I've sat through enough transformation post-mortems to know the standard deflection: The platform didn't integrate well. The vendors overcommitted. The team wasn't ready. All plausible. All wrong. The real failure happened weeks or months before a single line of code was written—and you didn't even notice.
The research is unsparing. 70% of digital transformation initiatives still fail to meet their objectives in 2026, despite years of effort and trillions spent. Failed efforts are estimated to cost organizations an astonishing $2.3 trillion per year. But here's what gets lost in those aggregate numbers: Gartner found that only 48% of digital initiatives meet or exceed their business outcome targets. The majority fail not because the technology didn't work but because nobody defined what success looked like before the programme started.
That's not a technology problem. It's a definition problem.
The Baseline Illusion
You can't measure what you didn't measure before. A transformation with no pre-programme measurement can never prove it worked—you will be arguing from anecdote. Yet that's exactly where most programs stumble: they adopt new technology, the team executes on milestones, the platform goes live—and then leadership asks, Did this actually work? Without a baseline, the answer becomes a Rorschach test. The CFO sees complexity. The COO sees opportunity cost. The CTO sees technical debt reduced. Everyone is right. No one knows.
Collecting baseline data helps you. Baseline data refers to the data you have at present. To know how to measure digital transformation, you want to know how the numbers currently look. It gives you a "before picture" so you can put together a "before-and-after" vision. Yet fewer than one in three programs capture this rigorously before starting.
The consequence isn't abstract. Only 30% of organizations can accurately measure the return on their digital transformation investments, even as these initiatives consume 12–15% of enterprise IT budgets. That gap between spending and measurement is where most transformations quietly fail.
The Metrics Trap: Activity vs. Outcome
Here's the insidious part: you can feel progress without having any. Activity metrics masquerade as outcomes. Counting the number of training sessions delivered or communications sent tells you nothing about whether people changed their behaviour. These metrics are easy to collect, which is precisely why teams default to them.
Your dashboard shows 95% of users trained. Your governance steering committee marks the implementation "on track." Your vendors bill their milestones. But have order-processing times actually dropped? Do your frontline people actually use the new system, or do they work around it? Is customer satisfaction up? None of this is on the dashboard.
Success Before Go-Live
The fix is structural, not aspirational. Define success criteria before you acquire the technology. Not platitudes—specifics.
Business and technology leaders should define value before investment, establish credible baselines, and remain jointly accountable for results throughout the transformation lifecycle. This means:
-
Capture the current state: Collect 3-6 months of historical data before setting targets. How long does the process take now? What's the defect rate? What's the time-to-fill for a critical role? How many escalations does a customer inquiry generate? Write it down.
-
Distinguish leading from lagging indicators: Leading indicators move first. They measure the behaviours and adoption that cause the outcome. If your target is faster order processing through a new tool, the lagging indicator is average processing time; the leading indicators are what percentage of orders actually go through the new tool, how many staff have been trained, and how often people fall back to the old workaround. Measure adoption during the program, not after it ends.
-
Assign owners and decision rights: Assign accountability for each benefit to specific business leaders, not just the project team. The finance director owns cost reduction. The operations director owns cycle time. They don't hand off accountability at go-live; they own the outcome.
-
Review early and often: Waiting until post-implementation to assess adoption means you have no opportunity to course-correct. By the time the data confirms a problem, the project team has moved on.
The Pattern
KPMG's 2026 Global Tech Report found that 74% of organizations report AI use cases creating business value, but only 24% achieve ROI across multiple use cases. The gap is not opportunity. It's accountability. The 24% defined what they were measuring before they started. The 74% got distracted by novelty.
Prosci's benchmarking data shows that 88% of projects with excellent change management met or exceeded their objectives, compared to just 13% with poor change management. That is a sevenfold difference in outcomes. That sevenfold gap isn't bought with better tools or bigger budgets. It's clarity about what success looks like, captured in numbers, owned by people with skin in the game.
Your next transformation won't fail because of the platform. It will fail because the CFO and the COO were never asked—or never answered—what they'd actually measure to know it worked. Spend the first six weeks of your program answering that question, not arguing about licensing models.
Everything else is noise.