessed: 13-Nov-2018]
 IBM. “2015 Corporate Responsibility Report” 2015. [Online] https://www.ibm.com/ibm/responsibility/2015/at_a_glance/our_approach.html [Accessed: 13-Nov-2018]
 Springer. “Encyclopedia of Database Systems” 2009. [Online] https://researcher.watson.ibm.com/researcher/files/us-dchamber/SQL-encyclopedia-entry.pdf [Accessed: 13-Nov-2018]
 Julie Bort. “Ellison Grew Rich From IBM’s Idea” 2014. [Online] https://www.businessinsider.com/ellison-grew-rich-from-ibms-idea-2014-9 [Accessed: 13-Nov-2018]
 Chris Dannen. “Iterative Capital Thesis” 2018. [Online] https://iterative.capital/thesis [Accessed: 13-Nov-2018]
 Karim Lakhani, Jill Panetta. “The Principles of Distributed Innovation” 2007 [Online] https://www.mitpressjournals.org/doi/pdf/10.1162/itgg.2007.2.3.97 [Accessed: 13-Nov-2018]
Student comments on IBM’s Deja Vu in Disruption
I think IBM actually agrees with you. They made some of their enterprise blockchain software open source on the Linux Foundation. https://www.linuxfoundation.org/blog/2018/09/how-ibm-is-using-open-source-for-a-greater-good/.
I think the risk with open innovation on ERP systems or infrastructure systems is security, how do you make sure that the programs created are actually secure for your end customers when you have no control of the innovation?
As someone who knows very little about open innovation, I find your article extremely thought provoking. All the companies that I have worked for have been extremely diligent in protecting their IP, so the thought that IBM should operate “without enforcing patents, intellectual property, and for-profit infrastructure” to be completely terrifying. Although if the upside of doing so is that they get greater innovation than they otherwise would, then maybe then it’s worth it. But in this scenario, I can see how society would benefit, but if they won’t be able to control or monetize it, how can IMB really capture the value?
Thanks for sharing, Jason. I believe that the most relevant business case for open innovation at a behemoth like IBM pertains to driving engagement and product awareness. While other firms may use open innovation to, well, innovate, I believe that IBM is less inclined to step outside of their organization for such purposes, as they are aiming to insulate their credibility and potential future success “claims.”
Thanks for sharing! Like MM, above, I wonder a lot about the potential risks of open innovation in a private business context. In my work experience, there is tremendous value in having access to IP that others don’t. In order to pull this off, would you have to change the IBM model? Or would you simply protect the base platform with some sort of patent? Or maybe it would be more incremental changes and not expose the bulk of the Watson IP? My sense talking to a few of the people in the section is that open-source is often used in academic settings and not really for business where you risk a very valuable underlying asset, but truly not sure.
Thanks for the article! Like some of my section-mates, I also have a limited understanding of open innovation and initially think that making its most valuable IP would present a real risk to the value of the company. I may be missing the point, but it seems that the reference to the “relational database” article that spawned Oracle, a major competitor of IBM, may be why they are more closed with their technical advances. That said, I recognize that the culture in tech is very supportive of open innovation. Therefore, to acquire technical talent, I would consider selecting several strategically valueless, yet highly interesting, pieces of research and releasing it to the broader community. This then starts to create a reputation for open innovation, helping with recruiting while still protecting IBMs valuable IP.
Interesting article – however, I think the call to abandon a for-profit infrastructure is too extreme, especially for a company that only relies on monetizing technological advances.
However, IBM has actually already made large strides into the open innovation space just in the last few days. They acquired Redhat for 34bn dollars and have announced that they will stay true to the open innovation background of Redhat to provide an open approach to cloud.
Your point that Machine Learning and AI might go the way of TCP/IP protocol is interesting. Like the internet infrastructure we could think of ML/AI systems as a platform that facilitates, but does not directly contribute to value generation. That is, there is orders of magnitude more value in the amount of consumer data that IBM holds than the actual ML/AI algorithms that actually can process it. If IBM focuses on strengthening it’s collection and implementation of user data as a service and opens up its IBM Watson system to the open-source community it would probably be in a much better position.