Why one digital transformation in two fails, and how to avoid it
In 2026, digital transformation is no longer optional for French small businesses: according to the France Num barometer (from the French government’s small-business digitization program), 82 percent of them are investing in digital this year, with an average budget close to €75,000, up sharply from 2024. Artificial intelligence alone captures about 42 percent of these budgets, ahead of cybersecurity and process automation. Yet behind this acceleration, a stubborn reality remains: a significant share of these projects never achieve the expected results, for lack of method.
The most common trap is confusing digital transformation with buying tools. A small business signs up for a CRM, rolls out an ERP or launches an AI project before diagnosing its processes, mapping its data and preparing its teams. The result: underused software, data still in silos, employees working around the new tools, and an owner who concludes that digital costs a lot without paying off. In fact, the number-one obstacle cited by 72 percent of French small businesses isn’t money, but the lack of in-house skills to lead and absorb the change.
This article details the 6-step roadmap I apply in my fractional CIO engagements to make a small or mid-sized company’s digital transformation succeed. Each step is documented with its objectives, method, deliverables, duration and pitfalls. This method is part of our series on digital transformation for small businesses. For the upfront technical inventory, rely on our guide to IT audits for small businesses, and for strategic oversight, on our guide to the fractional CIO / CIO as a Service.
The 6 steps of a digital transformation roadmap
The roadmap is organized into 6 steps that run partly in sequence and partly in parallel. Steps 5 (change management) and 6 (management) run continuously, from the first project delivered. Here’s the overview with the typical duration and key deliverable of each.
| Step | Objective | Duration | Key deliverable |
|---|---|---|---|
| 1. Maturity assessment and vision | Assess the current state and set the direction | 3-5 weeks | Maturity radar + target vision |
| 2. Scoping and prioritization | Build the project portfolio | 3-4 weeks | Roadmap + budget |
| 3. Pilot project | Prove value on a limited scope | 6-10 weeks | Pilot in production + measured ROI |
| 4. Technical foundation, data, security | Deploy the structural building blocks | 4-9 months | Target IT systems + data + cybersecurity |
| 5. Change management | Bring teams on board and train them | Ongoing | Adoption plan + training |
| 6. Management and continuous improvement | Measure ROI and iterate | Ongoing from the 1st project | Dashboard + reviews |
The guiding principle for 2026: stop aiming for one big 3-year transformation plan, and instead run a series of 2- to 6-month projects that each produce measurable value. This wave-based approach reduces risk, partly funds what comes next through the gains achieved, and keeps teams on board through regular wins.
Assess digital maturity and set the vision
Typical duration: 3 to 5 weeks. Lead: fractional CIO or transformation consultant. People involved: owner, department heads, IT, finance.
Objective
Establish an objective picture of the company’s digital situation and define a clear target vision for 18-24 months. Without this foundation, any roadmap is built on assumptions and leads to tool choices disconnected from real needs. The assessment answers a simple question: where do we really stand, and where do we want to go?
Method
Conduct 6 to 12 interviews with business leads and observe the field to map key processes, daily pain points and time-consuming tasks. In parallel, inventory the existing IT systems (tools, licenses, level of integration), assess data quality and accessibility, measure the cybersecurity level and gauge the teams’ digital skills. Then place the company on a 4-level maturity scale, along the lines of the France Num barometer, which distinguishes mature, dynamic, high-potential and reluctant companies.
Build a maturity radar across 5 to 6 axes: process digitization, data use, customer experience, cybersecurity, digital culture and skills, infrastructure. The radar visualizes the gaps between the current situation and the target, and brings natural priorities to the surface. Finally, formalize with the owner a target vision aligned with the company’s strategy and measurable objectives (for example: cut order processing time by 30 percent, centralize 100 percent of customer data, reach a compliant security level).
Deliverables
- Digital maturity assessment report
- Maturity radar across 5-6 axes with gaps to target
- Map of processes and priority pain points
- Inventory of IT systems, data and security level
- 18-24-month target vision approved by the owner
- Measurable objectives aligned with company strategy
Pitfalls to avoid
Don’t skip the assessment to go faster: it’s the founding mistake that dooms 50 percent of transformations. Don’t leave the inventory solely to a software vendor, who will naturally steer you toward its own solution. Don’t limit yourself to technology: digital culture and skills weigh as much as infrastructure. Don’t set vague objectives (“be more modern”): without a measurable indicator, you’ll never be able to prove ROI. In France, consider using the free France Num assessment as a structuring starting point.
