You have probably heard that 70 percent of digital transformations fail. It is one of the most quoted numbers in consulting, and it is usually wrong in a useful way.
The figure most people mean comes from a 2020 Boston Consulting Group study, which found that 70 percent of digital transformations fall short of their objectives. That is not the same as failing. In BCG's data, about 30 percent fully met their targets, roughly 44 percent created some value but missed the goal, and around 26 percent produced little or nothing. McKinsey's 2018 survey set a stricter bar and found that only 16 percent of respondents said their transformation had both improved performance and kept it improved.
Either way, the odds are not great. The more interesting part of the research is what the successful 30 percent did differently, because none of it is about choosing the right software.
What the research says separates winners
BCG identified six factors. Companies that got all six right raised their success rate from 30 percent to 80 percent.
- An integrated strategy with clear goals. The program is tied to specific business outcomes, not a list of systems to install.
- Leadership commitment from the CEO through middle management. Not just a kickoff speech. Managers who run the day-to-day work have to own the change.
- High-calibre people on the program. Companies put their best operators on it, not whoever happens to be free.
- Agile governance. Decisions are made quickly, scope is adjusted as the team learns, and problems surface early.
- Effective monitoring of outcomes. Progress is measured against business results, not milestones.
- A business-led, modular technology and data platform. Systems are chosen and phased to serve the business priorities, and can be changed piece by piece.
Read that list again and notice how little of it is technology. Five of the six factors are about goals, people and decision-making.
The failure patterns we see most often
In mid-sized companies, the kind with somewhere between 50 and 2,000 people, the research plays out in a few recognisable ways.
"Digital" as the goal. A program called "go paperless" or "move to the cloud" has no finish line and no way to know if it worked. Compare that with "cut order-to-invoice time from six days to one". The second one tells everyone what to build and when to stop.
Buying the platform first. A company signs a large ERP or CRM contract, then tries to work out what to do with it. The software ends up configured around old habits, with spreadsheets running alongside it within a year.
The big-bang go-live. Everything switches over on one weekend. When something breaks, and something always breaks, the business loses confidence in the whole program.
No owner in the business. IT runs the project, the sales or operations head treats it as IT's problem, and adoption quietly dies after training week.
Ignoring the people doing the work. The warehouse supervisor or the accounts clerk knows exactly where the current process breaks. If they are not in the design sessions, the new system will break in the same places.
A practical 90-day start
Here is the approach we use when a company asks us to help it begin, or restart, a transformation.
Weeks 1 to 3: pick the target. Map the two or three processes that cost the most in money, time or cash flow. Typical candidates are order-to-cash, procurement and inventory, field service, or customer onboarding. Agree on three to five metrics and record today's baseline.
Weeks 4 to 6: design the change with the people who do the work. Sit with the teams. Decide what the new process looks like before choosing or configuring software. This is where most of the value is found, often before any code is written.
Weeks 7 to 12: ship a first phase. Put one process live on connected systems, with a small group of users, and measure it against the baseline. A common first phase for our clients is quoting and invoicing in an ERP such as Odoo, with WhatsApp notifications for customers and a live dashboard for management.
At the end of 90 days, leadership has a real result to judge, and the next phase is planned on evidence rather than enthusiasm.
What to measure
Good transformation metrics are ones a CFO or operations head already cares about:
- Days sales outstanding, or how fast you get paid
- Order cycle time, from enquiry to delivery
- Inventory accuracy and inventory turns
- Hours of manual data entry per week
- First-response time for customers
- Error or rework rates in a key process
Avoid metrics like "number of users onboarded" or "modules live". They tell you activity happened, not that the business improved.
Where AI fits
Every transformation conversation in 2026 includes AI, and it should. But AI agents need clean data and systems they can act in. An agent that chases overdue invoices needs an accounting system with accurate balances and an API it can use. An agent that answers stock queries needs inventory data someone trusts.
That is why we usually sequence it this way: connect the core systems first, then add AI agents to specific workflows once the data underneath is reliable. Companies that skip the first step often end up with impressive demos that cannot do real work. We cover that problem in more depth in our guide on taking AI from pilot to production.
A quick self-check
If you are planning a transformation, or are partway through one, ask:
- Can we state the goal as a business number with a date?
- Does a business leader, not IT, own the outcome?
- Are our best people working on it, at least part-time?
- Will something go live, for real users, within 90 days?
- Do we review results against the baseline every month?
If any answer is no, fix that before spending more on software.
How Appmetry helps
Our digital transformation work is built around those phases. One team handles process design, ERP and CRM, custom apps, WhatsApp and AI automation, so nobody has to coordinate four vendors. For larger organisations with multiple business units, our enterprise transformation programs add portfolio planning and change management across the group.
If you want a second opinion on a program that has stalled, or a plan for one that has not started, talk to us. We will tell you where we would begin.
Sources
- Flipping the Odds of Digital Transformation Success, Boston Consulting Group, 2020: the 70 percent "fall short" finding, survey of 825 senior executives, the six success factors and the 30 to 80 percent improvement.
- Unlocking success in digital transformations, McKinsey Global Survey, October 2018: 16 percent of respondents reporting improved and sustained performance.
- Do 70% of Digital Transformations Fail? Tracing the Number, Reliable Media, 2026: breakdown of BCG's 30, 44 and 26 percent outcome groups and the history of the 70 percent claim.
- Study of 900 digital transformations: only 30% are successful, Consulting.us: independent coverage of the BCG findings.