The Risks and Rewards of Bimodal IT
Today's tech challenges increasingly require both traditional and experimental solutions. Here's how a bimodal approach can impact the IT organization.
IT organizations are increasingly combining a conventional approach to core systems with a more agile, exploratory approach to projects with greater uncertainty — a pattern that research firm Gartner has described as "bimodal IT."
"Bimodal is a marriage of two distinctly different but coherent approaches to delivering business change," said Simon Mingay, a Gartner research vice president, during a recent webinar presentation. "Mode 1 is the traditional rock-solid approach that we all know and love and Mode 2 is more iterative and nonlinear and tends to be applied to projects of significant value to the organization as they develop a digital capability at the pace they need to adopt."
Mode 1 tends to involve core applications such as enterprise resource planning and customer relationship management, while Mode 2 is initially applied to small but important projects — typically things that customers will touch, feel and see. "Part of the initial challenge is to get it recognized in order to get buy-in from executives to expand and fund it," Mingay said. "So organizations tend to do it in something that will be seen and recognized."
A Risky Move
The concept of breaking out teams to work in new ways on issues that hold a lot of uncertainty is becoming more commonplace in larger private-sector organizations. Gartner estimates that 40 percent of enterprises are on the journey toward bimodal, and within the next 18 months it expects closer to 75 percent to be working toward it. "The key value proposition is that it provides the organization with a tool to better manage uncertainty," Mingay said. "You may understand the problem, but you aren't sure what the right solution is."
But Mingay admitted that the move is perilous. "Fifty percent of those organizations will run into trouble with it," he said. "Bimodal is not risk-free by any stretch of the imagination." In fact, he titled his presentation "Bimodal IT: How to Have It All Without Making a Mess."
So what are some of the potential landmines? Mingay outlined several risk factors facing any CIO starting down this path.
1) Lack of communication. If you create separate teams, the two modes can become disconnected and have poor communication. It is important that the two modes have an open, transparent, collaborative style and work together to align priorities, he said. The Mode 2 team will eventually turn to the Mode 1 team for resources, and the two groups need to be able to work smoothly with each other. Mingay used a metaphor of different types of runners: Mode 1 created IT employees who were like marathon runners, while the digital era is requiring more sprinters. "Part of the complexity is that you need to convert some marathon runners into sprinters," he said. "And marathoners and sprinters have different world views, so it is a challenge to get them to cooperate."
2) The timid middle. Some organizations will fail to fully embrace the Mode 2 approach. "Mode 2 can be intimidating," Mingay said. It requires cultural and behavioral change and there is a temptation to 'de-risk' it by applying traditional approaches to project management — and you end up in the timid middle, not one thing or the other. Then you tend to get the disadvantages of both." If you can see value in the bimodal approach, then embrace it in entirety, rather than try to fudge it, he added.
3) "Us vs. them" mentality. For large organizations that actually bifurcate their project teams, it is important to maintain equity between the two teams in terms of rewards and recognition. "It is important to avoid an 'us vs. them' mentality," he said. The two teams are more likely to cooperate if they feel they are being treated in an equitable way. CIOs need to be careful about management language, he added. You don't want to be celebrating this new bright, shiny object and ignoring the people who are doing the hard work of keeping the business running.
4) Interfacing with core systems. Initial Mode 2 projects tend to be quick wins involving topics like social media that don't rely on access to core applications, but Mingay warned that you will run out of those projects eventually and will need to start interfacing with core systems. He recommended renovating the core systems so that Mode 2 teams can build or expand on them in a service-oriented architecture or through application programming interfaces. You want to make sure the Mode 2 team doesn't end up running into an inflexible core.
What Bimodal Doesn't Mean
To tighten up the definition of bimodal and clear up uncertainties, Mingay gave some examples of what bimodal is not.
"It is not just dividing the organization in two," he stressed. It is about distinctly different styles of approach to business change, in which Mode 2 is more iterative and exploratory.
Bimodal is not just about agile software engineering. Although a more iterative approach is one of the core concepts, that is not enough. It is possible to use agile development but apply it in a Mode 1 way, Mingay noted.
Bimodal is not the central IT group vs. business unit groups. "It is the approach, not the organizational structure, that is different," he said. Nor is bimodal the same as "shadow IT," in which non-IT groups develop their own software in the shadows. "Those people may be creating value, but that doesn't make it Mode 2," he said. If it is truly shadow, it breaks the first golden rule of bimodal, which is that it is aligned, collaborative and transparent."
Other examples of the way Mode 2 is different from Mode 1 include governance and sourcing, he said. In terms of governance, both sides have a plan. In Mode 1, as you go through a project you check how well you are sticking to the initial plan. In Mode 2, you are continuously checking against reality, he said. "You get direct feedback from the field and users about how something is actually being used to determine what that tells you about the next iteration."
In terms of sourcing, in Mode 2 you need a way to engage with vendors that might not be the usual suspects, Mingay said. If you are working on mobile and social projects, for instance, you may want to work with smaller, boutique firms with niche capabilities. "You can't go through a six-month RFP process on something that is exploratory," he said. "That will just kill exploration stone dead."
Getting Started
So how does Mingay recommend you get started? First, pick projects that have little impact on Mode 1. "Don't start with transactional systems. That may be biting off more than you can chew," he said. Get started on something without having to go and ask for resources from the Mode 1 team. There is no such thing as an island project, but on the continuum of things ranging from lots of moving parts to relatively simple, start on the simple end, he suggested. Work on something that can be completed within three to six months and that creates external value — something that can be touched by stakeholders. Start with small teams of five to 12 people, he added.
Bimodal presents different options for organizational structure. Mingay noted. Gartner tends to see enterprises splitting up their org charts, not because it is the ideal picture but because their experience shows them that trying bimodal within existing team structures slows down the pace of change. The Mode 2 project people "don't get the freedom they need to break the rules they need to break, and the effort comes to a grinding halt," Mingay said.
Some enterprises have two separate organizations reporting to the CIO, he said, and in other cases the Mode 2 team sits outside of IT and reports to someone with a title such as chief digital officer. "That can work if the two modes have a transparent style," he said. But if the catalyst is that the CEO is frustrated with a perceived lack of response from the CIO, then the organization could face significant costs or risks down the road.
Mingay said you should not overthink bimodal upfront. You should start before you think you are ready, whether it is setting up an innovation lab or adapting a policy of flexible sourcing with small vendors. "Don't think too much about what capabilities you will need and don't overplan," he advised. "There is an ultimate irony if you end up using a Mode 1 approach to this. Find business sponsors willing to engage with you, get started and evolve from there. The important thing is to get going."