Scope the roadmap and prioritize projects
Typical duration: 3 to 4 weeks. Lead: fractional CIO with the leadership team. People involved: owner, business leads, finance, IT.
Objective
Turn the vision into a portfolio of concrete projects, sequenced over time and budgeted. The roadmap must answer three questions: what do we do, in what order, and with what budget? The challenge is to avoid both scattering (launching everything at once) and paralysis (a plan so ambitious it never gets started).
Method
List the candidate projects that came out of the assessment: automating administrative tasks, centralizing customer data in a CRM, modernizing the ERP or moving to the cloud, redesigning the website and customer journey, targeted AI projects, upgrading cybersecurity. Score each project on five criteria: measurable business value, technical feasibility, cost, risk level, and dependencies on other projects (for example, making data reliable before rolling out an AI project).
Sequence the roadmap in 2- to 6-month waves, putting 1 or 2 high-return, low-risk quick wins first to build momentum and partly fund what follows. Estimate the overall and per-project budget, including not only licenses but also integration, data migration, training and change management (often 25 to 40 percent of the total cost). Identify the available funding (in France: the France Num assessment and programs, regional grants, the training tax credit). To scope the budget for an ERP or CRM project, rely on our analysis of the cost of a cloud ERP migration.
Deliverables
- Portfolio of scored and prioritized projects
- Roadmap sequenced in 2- to 6-month waves
- Business case per project (investment, gain, ROI)
- Consolidated multi-year budget with identified funding
- Project governance (sponsor, committee, management cadence)
- Formal approval by the leadership team
Pitfalls to avoid
Don’t launch everything in parallel: a small business rarely absorbs more than 2 to 3 projects at once. Don’t neglect dependencies: rolling out an AI project on siloed data is doomed to fail. Don’t underestimate the human cost: change management is the most overlooked and most decisive budget line. Don’t build a fixed 3-year plan: the roadmap must be revised with each wave based on results and emerging opportunities, AI in particular.
Launch a pilot project to prove the value
Typical duration: 6 to 10 weeks. Lead: internal transformation lead. People involved: pilot business team, IT, integrator, executive sponsor.
Objective
Start the transformation with a quick, visible win. The pilot project takes the first high-impact, low-risk project, rolls it out on a limited scope, and proves the value concretely before any broader rollout. It’s the step that turns theoretical buy-in into real confidence, among the teams as well as the owner.
Method
Choose a contained scope: one team, one process, one branch or one product line. Define success indicators before starting (time saved, fewer errors, user satisfaction, savings achieved). Deliver a usable first version quickly rather than a perfect solution, then iterate with real users. Measure the actual gain over 4 to 8 weeks of use and compare it with the previous situation. A good first pilot in 2026 is often automating a time-consuming administrative process or centralizing customer data; on these topics, see our guide to no-code automation and our 2026 small-business CRM comparison.
The pilot also serves to break in the project method: meeting cadence, user involvement, collaboration with the integrator, data management. The lessons learned here sharply reduce the risks on the heavier projects that follow.
Deliverables
- Pilot solution in production on a limited scope
- Success indicators defined and measured
- Measurement of real ROI compared with the previous situation
- Structured user feedback
- Reasoned decision to roll out or adjust
Pitfalls to avoid
Don’t choose an overly ambitious or risky pilot: a first visible failure kills momentum for months. Don’t aim for perfection: a usable, imperfect pilot beats a complete solution that never ships. Don’t forget to define indicators before starting, or the gain will stay invisible and impossible to sell internally. Don’t do a big-bang rollout across the whole company after the pilot: expand in controlled waves.
Build the technical foundation, data and security
Typical duration: 4 to 9 months depending on scope. Lead: transformation lead and IT. People involved: integrators, vendors, CISO, business teams concerned.
Objective
Deploy the structural building blocks that support the transformation over the long term: infrastructure and cloud, core business tools, data integration, and cross-cutting security. It’s the most investment-heavy step, and the one that determines the company’s ability to make use of AI and analytics afterwards. A poorly laid foundation limits everything that follows.
