A conversation with Jim Colson, Independent Advisor, IBM Fellow — Emeritus and former VP/CTO of Watson Customer Engagement
This is part of my series on thought leaders in the innovation space. Check out the other articles here.
Growing up in Detroit, Jim Colson almost got his start as a mechanical engineer in the automotive industry. After learning that his dream job on the Robotics team at Texas Instruments (TI) wouldn’t be open for another year, Jim accepted a job doing vibration analysis for Rockwell International. During subsequent interviews in the automotive industry, Jim, unfortunately, learned the term layoff. “I interviewed with Cadillac and they said, ‘Here’s your offer and a one year guarantee against getting laid off.’ I said, ‘That’s great. What does ‘laid off’ mean?’”
The reality of layoffs quickly had personal significance when he learned that Rockwell International decided to downsize its workforce and revoke all college offers, including Jim’s. He turned again to TI and learned that they had opted to begin the robotics project a year early and offered him the job. “That’s what got me in Austin, that’s what got me to TI, that’s what got me into robotics. I’ve been in innovative opportunities ever since.”
After spending time in Robotics at TI, Jim realized: “the more interesting problem was — how does an application engineer write a program for a robot when they don’t understand how motion planning actually works?” This led him to move to IBM and begin working on a team building a menu-driven programming system. “I added in the ability to be computer vision guided so we could find things on a conveyor belt or loosely packed in a cardboard box, appropriately find the right position and orientation to pick them up and do the assembly without having to precisely position the objects for the robot to pick up. That got me into computer science.”
As it became clear to Jim that network connectivity was becoming ubiquitous even for non-computer devices, he and three colleagues wrote the seminal white paper for IBM called “The Tier Zero Strategy” about the phenomena now referred to as the Internet of Things (IoT). The white paper questioned how IBM would respond to the emergence of what is now known as IoT devices. Following the white paper, a division was created and Jim was appointed one of the first four employees. “That was probably the closest thing I’ve had to a ‘sit and think’ job, but we had revenue and product targets. We had ecosystems to build and partnerships to establish. There were a lot of business metrics associated with that work.”
The limits of innovation measurement
Even very early in his career, Jim notes the presence of measurement. He shared that useful measurement can change based on an organization or program’s maturity. “It’s very hard to get [measurement] ‘right’ where ‘right’ means: repeatable with direct correlation to downstream success measures. Early in the innovation cycle, a number of ideas per employee is reasonable, but ultimately ideas need to be realized and that leads to direct success measures like revenue, profit, market share, etc.”
When I asked Jim about innovation programs that are shut down due to a lack of immediate ROI, he first acknowledged reality before offering a trade-off. “It’s hard to instill patience in somebody who is inherently impatient.” One way to mitigate the loss of information that comes from a program shut-down and contend with impatience is through patent filing. “Somebody will have an innovative idea. They’ll be able to describe its conceptual model sufficient for somebody skilled in the art to implement it, but they don’t have the time, resources, or the patience to implement it themselves so that becomes an invention disclosure. You submit that for potential protection from a patent standpoint, and if that idea, which really does represent a conceptual model and innovation, turns out to be valuable then there’s economic gain from that idea even if you never physically manifested it yourself. It’s a way to approximate instant gratification as compared to the actual realization and physical manifestation in the industry because you don’t have the resources to manifest all ideas.”
When it comes to funding the execution of innovative ideas, Jim points out that venture capital funding has an advantage over internal innovation programs in their increased tolerance of risk due to the economy of scale. In measuring the impact of innovative ideas and programs, Jim calls attention to the reality of innovation through some personal experiences. Through the Pervasive Computing division at IBM (today known as IoT), Jim and his team created a lot of IoT patents and technology innovation that is ubiquitous today. “The syncing of calendar entries and address book entries from your mobile device to the cloud and to other mobile devices was done through an organization that we helped create and lead, which was widely participated in called SyncML, Sync Markup Language. Virtually all the synchronization that occurs today is derived from that core intellectual property that we innovated in that body.”
