Why digital transformations fail before they even start
Digital transformation initiatives rarely fail for lack of vision. In 2026, most leadership teams know where they want to go: a unified cloud platform, generative AI embedded in business processes, a product organization rather than a project one. What derails these ambitions is almost always the scale of the first step, a program that's too broad, too many dependencies activated at once, a starting point chosen by gut feeling rather than method.
Launching a transformation that builds momentum, demonstrates tangible value, and absorbs the inevitable setbacks without putting the entire business at risk means letting go of two natural instincts: the big, all-encompassing plan that looks reassuring on paper, and picking the most visible or most demanded project internally. The right approach is to methodically select an "exemplar", a pilot initiative that acts as a scout for the whole transformation journey.
An exemplar is not simply a technical pilot. It's a strategic probe: a focused initiative with a defined scope, designed to test new ways of working, new technology building blocks, and new team structures under real conditions. It serves as a source of learning ahead of a large-scale rollout, and it surfaces unknowns, funding, stakeholder buy-in, organizational debt, before they become blockers on an entire program.
Our experience on the ground confirms this: the success of an exemplar-led transformation depends directly on the rigor of the selection process. You need to move beyond guesswork and pet projects to identify the initiative with the best balance of business impact, feasibility, and organizational relevance. Choosing the wrong exemplar is never neutral: an early failure on too ambitious a scope feeds skepticism, a topic too niche speaks to no one outside the immediate team, and a project too trivial never convinces anyone to go further.
From gut feeling to methodical selection: the discovery phase
Simply asking "what are your top project ideas?" almost always produces the same list: pet projects, ideas that sound appealing on paper, and loud complaints from the people most present in meetings, what's known as recency bias. A more robust approach involves running a structured exploration phase, which we call "discovery", to map the entire business and technology landscape before narrowing the field of possibilities.
Mapping the customer journey
By tracing customer interactions across every touchpoint, from the first digital contact to post-sale support, you surface friction points, inefficiencies, and the highest-value opportunities for digital intervention. Cross-functional workshops are decisive here: they let you map the stages of the journey across different business units, reveal hidden dependencies, and expose areas where the scope of several initiatives overlaps, which limits duplicated effort across the organization.
Modeling the business domain
Domain modeling means building a shared understanding and representation of the business domains, the core products, services, and capabilities the organization offers. This exercise almost always reveals a surprising lack of common language between departments, but it's essential for identifying product boundaries and their mutual dependencies. It also helps pinpoint strategically critical products and the internal platforms that support them, an exercise that sits at the heart of product cartography approaches.
These two techniques are just a starting point; other frameworks, value stream mapping, Wardley mapping, event storming, can be substituted depending on the organization's maturity. What matters is arriving, after a few sessions with the right stakeholder groups, at a deliberately broad overview: at this stage, you need to capture widely, from customer-facing improvements to internal process optimizations, without narrowing the scope too early. Every domain identified still needs to be explicitly tied to business objectives and strategic goals, or the selection that follows will lose its objectivity.
The three selection lenses: impact, feasibility, relevance
Once the inventory of candidates has stabilized, the real selection work begins. Each candidate is run through a set of objective criteria, not gut feeling, but disciplined evaluation. We build on the classic impact-versus-effort prioritization approach and add a third lens that's essential for an exemplar: relevance.
Business impact: does it actually matter?
This first lens is used to narrow the long list down to a prioritized shortlist. It means identifying the relevant business metrics for each domain and ranking them, using absolute scoring (A is better than B) or relative scoring (A generates more revenue, B protects more market share). The goal is to isolate three to five domains with the potential to "move the needle", quantifying impact wherever possible, expected revenue increase, cost savings, customer retention. A good candidate has a measurable impact within three to four months, without being so business-critical that an experimental failure would be devastating.
Feasibility: can it be delivered within a reasonable timeframe?