Method
Deploy in waves, never all at once. First, modernize the infrastructure: move to the cloud for flexibility and resilience, gradually retire aging servers, unify collaboration tools. Second, roll out core tools according to priorities: ERP for integrated management, CRM for customer relationships, industry-specific solutions. To decide between cloud and on-premises, see our guide to cloud ERP migration.
Third, treat data as an asset: integrate systems to break down silos, make data reliable and clean it up, set up flows between tools. Without this layer, the company accumulates tools that don’t talk to each other. Fourth, secure across the board: tested backups, multi-factor authentication, access management, GDPR compliance and, for the companies concerned, compliance with the EU’s NIS2 directive. On this point, see our 2026 guide to NIS2 and small-business cybersecurity. Always favor interoperable solutions and limit vendor lock-in to preserve the company’s freedom.
Deliverables
- Infrastructure modernized and migrated to the cloud
- Core tools deployed (ERP, CRM, collaboration)
- Integrated, reliable data and automated flows
- Cybersecurity foundation (backups, MFA, compliance)
- Technical documentation and operating procedures
- Interoperable architecture limiting vendor lock-in
Pitfalls to avoid
Don’t pile up tools that don’t communicate: data integration matters as much as the tools themselves. Don’t postpone cybersecurity: a digital transformation increases the attack surface, so security must be built in from the start. Don’t sign contracts with heavy lock-in: require reversibility and data export. Don’t neglect the quality of existing data: migrating bad data only moves the problem.
Lead the change and build up the teams’ skills
Typical duration: ongoing, alongside every step. Lead: executive sponsor with HR and the transformation lead. People involved: all employees, business ambassadors.
Objective
Ensure that the tools deployed are actually adopted and create value. It’s the most decisive step: technology produces no return if teams don’t use it or work around it. With 72 percent of French small businesses citing a lack of skills as their top obstacle, change management isn’t an extra; it’s the core of success.
Method
Treat adoption as a project in its own right, from the start and continuously. First, a visible executive sponsor who embodies the transformation and gives it meaning: without commitment from the top, teams don’t follow. Second, identify ambassadors in each business area, local go-betweens more credible than a memo from management. Third, communicate continuously about the why and the concrete benefits for each person, answering fears honestly (fear of being replaced, overload, loss of bearings).
Fourth, train by profile and by real use, beyond the one-hour demo: hands-on sessions, accessible support, gradual skill-building, and, in France, tapping the training tax credit and OPCO training funds. Fifth, measure real adoption (tool usage rate, satisfaction, reported pain points) and correct as you go. Digital culture is built over time, not in a single seminar. To structure AI skill-building specifically, see our AI guide for small businesses: where to start.
Deliverables
- Formal change management plan
- Identified executive sponsor and network of ambassadors
- Long-term internal communication plan
- Training paths by profile with funding secured
- Adoption dashboard (usage rate, satisfaction)
- System for reporting and handling pain points
Pitfalls to avoid
Don’t reduce change management to a single training session at the end of the project: it starts at step 1. Don’t ignore the teams’ fears: denying them fuels passive resistance. Don’t neglect the owner’s role: an absent sponsor dooms the project. Don’t measure deployment (“the tool is installed”) but adoption (“the tool is used and liked”). Don’t forget to promote and fund training, a mechanism small businesses often underuse.
Manage, measure ROI and improve continuously
Typical duration: ongoing, from the first project delivered. Lead: fractional CIO or digital steering committee. People involved: owner, business leads, finance, IT.
Objective
Turn the roadmap into a continuous improvement cycle governed by value. Digital transformation has no end date: it becomes a permanent capability of the company to evolve. Management ensures that each project delivers what was planned, and that priorities adjust to results and new opportunities, foremost among them AI in 2026.
Method
Set up a management dashboard combining business indicators (revenue, productivity, customer satisfaction, lead times) and digital indicators (adoption rate, availability, incidents, costs). Convene a digital steering committee on a regular cadence (monthly, then quarterly) to track progress, measure consolidated ROI and decide what comes next. Systematically compare the real gain with the initial business case to maintain the approach’s credibility. Revise the roadmap with each wave based on the results obtained and emerging opportunities, taking into account the rapid progress of AI and automation. To run this function sustainably without hiring a full-time IT director, bringing in a fractional CIO / CIO as a Service offers the best cost-to-expertise ratio for most small businesses.
