Updated October 13th, 2017
Bimodal IT promises to quickly deliver the benefits of a digital innovation center without having to face the challenge of addressing IT’s legacy organization and processes. But is it too good to be true? And how does it compare to a full-scale Agile transformation? This white paper looks at bimodal IT and offers timely guidance.
Anyone monitoring the voluminous flow of words advising CIOs how to “align with the business”, “deliver more strategic value”, and “implement a modern digital strategy” has witnessed the debate about “bimodal IT”. Gartner introduced the concept of bimodal IT in 2014, recommending that CIOs run two distinct modes of project execution in their corporate IT shops. Mode 1 is deliberately slow, “traditional and sequential, emphasizing safety and accuracy”. Mode 2 is fast and responsive, designed to incorporate customer voice, learn and iterate towards better products quickly.
Predictably, several large traditional consultancies immediately offered CIOs their off-the-shelf bimodal or “digital” implementation services, promising quick and easy ROI by using value-driven, customer-centric, Agile approaches on a few projects, while conveniently leaving slow legacy structures and processes unchanged for most of the organization.
The need for dramatic improvements in user focus, accelerated time to market, and improved responsiveness is common in Fortune 500 IT organizations. Market-dominant firms tend to become inwardly focused and bureaucratic, less innovative, slow and risk averse. Disengaged from customer voice, IT ships poor products full of unused features. Employees disengage as knowledge workers and become mere order takers, their work lacking purpose and their mission increasingly obscure. Ideas for new products or business model innovation is stifled in the interest of avoiding the “cannibalization” of cash cows. Organizational performance suffers, producing slower revenue growth and increasing threats to market share. The benefits of bimodal IT, therefore, are quite attractive to these types of organizations.
Introduced by Gartner in 2014, bimodal IT encourages CIOs to separate their organizations into Mode 1 and Mode 2. Mode 1 is for the “systems of record” and the people who support them using legacy waterfall processes, delivering slowly and sequentially in order to ensure safety and accuracy. Mode 2 projects are delivered incrementally with frequent rapid deliveries to incorporate customer feedback and support adaptation.
Is BiModal IT a Breakthrough Insight or a Convenient Half-Truth?
By promising to quickly deliver the benefits of a digital innovation center without having to face the challenge of addressing IT’s legacy organization and processes, bimodal IT almost seems too good to be true. For technology organizations considering investing in a significant performance improvement initiative, I have prepared a comparison of the relative merits of the siloed bimodal approach typically espoused by consultants to a more holistic enterprise-wide Lean/Agile transformation approach, in which bimodal (or multi-modal) IT is a transitional state in the journey to a high performance organization.
The Benefits of Bimodal
Bimodal IT is attractive to IT organizations facing problems of speed and responsiveness, and the approach can deliver modest benefits, at least for the Mode-2 portion of the portfolio. These benefits include:
- By increasing focus on incorporating customer/user feedback, you can deliver better products than IT organizations typically deliver, according to industry statistics.
- Faster cycle times at all stages of the delivery lifecycle speed up time to market.
- You deliver more value per dollar invested.
- Risk is reduced and ROI increased because you break large projects into smaller increments that deliver value early and often and that surface problems earlier with less sunk investment.
- By leveraging more modern technologies and focusing more on innovation, you can rapidly experiment using validated learning, “fail fast” mentality, and quickly pivot to better solutions, like a Lean startup
- Acquisition and retention of top talent improves because individual passions are engaged when they are empowered to deliver compelling solutions that thrill users and drive markets. In other words, to develop software like a real Silicon Valley tech company!
The Problem Is Real
The appeal of bimodal IT stems from a problem almost as old as IT itself – older technology systems in “support” or “sustainment” mode typically contain high levels of technical debt, making changes slow, risky, and unpredictable. CIOs know their organizations and legacy applications are devilishly difficult to accelerate and improve, but pressure from line-ofbusiness leaders to support innovation and deliver digital solutions more rapidly is increasing daily. Line-of-business increasingly bypasses IT by outsourcing development or by building shadow-IT organizations. This threatens corporate IT’s influence and resources.
