In many companies, several employees are already using artificial intelligence in their work. However, very few management teams actually make the decision to structure AI usage. The result: AI is used throughout the organization, but rarely in a coherent manner. If you haven’t yet read How to Structure AI Usage to Strengthen Your Team, we strongly recommend you do so before going any further.
In the majority of SMEs, team leaders and managers initiate the first steps. On your end, you probably started by observing how your colleagues were using AI individually. You identified certain practices that work better than others, set a few rules that you consider important, and tried to channel the use of tools available in your Microsoft 365 environment, such as Copilot. You also identified a few useful use cases for your team and tried to measure what really works.
But despite these efforts, it is becoming difficult to go further alone.
The IT department does not always have the resources to configure the necessary security policies, structure data access, or create clear benchmarks for AI use. Procedures remain variable from one person to another, results are not always comparable, and certain practices must be redone or adjusted regularly. Meanwhile, management does not always see why AI deserves more consideration, even though the lack of an AI framework creates very different practices from one team to another.
This is generally when a manager realizes that experimenting is no longer enough.
To go further, you must be able to explain the situation clearly, demonstrate the value of investing time—and sometimes a certain budget—to structure AI usage within the company. And to achieve this, you need a clear approach to present to management.
This is exactly the role of a “business case template“.
To help you organize this process, we have prepared a business case template that allows you to present the current situation, the associated risks, and a realistic pilot project to propose to management. In this article, we will use it in the context of AI, but this same template can also be used for other challenges managers face with Microsoft 365 and digital tools.
👉 Consult the business case template
If you haven’t already, consult our article on structuring AI in teams, which explains how to set boundaries, choose a common tool, train on best practices, and measure adoption.
Your Management Probably Underestimates the Growing Gap
From management’s perspective, AI is already “accessible.” Colleagues are using it, results are coming in, and no one is complaining. Why invest in something that seems to work on its own?
What management fails to measure is the difference between scattered use and adoption that creates value. According to the McKinsey State of AI 2025 report, 92% of companies plan to increase their AI investments over the next three years, but barely 1% consider their deployment to be mature. The gap between “we use AI” and “we derive measurable value from it” is enormous, and that is exactly where your team stands.
The Microsoft Work Trend Index 2025, based on a survey of 31,000 workers in 31 countries, reveals that 90% of advanced AI users find their workload more manageable thanks to the tool. However, the same report shows that 75% of workers are now using AI, often without oversight from their organization. This is a signal that management needs to hear: your colleagues are not waiting for permission. The question is not “should we regulate AI,” but “should we structure its use before an incident forces us to do so.”
For management to act, you must show them the cost of not structuring anything and the return on doing it correctly.
The Five Questions Your Business Case Must Cover to Convince Management Before Investing to Structure AI Usage
An argument that works with management answers five questions. No tech jargon. Just facts, a plan, and a reasonable request.
1. What have you observed in your team? Name concrete behaviors. “We use too many different tools” is not enough. “Four out of six colleagues use a different AI tool. The accounting technician copies financial data into ChatGPT. The representative uses an Outlook add-in that the IT department never approved. No one validates results the same way.” This is the kind of picture management can process. The results of our self-diagnostic can help support this finding if you have already completed it.
2. What is the current cost? Translate risks into concrete impacts. Every free tool used by a colleague is a channel through which company data leaves the controlled environment. Every unvalidated result that ends up in a report or a client email is a risk to the team’s credibility. And the time lost by everyone reinventing their own prompts and repeating the same trial-and-error instead of building on what already works is productivity evaporating every week. McKinsey observes that organizations most advanced in integrating artificial intelligence invest five times more in structuring AI usage and scaling than in the technology itself. The tool is not the problem. It is the lack of structure around the tool that is costly.
3. What is the proposed solution? The goal is to move from individual, unmanaged use to structured integration with a common tool, clear rules, and training adapted to each role. The tools are already available in Microsoft 365: Copilot integrated into Excel, Outlook, Teams, and Word, with data remaining within the company environment. SharePoint to centralize procedures and usage rules. The Microsoft 365 admin center to structure AI usage by person and by application. It is an AI framework using the tools you are already paying for.
4. What is the recommended pilot project? Start with a single team and a limited number of scenarios. For example, structure the use of Copilot in the accounting department for three specific tasks: budget variance analysis, explanatory notes on journal entries, and summaries of discussions with auditors. You define the rules, you train the team, and you measure during a month-end closing cycle. Concrete results over four to six weeks speak louder than a theoretical twelve-month strategy. The goal of a pilot project is to test how to integrate AI into specific daily work tasks.
