You bought the tools. You have not changed the work.

In September Microsoft published the second edition of its playbook Becoming a Frontier Firm. It is not a sales brochure but a record of its own AI transformation, built on a review of more than a hundred internal cases. The most important finding is surprisingly non-technical: AI transformation cannot be run like a product launch or an IT project. It is business transformation, and it has to be guided by clear business goals set by leadership.

The playbook rests on five pillars: business ambition and goals, a system that spreads AI into everyday work, investment in people, codifying the company's own advantage and controls, and security. Each deserves its own blog. Here I focus on three that I see as the biggest gap in Slovenian companies.

The first is the direction of decisions. Microsoft cascades goals from the top down; it does not collect use cases from the bottom up. Every goal has a named owner in leadership, accountable for the result, not the activity. For every initiative they ask: if the only goal is lower cost, what bigger opportunity are we missing? Their summary is clear: efficiency is the floor; employee and customer experience, innovation and growth are the goal. In practice I see the opposite. Companies collect use cases, count active licences and hope a strategy will assemble itself. It will not.

The second is the way work is done. The playbook sums it up in a formula: CI + AI = CA. Continuous Improvement and AI together create Capability Add. Continuous improvement removes errors, delays, lost context at handoffs and unnecessary searching. AI adds what was not possible before: prediction, a wider range of tested solutions, real-time sensemaking. The first alone gives you a leaner version of today's company. The second alone gives you a faster bad process. Only together do they create something new. That is why Microsoft simplifies the workflow before adding agents. A rule I have repeated for years: simplify before you accelerate.

The third pillar is people, and the playbook itself calls it possibly the most important. One data point every CEO should know: according to the Work Trend Index (May 2026), manager role-modelling is the single biggest predictor of AI adoption in a team. Where managers actively use AI, recognised AI value is 17 points higher. And another: organisational interventions have twice the impact on AI effectiveness of work on individual mindset and skills. A leader who delegates AI use to the team does not move the organisation. A leader who shows it in their own work does.

This matches my experience from Microsoft, Zavarovalnica Triglav and my work with clients. The most common pattern I see: licences are bought, pilots are running, and the way work is done stays the same. Not because the technology does not work. Because nobody decided which business result must be different a year from now, and who is accountable for it.

Microsoft's approach was built in an organisation of more than a hundred thousand people, with its own transformation office in every function. A Slovenian company of two or three hundred people needs the same logic at a different scale: one small team, a clear sequence and measurement that tells leadership the truth. At Trnovec Consulting we have written that sequence into the AUXANO methodology. It starts with assessing leadership readiness, not with choosing a tool. Culture and leadership first, then goals, then redesigning the work.

The test I suggest to every leadership team is simple. Which business result will be different in twelve months because of AI, and who in leadership is accountable for it? If the answer does not come within five seconds, you do not have an AI strategy. You have a list of tools.

The tools are equally available to everyone. The way of working is not.

Source: Microsoft, Becoming a Frontier Firm, Our Frontier Playbook, 2nd edition (2026); Microsoft Work Trend Index (May 2026), as cited in the playbook.