“The syncing of calendar entries and address book entries from your mobile device to the cloud and to other mobile devices was done through an organization that we helped create and lead…”
Another example: “Deducing traffic by triangulating the position of cellular devices en masse on the highway, I patented with three colleagues and now it’s used everywhere. We implemented it, but we really couldn’t monetize it because it was just too far afield from a business standpoint to where the company was at the time. Had we been doing it now where there is a focus on IoT, maybe it would have been different.”
These examples illustrate that innovation is often disruptive and even great ideas fall apart because they don’t have a method of realization within the organization in which they were first gestated. While business goals and measurement are integral to innovation, it’s important to also recognize that not all aspects of innovation are fully measurable. The class libraries Jim and his team built for their clean room Java virtual machine, for example, were the foundational libraries used in Android.
“Innovation doesn’t necessarily get manifested and delivered through the organization in which the innovation occurred.”
Conceptual models in innovation and design thinking
In order for an innovative idea to be measured, it must first pass the test of understanding. Jim finds that taking an idea from the verbal stage to written form is the first step to vetting its value.
“I find that a verbal idea has to go through the crucible of writing. No matter how many times you say it to somebody, once you actually write something down, that crucible will force you to recognize whether the idea is cogent or not.”
Jim points out that the need for articulating an idea in writing is at the core of intellectual property as well. “When you write down a disclosure for potential patent submission, you have to write those kinds of things down to say this is the conceptual model of this activity.”
The writing stage then leads to the creation of a conceptual model. Jim finds that people practicing design thinking can get stuck on the visual aspects of a concept before exploring its functional value. “I don’t even like the phrase ‘look and feel’. If you interpret that phrase correctly it has proper meaning, but too often it’s interpreted to mean just the visual veneer of some conceptual idea.” Jumping to a visual representation too soon can have unintended consequences of causing engineering teams to check out of the design thinking process.
A conceptual model describes key artifacts of an idea, the relationships to those artifacts, and their movement or velocity — not just their starting point but the change in state over time as they are manipulated. A conceptual model for banking, for example, could be the idea of making a payment with a credit card. It’s not about the specifics of swiping a mag stripe, entering a number on a website, or inserting a chip. The model describes the overall concept — that you have a bank account with a certain amount of money in it. You may also have a relationship with another completely different bank who has given you the notion of credit based on your history of paying debts. When you make your purchase the credit card company makes the payment on your behalf knowing they’re going to send you a bill to pay them later and understanding that you will give them money out of your account to cover the purchases. Describing this process and flow is a conceptual model.
Jim learned how to best articulate his thoughts about conceptual models in the design process from a distinguished engineer at IBM named Carolyn Hyink. She gave him a conceptual model that, generally, describes any system design. A design is three-fold including a conceptual model of an idea (like the banking example above), a set of interaction models used to manipulate and move the conceptual model forward to execute its business process, as well as a set of visual models used with the interaction models to manipulate the conceptual model. “Thanks to Carolyn, once I had my own conceptual model around what a design was, then I could embrace the other things that were happening around design thinking that I was dismissive of because I looked at those as being only surface deep so to speak.”
Jim advocates starting with the conceptual design early in the process of innovation as a way to engage engineers and as a way to foster better communication between design and engineering teams. “Now I can effectively converse with designers. They may not have the vantage point in terms of all three of those elements of a conceptual model, but I can much more easily connect with folks in the design realm around what they’re trying to achieve. If I feel like they have a blind spot, I can, in a more informed, nuanced way, ask questions to tease out whether it’s truly a blind spot or it’s just not being articulated. Conversations are richer, the outcomes are more productive, and I think that when engineers get their head around that topic the engineering community itself will completely embrace design thinking.”