By promising to quickly deliver the benefits of a digital innovation center without having to face the challenge of addressing IT’s legacy organization and processes, bimodal IT seems to promise results that are almost too good to be true.
The IT press has published a great deal of analysis of bimodal IT, attempting to articulate implementation specifics and address practical considerations of bimodal IT omitted by Gartner.
In “What Gartner’s Bimodal IT Model Means to Enterprise CIOs”, CIO Magazine despaired that “Gartner has thrown in the towel”, abandoning hope that traditional IT can ever fundamentally improve. The author, Bernard Golden, illustrates concerns about the human consequences of creating “fast” and “slow” silos in an IT organization: “The implication of bimodal IT is that traditional IT is facing an awful future – one where it’s consigned to keeping the lights on for legacy applications but precluded from participating in what’s viewed as critical to the future of the company at large… This is one place that I believe Gartner fails to comprehend what this bimodal reality brings – or at least prefers to not address it explicitly to avoid frightening its clients. As this bimodal reality sets in, one can expect many companies to experience huge conflict as the two camps engage in pitched battles for influence, resources, and power.”
Forbes, in “Bimodal IT: Gartner’s Recipe For Disaster”, suggests that bimodal IT is merely “old wine in new bottles” doomed to be ineffective because “The central challenge with bimodal IT… is that it encourages IT management to shift their transformation efforts away from the slow, mode-1 IT. Transforming traditional IT is difficult, Gartner would seem to imply, so don’t bother. Gartner may be telling CIOs what they want to hear, but not what they need to hear.” “What many organizations are finding is that for digital transformation to be successful, it must be end-to-end – with customers at one end and systems of record at the other.” Is Gartner providing us a breakthrough insight to rapidly and inexpensively achieve order of magnitude improvements in IT performance, to win in contemporary markets? Or is Gartner pandering to laggard CIOs, reassuring them that their slow inefficient organizations are now respectable “mode 1” shops? Gartner prescribes that mode 1 should be “IT-centric” and “removed from the customer”. But is having most of IT slow and disengaged really the new state-of-the-art?
“Gartner may be telling CIOs what they want to hear… not what they need to hear” -Forbes
Keep the Goal in Mind
If your technology organization is considering investing in a significant performance improvement initiative, don’t forget the core reason for doing it: to deliver more value more quickly to users and stakeholders, internal and external. It’s imperative that we assess the merits of the siloed bimodal approach in comparison to a holistic enterprise-wide Lean/Agile transformation approach, in which bimodal (or multi-modal) IT may be a transitional state. In comparing the advantages and disadvantages of these two approaches, I’ll focus on the following points:
- Big benefits require optimizing the entire value stream.
- There are more than two categories of work, thus more than two modalities.
- Quality is not optimized by isolated silos delivering slowly.
- Creating more silos in IT to implement bimodal hurts more than it helps.
- Techonology organizations don’t achieve leaps in effectiveness without courageous leadership.
Siloed Bimodal IT
Given that organizational silos, and the parochial behaviors and local optimizations that result from them, have been identified as a significant impediments to organizational alignment and efficiency, it seems odd that Gatner should propose creating a new one at this stage of IT’s evolution. While it’s clearly quicker to stand up a new silo, like a lovely little hothouse garden in the arctic waste, is this a sustained improvement in the long term? Or does it create more problems than benefits in the long term?
1. Value Stream Optimization
Siloed bimodal IT narrowly addresses the project delivery process, mainly in the development phase. In other words, it’s a local optimization, and only for mode 2 projects. In Mode 2, detailed requirements, coding, test, and deployment may leverage Agile processes, but bimodal explicitly avoids addressing critical systemic impediments up and down the value stream. For example:
- IT’s legacy budgeting, governance, and portfolio management processes often lock in projects’ sequence and scope a year in advance. When markets move, users’ needs change, or exciting new opportunities emerge, it is difficult to adjust and adapt.
