A digital-transformation stage diagram can make the route look clear. An arrow reaching the next box does not mean an organization knows how to get there.
I think of a stage diagram I saw at a conference. Consulting and entrepreneurship have made me attentive to what happens after tools arrive: how work changes, who learns, and how concerns are addressed. Drawing a route does not answer these questions automatically.
1. Beyond the roadmap lies actual work
I do not want to understand transformation merely as buying a system or starting a project. If the aim is better work, we need to understand the existing process, the problem, and the change we hope to see.
That does not make technology irrelevant. Capabilities, compatibility, data quality, costs, and security conditions may determine whether an approach is usable. Organizational coordination and technical limits need consideration together.
Cases I have encountered can prompt questions, not estimate failure rates for all businesses. A project stopping, changing direction, and a company closing are different outcomes; one word, “failure,” should not collapse them.
2. Organizing experience also requires conditions for learning
Suppose someone familiar with operations must move an established way of working into a new system. This involves more than pressing buttons. Whether the interface is understandable, how information is organized, and how duplicate entry burdens are handled deserve examination.
This is a hypothetical implementation situation, not a reconstruction of a client’s history. It reminds me that unfamiliarity with a system does not prove refusal to improve. Familiarity with an older method does not make knowledge impossible to organize or pass on.
I value industry understanding, technical judgment, coordination, and continuing learning. These are my hiring preferences. They need neither a claim that hardly anyone has these abilities nor a general judgment that education stops learning. Nor must one person necessarily supply every capability.
3. Do not label reasonable concerns as obstruction first
Whose data is involved, whether a system is secure, whether architecture fits, and how performance will be assessed all require concrete answers. Questions from several departments do not by themselves prove a joint defense of power.
Changes in responsibilities and authority may create conflict; this is one interpretation that concerns me. Judging a particular case still requires actual conduct and decision records. A statement of “security concerns” alone does not establish its speaker’s motives.
Conversely, repeating that risk is high does not resolve it. Proponents should offer a proposal, while people raising concerns should help explain which risk is involved, under what conditions it arises, and what evidence would permit a decision.
4. Data access needs scope and responsibility
I would not make willingness to hand over every record a test of commitment to transformation. Asking what information a task needs, who may use it, for what purpose, and what should not be handed over is more concrete than demanding universal transparency.
Likewise, easier access does not automatically make judgments fair or decisions traceable. What gets recorded, who can check it, and who handles errors all need arrangements. A new tool cannot make those agreements for an organization.
The trade-offs in thinking about collaboration through tools apply here too. Cooperation need not abandon judgment, and retaining necessary capabilities need not reject cooperation. Scope and continuing responsibility need explanation.
5. Turn concerns into decisions that can be made
I want to organize abstract support or opposition into four questions:
- Which part of the work needs improvement, and what would count as helpful?
- What risks and limits are involved, and what evidence is needed to judge them?
- Who will try, verify, and support the approach, and how will resources be arranged?
- When will results be reviewed, and on what conditions should we continue, adjust, or stop?
This is the author’s proposed way to discuss projects, not a transformation-success model or a guarantee that every concern can be removed or every project should continue.
If compatibility needs checking, for example, specify which part to test and who will confirm it before discussing whether the investment is worthwhile. This is a possible approach, not a process every company already has. If evidence indicates unsuitable costs or risks, stopping can also be a reasoned decision.
6. AI adoption still needs concrete questions
With AI, we can still ask what task it should assist, whether inputs are suitable, who judges results, and how errors are handled. I consider these worthwhile questions, not evidence that every AI tool fits the same arrangements.
I no longer encourage change by saying technical problems can always be solved, or explain difficulty by saying people must be the biggest obstacle. Choosing to pursue change means attending to conditions between technology, work, and people, rather than labeling someone backward first.
Keep the roadmap. The next task is to explain the work between its arrows, allowing support and concerns to be examined and making someone responsible for decisions.
Materials and limits (revised through November 4, 2026)
The conference diagram and consulting and entrepreneurship background retain the original account. Slides, client records, and company outcomes were not recovered and cannot establish a failure rate. Operations and compatibility passages are explicitly hypothetical or possible approaches, not added client history. The four questions, data scope, and hiring preferences are author proposals, not verified systems, technical or security findings, legal guidance, or success-rate claims.
Revision note: Case counts and completion rates, unchecked ERP details and dialogue, generalizations about education and talent, definitive power-sabotage claims, automatic AI transparency, and five-year elimination predictions were removed. Transformation remains more than buying tools, while technical, security, and data concerns are restored to consideration.


💬 Comments
Loading...