In addition to engaging engineering teams and improving communication, conceptual modeling also prevents teams from designing for a miracle.
“If you don’t really have some notion of what the conceptual model is and some reality check on what’s possible and not possible, it’s very easy to start on the outside in and design a miracle without even realizing it.”
Loose structure and a career path for innovation
In structuring innovation programs, Jim believes there are elements of culture and structure at play. To start, career paths must foster innovation. “Demonstration of innovative thinking should be rewarded just like delivering on your day job. The way you motivate people is by rewarding them in their career path that the more they swim outside of their swim lane, the more senior they can become because they are demonstrating that they’re working on behalf of the firm and not just doing a job.”
“The way you motivate people is by rewarding them in their career path that the more they swim outside of their swim lane, the more senior they can become…”
As seniority increases, reliance strictly on day job duties should be diminished in gauging job performance. “If you want to be part of a company that’s going to be innovative and transform itself multiple times in its life, you’ve got to be thinking about the next thing. Your leaders have to demonstrate that they’re thinking about the next thing and, at a minimum, at least communicating that and sharing that with others so that those ideas can build on each other and maybe find a way outside of a given budget and execution plan to execute and make something happen.”
Jim encourages organizations to allow motivated individuals to form teams around valuable ideas that catch a spark and drive innovation. “There’s some top-level guidance that has to be established to foster that mentality, but it’s almost purposefully void of any particular structure because you’re allowing people to do the very innovative self-discovery, self-forming of teams and execution that would occur in the wild outside of the place you work. That’s how startups happen.”
Barriers to innovation in large organizations
The day job is one barrier Jim sees to innovation in large organizations. “We’ve got more things to do than we have the time or people to do them.” Tasks of a day job are also more likely to be over-prescriptive when it comes to execution. Rather than telling teams what their tasks are and how to accomplish them, Jim advocates setting goals for teams and letting them innovate around achieving those goals as a way to demonstrate their ability to be creative in their execution. “You have to relax those constraints, and let people be creative.”
Jim also mentioned the influence of memory within an innovative organization. When it comes to evaluating individuals, Jim believes a more transient memory is appropriate. “If somebody has a string of ideas that don’t go anywhere, it shouldn’t become an anchor to them and it shouldn’t become a negative on their career aspirations.” As long as an individual continues to fulfill their day job and shows improvement Jim sees positive progress. “Sort of blur your eyes and recognize this person had several ideas, they’re trying. They’re motivated, they have passion, they definitely want to swim outside of their swim lane. They want to get something done. To that point, I say that’s sort of a transient memory.”
The flip side where memory needs to be more permanent comes in when tracking innovative ideas. “There needs to be some persistence of those ideas and some ability to find them, index them, and recognize them so that an idea that was created, evaluated, and dismissed isn’t ginned up from ground zero again. It might be reevaluated later because the timing was wrong. [Organizations] shouldn’t just cast away all ideas that are not being pursued at this point in time because history will inevitably be repeated. Part of the mitigation of this issue is motivation for invention disclosures we discussed earlier.”
Don’t innovate for buzzwords
“What I see happening, which often ends in complete disaster, is somebody will get so enamored with a technology that they’ll just try and make it fit everything.”
Innovation has to have a value proposition and a vision for some desired outcome. Innovating around the latest, popular technology is one area where Jim sees well-intentioned efforts go astray. “What I see happening, which often ends in complete disaster, is somebody will get so enamored with a technology that they’ll just try and make it fit everything.” When people tell Jim they need to implement AI or machine learning his first question is why. If they can’t articulate a reason, they’ve typically fallen prey to innovating for buzzwords. “They don’t look at it as a tool in the toolbox to use to drive an outcome. They just feel they have to do AI and machine learning. Some of those conversations have been so contorted that when I probe on it, they don’t even know what AI and machine learning are. They just know the buzzwords.”
If you want to read my other articles about innovation experts and practitioners, please check them all out here.
