Why Digital Transformations Fail and How to Fix It

Changing your operating model to improve the odds of success?

| “We can never do merely one thing.” – Garret Hardin?

Digital transformation continues to be a top priority for organizations. Investment in digital transformations is expected to reach $3.4 trillion worldwide in 2026, according to International Data Corporation’s Worldwide Digital Transformation Spending Guide. Yet only 30% of digital transformations succeed in meeting their objectives, according to BCG research. One of the most influential factors determining success or failure is the type of operating model an organization uses in its transformation.??

CEOs kick off a digital transformation by creating a centralized digital team responsible for digital transformation projects. That makes intuitive sense because they want to dedicate proper resources and ensure accountability. Their argument is that a centralized digital organization can maximize and incubate limited technical expertise, decrease enterprise risk through standardized protocols and architecture, reduce overall cost, deliver both operational efficiencies and effectiveness (e.g., increase speed to market, deliver higher quality deployments), build a richer ecosystem of partners, and unlock innovation.???

Yet, in adopting this model, leaders often fail to account for the complex psychological dynamics of centralization. A monolithic centralized organization can lead to unintended consequences, creating a static, insular, and hierarchical structure that impedes progress and stultifies healthy competition.?

The law of unintended consequences?

A centralized model can unintentionally create disincentives, confusion, power struggles, and misdirected effort and resources. Although the goal is to maximize technical expertise, centralization can insulate expertise, which results in gaps in end-user knowledge and skills and becomes apparent at the critical point where digital technology must deliver business value. In other words, you may now have the screws, drills, and planks, but don’t have the knowledge to build the house. Standardization and cost-control efforts can lead to inadequately allocated or misdirected resources. In a multi-brand scenario, smaller brands/businesses feel ignored while bigger brands/businesses complain that solutions are not customized to their needs. Centralized teams become persistent bottlenecks. The organization may then react by increasing resources as a solution to operational problems, leading to the ballooning of central teams. The bigger teams in turn add more layers of administrative tasks—directing, clarifying, checking work—but more work isn’t necessarily completed, and the core problems persist.??? ?

As problems multiply, managers in monolithic organizations try to counter the negative narrative. They become less transparent with roadmaps, timelines, and deliverables. They may blame failures on the purported digital immaturity of their business counterparts. They may hail marginal improvements as triumphs. Culture deteriorates and becomes hostile. Organizational politics and positioning replace open, honest, and direct communication. Courting the leaders of the digital transformation project has become more important than delivering business or customer value. Apathy grows because business teams feel unable to get the right support and central teams come to see their job as a digital help desk.???

Most CEOs are either unaware of these problems or unwilling to address them until they reach a crisis. CEOs often avoid arbitrating between central digital and business teams. Yet, by distancing themselves, CEOs lose the ability to identify and potentially solve problems before they become unmanageable. Even when CEOs recognize there’s a crisis, they may avoid facing it head-on because that would require acknowledging significant gaps in digital capabilities that can risk board and investor support. ?

Why technology is not a lightbulb?

| “New metaphors are capable of creating new understandings and, therefore, new realities.” – George Lakoff?

A monolithic approach treats technology like a utility. A goal of technology is cost reduction through scale. A provider of electricity has a high starting cost but lowers costs and increases efficiency by expanding its network. The more nodes you link, the lower your per-node cost. However, a more useful and productive metaphor is to think of technology as blank canvas. An empty canvas has a low starting cost, but value can accrue depending on what you put on it.??

This analogy helps reframe the conceptual framework for transformation. Turning a blank canvas into valuable artwork (similar to using technology to add value to your organization) requires a diversity of knowledge and skill, from technical to design to creative. This is the inverse of a utility where value is generated by focusing on technical skill sets. In a work of art, value emerges as you create it. In other words, you find emergent properties. With a utility, unplanned events—like surges—have negative consequences. Finally, a valuable work of art requires artists who are bold experimenters, risk takers and learners. A utility uses a team of technicians focused only on precision and efficiency.?

7 steps to de-risk digital transformation??

By using the art metaphor, leaders can take practical steps to develop an integrated modular operating model that doesn’t have the typical problems of a monolithic centralized digital organization.?

1. Risks and Incentives: Create a meaningful and clear upside for achievement and downside for poor performance at both organizational and individual levels.??
2. Feedback and Measurement: Establish a tradition of consistent assessment and feedback. Like risks and incentives, effective feedback should happen at multiple levels: function, team, leader, and individual.???
3. Modularity and Teaming: Break large projects into smaller, flexible parts and build discrete, multi-disciplinary teams that have a clearly defined mission. For each team, identify an executive sponsor with the responsibility to deliver.???
4. Iterative Test and Learn: Create a cycle of consistent improvement, starting with a “minimum viable product” (MVP). Build, refine, and iterate to unlock more value.??
5. Rule of Law: Anticipate conflicts and codify how to escalate problems and expedite resolutions.????????
6. Competition: Build a rich ecosystem of services provided by both centralized teams and outside third parties.??
7. Knowledge Exchange: Institute mechanisms to create and deepen collaboration, share knowledge, and foster trust.???

Having only one or a subset of these system components is not enough; in fact, it could be damaging. This integrated system needs to be in place first before the transformation starts, yet most CEOs miss this critical planning step.????????

Modular op-models in action?

Our team supported a full-scale digital transformation at a leading consumer packaged goods provider. The company invested $500 million over five years and expected sizable return on capital.?Evolving customer experience was at the core of the transformation. We advised the Chief Strategy Officer (CSO) to launch an acceleration program early in the transformation.???

To encourage modularity and teaming, we helped the client establish an accelerator pod of ~8 people to drive the work. We brought together a diverse multi-disciplinary team of experts on personalization, media, analytics, consumer data and technology. To align risks and incentives—as the brand still owned delivering their quarterly plan—a marketing director for?the brand became the pod’s “quarterback,” retaining control over defining use-cases and aligning budgets.? Members of the centralized team became embedded in the pod by supporting the quarterback; both the individual as well as the brand’s success became their primary focus, aligning incentives at two critical levels. The CSO became the executive sponsor to whom we could directly escalate problems.?

To build a testing muscle, we started with a set of “quick wins” use-cases—potentially high impact but lower investment. To ensure effective measurement, we standardized metrics and sized potential return for each use-case. We had a rich set of service providers to support use-case activation, which encouraged competition. Use-cases that proved effective were scaled iteratively, which built trust as value was unlocked. To cultivate knowledge exchange, we established forums to share successful use-cases across the portfolio. This exchange was critical to scaling the program as excitement grew, and other brands saw the opportunity of launching their own accelerator pods.?????

This approach proved successful. In twelve months, the central team launched accelerator pods across four brands, launching approximately 70 use-cases to deliver a forecasted $50 million in operating margin.?

The question is not whether to transform?

| “White. A blank page or canvas…So many possibilities.” – Stephen Sondheim?

Digital transformations can be either a cumbersome burden or a liberating force in an organization. When it’s a burden, it’s not because of the technology but rather the operating model. If CEOs adopt a modular approach that counters the risks inherent in centralization, they will unleash the ownership and creativity at the heart of all successful transformations.?????????

?

By Bharat Khandelwal, Shishir Pathak, Abhishek Dalmia, Kunal Bhatia, Ganga Kannan, Shilpa Sharma, Henok Eyob, and Alex Barocas?