This lens forces an honest conversation across three dimensions. On the product side: is the problem well defined, are the vision and scope clear, is the link to business objectives explicit? On the technology side: is the stack ready for rapid decoupling, or does it need major rework first? Can the team operate with minimal dependencies on slower-moving teams? On the organization side: can you secure a dedicated product owner and tech lead, free up the right skills without major disruption, avoid a scattered start with staggered arrivals that keep resetting team dynamics to zero? Scoring here is relative, domain by domain, to stay honest about red flags: organizational obstacles and technical maturity are, in practice, the most frequent blockers.
Relevance: is there enough interest?
This third dimension, specific to exemplars, is decisive for change management. It assesses whether the chosen domain can serve as a compelling showcase, capable of inspiring other teams to adopt the new ways of working. Domains too obscure for most people to understand should be avoided, highly specialized back-office systems, advanced analytics, internal platforms or APIs whose impact stays invisible outside the technical team. A good candidate creates an "I want that too" effect: colleagues on other teams understand the problem, recognize themselves in it, and can translate the changes to their own context.
Making the call: from evaluation to final decision
More often than not, a good number of candidates can be ruled out quickly, without much research. The remaining candidates are ranked, and experience shows that three to four options with a clear recommendation is the optimal format for a focused decision discussion. These options will almost always meet the necessary criteria: there's no point chasing the perfect selection at the cost of analysis paralysis.
The decision itself shouldn't rest on gut feeling or personal preference, but on this rigorous evaluation across the three lenses. That doesn't rule out group discussion and leadership judgment, which remain necessary to arbitrate between close options and land on the best possible selection.
Once the decision is made, it needs strong leadership alignment and sustained commitment to stay the course. It's worth remembering that an exemplar is still an experiment: it can run into difficulties, or even fail on certain aspects. When things don't go as planned, the leadership team's role is to focus on learning and adjusting, not on questioning the transformation itself. This is often where the credibility of the whole program is quietly decided.
After the choice: the mobilization phase
Choosing the exemplar is a critical step, but it's only the starting point. The next step, which we call "mobilization", covers two stages: ramp-up, then kick-off, to prepare the team for its first delivery sprint.
Ramp-up: structuring the target team
This phase brings together a set of key activities to shape the future team: appointing critical roles (product owner, tech lead, relevant subject matter experts), defining the roadmap and target architecture, designing the team structure, naming the actual people who will make it up, and building a realistic transition plan. The feasibility assessment conducted during selection forms the foundation for this next level of detail, it's also the moment to secure funding beyond the initial project budget alone.
Kick-off: launching the team on solid ground
Once the full team is named and ready to start, an intense week is aimed at launching it: aligning everyone on the vision, roadmap, and goals, agreeing on ways of working, and starting to bond as a team. These are essential steps for building a genuinely high-performing unit, rather than a simple assembly of individual skills.
Once the team is up and running, a coaching and support phase ensures it gets the guidance needed to overcome early obstacles and embed new practices. During the exemplar's execution, leadership also closely tracks the difficulties encountered and any learnings relevant to the next wave of transformation. This support can be scaled back, then stopped, once the team has stabilized enough for its way of working to become the new normal, which marks the official end of the exemplar phase.
Choose your scout wisely
A digital transformation is a major undertaking. Its success doesn't start with a big, all-encompassing plan, but with a single, well-chosen step. Investing time and rigor into a structured discovery and selection process, leveraging tools like customer journey mapping and applying the three lenses of impact, feasibility, and relevance, helps ensure that first step is the right one.
It's not just about picking a project; it's about choosing the scout that will successfully guide the entire organization through the complexities of transformation, building the trust, capability, and momentum needed to reach the intended strategic destination. In our experience, organizations that treat exemplar selection with as much rigor as their final business case are the ones that turn an early local win into sustained momentum, rather than an isolated anecdote that's soon forgotten.
Disclaimer: The statements and opinions expressed in this article are those of the author(s) and do not necessarily reflect Adservio's positions.
STAY POSTED
Get our next analyses and field notes straight to your inbox.