Deliverables
- Management dashboard (business and digital indicators)
- Digital steering committee with a set cadence
- Consolidated ROI measured per project and overall
- Periodic reviews and documented decisions
- Roadmap updated with each wave
- Watch on emerging opportunities (AI, automation)
Pitfalls to avoid
Don’t consider the transformation finished once the tools are deployed: without continuous management, gains erode and tools become outdated. Don’t manage without measuring ROI: that’s what justifies the next investments to the owner. Don’t freeze the roadmap: adapt it to results and opportunities. Don’t multiply indicators: 8 to 12 well-chosen KPIs beat an unreadable dashboard.
The 5 key success factors of a small-business digital transformation
Beyond the 6-step method, five principles separate the transformations that succeed from those that get bogged down. They have less to do with technology than with how it’s managed.
- A committed executive sponsor. Transformation is a company project, not an IT project. Without visible backing from the owner, buy-in collapses.
- Step-by-step projects with fast proof of value. A series of 2- to 6-month waves beats a big, static 3-year plan.
- Data and integration before tools. Breaking down silos determines the value of everything that follows, AI included.
- A real investment in people. Training and change management represent 25 to 40 percent of the cost and 80 percent of the result.
- Management by ROI. Measuring each gain sustains confidence and funds the rest of the transformation.
This method isn’t reserved for large organizations. A very small business can apply a condensed version to 1 or 2 projects; a mid-sized company deploys it across a broader portfolio. What never changes: start with the assessment, prove value with a pilot, and treat human adoption as the number-one priority.
Frequently asked questions about the digital transformation roadmap
How long does a digital transformation take in a small business?
A structured 6-step roadmap typically spans 12 to 24 months to reach a solid foundation, but the first results come well before that. The assessment takes 3 to 5 weeks, scoping 3 to 4 weeks, the pilot project 6 to 10 weeks for a first proof of value, and the technical foundation 4 to 9 months depending on scope. Change management and oversight run continuously. The right approach in 2026 isn’t a single 3-year plan but a series of 2- to 6-month projects that each produce measurable value.
Where should a small business actually start?
Always with an honest maturity assessment, never with buying a tool. The right sequence: assess the current state, define a vision and measurable objectives, prioritize a first high-return, low-risk project, run it as a pilot to prove the value, then scale. In 2026, the most profitable first projects are automating repetitive administrative tasks, centralizing customer data and basic security, before any ambitious AI project. For French companies, the free France Num assessment is an excellent starting point.
What budget should a small business plan for digital transformation in 2026?
According to France Num, 82 percent of French small businesses are investing in digital in 2026, with an average budget of around €75,000. Reality varies: a very small business gets started with €8,000 to €25,000 on a first project, an established small business invests €40,000 to €120,000 a year across several projects, and a mid-sized company often exceeds €200,000 a year. The most underestimated budget line isn’t licenses but change management and training, which account for 25 to 40 percent of the real cost. In France, programs such as the France Num assessment, regional grants and the training tax credit reduce the bill.
What is the main reason digital transformations fail?
It’s almost never the technology; it’s human adoption. In 2026, 72 percent of French small businesses cite a lack of in-house skills as their top obstacle. Recurring causes of failure: a project run like an IT purchase rather than a company project, no executive sponsor, vague, unmeasurable objectives, training reduced to a demo, and a big-bang rollout. The remedy: a committed sponsor, step-by-step projects with fast proof of value, and a real investment in training and supporting the teams.
Do you need an IT director or an outside consultant to lead the transformation?
Most small businesses with 20 to 250 employees don’t have a full-time IT director and don’t need one permanently: they need part-time strategic leadership. An outside fractional CIO brings method, objectivity toward vendors and experience with comparable transformations, for €2,000 to €8,000 a month (French market rates) — far less than a hire. They work mainly on the assessment, scoping and structural choices, then on oversight, while execution is handed to internal teams and integrators. This outside leadership avoids ill-suited tool choices whose extra cost far exceeds the consulting fees.