The assessment that asks the right questions
95 out of 100 companies invest in AI and see none of it in their P&L. Not because the technology does not work, but because nobody analysed beforehand where the real leverage sits.
This page explains what most initiatives fail on, and why a situation analysis has to come from the outside. Scope, flat fee and process are on the assessment page.
The numbers are clear, and so is the cause
BCG surveyed 1,000 executives across 59 countries: three out of four companies fail to create scalable value from their AI investments. Across more than 300 initiatives studied, MIT arrived at a rate of 95 percent with no measurable effect on the P&L. Gartner expects that through 2026 around 60 percent of all AI projects will be abandoned for lack of a solid data foundation.
Not the models
MIT explicitly locates the cause outside model quality, in the missing anchoring within the organisation. Executives blame regulation or technology, the study shows something else.
People and processes
BCG puts a clean number on it: around 70 percent of the problems sit with people and processes, 20 percent with technology, only 10 percent with the algorithms. Yet those last 10 percent absorb a disproportionate share of the effort.
Data that does not hold
Without AI-ready data an initiative collapses regardless of how good the idea was. In 2025, 42 percent of companies abandoned most of their AI initiatives, up from 17 percent the year before.
Why an assessment has to come from outside
Your employees know best where things get stuck day to day. Those ideas are the starting point of every good assessment, but not its result.
Internal judgements almost always rest on yesterday's picture of what is possible. What an AI system can do today has shifted several times over the past twelve months. Anyone forming an idea based on what they saw a year ago will consistently aim too low. Or aim at a process that is not the expensive one.
Then there is the blind spot. Processes that grew over time are perceived from the inside as given, not as something you can design. That is exactly where the value sits. McKinsey tested 25 organisational attributes against actual bottom-line impact. By far the largest effect came from fundamentally redesigning workflows. Only 21 percent of users have fundamentally redesigned any workflows at all. Everyone else layers AI onto their existing processes.
A workshop puts the scenarios on the table, and that is a necessary phase. What tips the scale is the step after it: assessing the value behind each individual scenario. That requires a methodical process analysis and a view that does not already know the processes.
What an assessment delivers
- It takes your ideas seriously
- The demand is in the house. It is rarely articulated where the leverage is greatest.
- It challenges them consistently
- Every assumption is held against the current state of the technology. Some do not survive that.
- It goes deep
- The actual business case emerges once each scenario is costed out on its own merits.
How we proceed
Four steps that build on each other. The third is the one most people skip.
We record the starting position
Processes, data situation and the existing system landscape. And we listen to where management and employees each see the problem. Both views, because they are rarely the same one.
Every idea is held up to reality
Against what is actually possible today. Some assumptions do not survive that, and some ideas turn out considerably larger than expected. That is what we are here for.
We work the candidates through
Effort, data maturity, process depth, realistic benefit. This is where it is decided which use case carries and which one merely sounds good. This step turns a collection of scenarios into a basis for deciding.
You get a ranking, not a wish list
With the one use case you should start with, and the reasoning for why the others wait. Reasoned means: traceable for someone who was not in the room.
What you hold in your hands
Documented process analysis
Your workflows with the points identified where AI genuinely provides leverage.
An honest read on data maturity
Including the items that have to be solved first. Also when that is the less comfortable answer.
Prioritised use cases
With an estimate of effort and benefit, so that you can decide internally.
A roadmap for the first 90 days
Concrete enough to start on the Monday after. The result is yours, even if you do not continue with us.
The mid-market has a head start here
Large corporations run the most pilots and take nine months on average to move one of them into operation. Mid-sized companies do the same in around 90 days. Short paths, clear decisions, manageable structures: what often counted as a disadvantage in digitalisation is a genuine head start with AI.
Provided you start at the right point. That is exactly what the assessment is for.
On the sources
- MIT NANDA, «The GenAI Divide», 2025
- More than 300 public AI initiatives, 52 interviews, 153 survey responses. The 95 percent figure is contested, because the definition of success is narrow and the interview base small. The criticism targets the precision of the number, not its direction.
- BCG, «Where's the Value in AI?», 2024
- 1,000 executives from 59 countries and more than 20 sectors. Methodologically the broadest of the studies cited here.
- McKinsey, «The State of AI», March 2025
- Global survey. Of 25 organisational attributes tested, redesigning workflows has the largest effect on bottom-line impact.
- MIT Project NANDA: The GenAI Divide. State of AI in Business 2025 (PDF; the report is not published at a permanent MIT address, the link points to a mirror)
- Boston Consulting Group: Where's the Value in AI? (October 2024)
- McKinsey & Company: The State of AI. How organizations are rewiring to capture value (12 March 2025)
- Gartner: Lack of AI-Ready Data Puts AI Projects at Risk (February 2025)
- S&P Global Market Intelligence: AI Experiences Rapid Adoption but with Mixed Outcomes (2025)
Start at the right point
Every assessment begins with a free initial call of around 20 minutes. In it we clarify the task and whether an assessment is the right thing for you at all.
Book an initial call → See scope and flat fee →