- Traditional resource management models shuffle people between projects. They encourage starting projects over finishing them, overtaxing capacity, multi-tasking, generating waste, and harming employee engagement and accountability.
- Production Operations, environment provisioning and configuration, and deployment processes are usually owned by a disengaged silo that works tickets to a SLAs. They would rather avoid changes than embrace them.
- Organizational roles and structures are typically optimized for functional specialties in silos, not efficient value stream delivery. Incentives reward individual heroics over team cohesion and accountability. Competition and gamesmanship are rewarded more than collaboration.
- Lack of LOB alignment with delivery teams, reflected in poor accountability for feature prioritization and timely feedback, creates an “order taker” mentality instead of the partnership essential for rapid learning and adaptation.
- Traditional organizational culture and leadership don’t reward “fail fast” or rapid learning. In order to innovate and respond to market demand, we must rapidly experiment, measure, and pivot to deliver better solutions.
- In high-performance organizations, leaders value and support learning, exhibit a passion for continuous improvement, and empower knowledge workers in high-performance teams aligned to customer voice.
2. Multiple Modalities
Siloed bimodal IT offers two speeds. We’re to create a new mode 2 sub-organization, an island where “fast” can thrive somehow immersed in the ocean of “slow”. We establish “objective” criteria for which projects get voted on or off the island, yielding huge consequences for stakeholders and IT employees alike. Consider the wide range of systems and delivery models in your enterprise: old retiring applications in sustainment and maintenance, critical systems of record that vend services to customer facing apps, back-office systems that drive operational efficiency, consumer-facing systems that define your brand. Big IT delivers these systems through global system integrators, IT employees around the globe, and cutting-edge partners bringing valuable UX and Lean Startup skills to bear. Do two modes optimize all the combinations of stakeholders, goals, and assets in this matrix?
Siloed bimodal IT perpetuates a discredited myth: phased development performed in silos that infrequently throw huge batches of artifacts over the walls to one another somehow produces higher-quality product than well-aligned, collaborative, cross-functional teams who deliver smaller increments of value frequently. Contrary to the guidance in mode 1, decades of accumulated evidence from a range of industries demonstrates that quality, predictability, efficiency, and risk management are all negatively correlated to large batches, long-cycles, phase-separation, and functional silos. In other words, quality/stability versus agility is a false dichotomy.
Quality/stability versus agility is a false dichotomy.
4. More Silos, More Pain
Siloed bimodal IT suggests that walling off mode 2 projects and people in a separate operation will free them to deliver more value faster for a net benefit to the organization. There’s no need to change the bulk of the IT organization because mode 2 is effectively a different IT organization set apart, or a sub-organization that can ignore the people, processes, and technology of the IT of old. Bimodal really only appears viable if you believe the two modes require minimal interaction, and that mode 1 will not influence mode 2 to revert to status quo behaviors. This view ignores the inconvenient reality that many of the most valuable mode 2 projects will depend on mode 1 people and systems to deliver meaningful value to users.
5. The Need for Courageous Leadership
Siloed bimodal IT promises increased value delivery and delighted stakeholders by incorporating feedback faster, reducing waste, and improving alignment, and (most importantly) changing the approach on a minority of projects, and only for part of those projects’ lifecycles. The majority of the organization would proceed along unaltered. Everything in the organization has, for years, been optimized for “slow”, yet we are promised somehow that the inertia of the organization’s legacy processes, controls, policies, roles, incentives, and culture will not cause mode-2 delivery to resemble the dominant mode-1 characteristics a little more each day. Are we preserving outmoded legacy processes and behaviors that undermining our business goals out of fear, ignorance, or complacency?
Given these considerations, ask yourself: Is bimodal IT really a strategic path to optimizing IT value delivery? If not, what is the optimal path to achieve transformative performance improvements?