5. What does success look like after the pilot project? Do not promise a revolution. Promise documented gains. Do not promise that AI will transform the company. Instead, promise measurable gains and controlled risk. For example, that 100% of the team uses the approved tool instead of external tools. That each colleague has integrated two or three targeted tasks into their weekly routine. That the time spent on certain repetitive tasks decreases in a documented way. And that no confidential data passed through an unauthorized tool during the exercise.
To help you organize this process, we have prepared a business case template that allows you to present the current situation, the associated risks, and a realistic pilot project to propose to management.
👉 Consult the business case template
Choosing the Right Approach to Demonstrate Value Quickly
The choice of initiative determines the credibility of the entire case. Aim for three characteristics.
- Recurrence. Targeted tasks occur at least weekly. For example, preparing monthly financial reports or writing follow-up emails to clients.
- Observable Impact. Time savings or error reduction are visible to the naked eye. For example, the number of minutes the team spends writing analyses that could be drafted by Copilot in Excel.
- Comparability. You can document a “before” and “after.” For example, the preparation time for a variance report from one month to the next.
If you want to support your case before presenting it, a Microsoft 365 audit like the one offered by Grav-ITI evaluates the readiness level of the current environment for structuring AI usage.
After the exercise, document what has changed: the number of colleagues using the approved tool, time recovered on targeted tasks, and security or quality incidents avoided. A successful first initiative opens the door to the next one without needing to restart the argument.
Structured AI usage is one of the challenges many managers face. To better understand how this issue fits into a broader set of Microsoft 365 practices, we recommend starting with our main resource on optimizing Microsoft 365 for team leaders, which serves as the starting point for exploring the various issues addressed.
How to Present an Improvement Project to Management
A general manager wants a status report, a realistic plan, and a way to measure the result. Here is an example of phrasing adapted to the AI context.
“My team is using artificial intelligence, but everyone is working with a different tool and without common rules. Financial data is ending up in free tools that the IT department does not control. We are losing time because no one shares what works. I propose structuring usage around Microsoft Copilot, which we already pay for. We start with one team, three specific situations, over a month-end closing cycle. No new investment. At the end of the exercise, we measure time gains, compliance, and actual adoption, and we decide together whether to expand the approach.” This type of situation quickly raises AI governance questions, particularly regarding data access and work practices.
This type of proposal works because it is concrete, measurable, and does not require an act of faith.
Structuring AI Usage to Create a Leadership Lever
The behaviors you observe in your team are not unique. In the majority of SMEs, AI is already being used, but rarely in a structured way. Everyone experiments on their own, with different tools, without common rules and without a clear vision of what the team is actually trying to accomplish.
Meanwhile, a gap is beginning to widen.
Some teams are learning to better use AI within the scope of their work. They are beginning to understand where AI truly adds value and how to integrate it into their ways of working.
In other teams, AI usage continues to develop without rules or a clear framework. Colleagues each use their own tools and methods, and no one really knows which practices should be prioritized within the team.
In this context, it becomes easy to feel that performance is improving simply because AI is being used. However, when each person uses AI in their own way, without common rules or validation of results, the gains are often “artificial.”
In other words, the team is using AI… but no one can really demonstrate what the company is actually gaining from it.
This is exactly what we explain in our article on the risks of a performance illusion when AI usage is not managed in an SME.
Structuring AI in a Team is a Management Decision
This means deciding how the team will use AI to create, build, and improve its ways of working, while ensuring that these gains also benefit the SME.
Concretely, this involves a few simple choices:
- defining clear usage rules
- choosing a common environment for the team
- identifying useful use cases based on roles
- measuring gains obtained in daily work
To structure AI usage in your team and turn it into a true lever for the company, the first step is to clearly understand the current situation.
For this reason, we have prepared a self-diagnostic that allows you to identify where your team stands today, what the main risks related to AI usage are, and what initial improvements can be considered.
Once this profile is established, you can use our business case template to organize your findings and present a structured approach to management, including a realistic pilot project to structure AI usage.
👉 Start with the self-diagnostic, then use the business case template
The goal is not just to regulate AI use; it is primarily to show how your team can use AI to create, improve, and develop new ways of working for the benefit of the company.
Why Choose Grav-ITI to Structure AI Usage
In many teams, AI usage first appears spontaneously. Colleagues test different tools, discover useful applications, and begin integrating AI into certain tasks of their work.
For a while, this works.
But when AI begins to be used by several people, in several roles, or across several teams, the same questions quickly arise:
- which tools should be used in the company
- what data can be used with AI
- how to validate the generated results
- and above all, how to transform these individual experiments into real gains for the company
This is often when team leaders seek to structure AI usage.
Grav-ITI supports both management and team leaders who wish to structure the use of Microsoft Copilot within their Microsoft 365 environment. The goal is not to introduce a new technology, but to help teams use Copilot to create, improve, and develop their ways of working, while respecting the company’s security and governance rules.
Concretely, the process allows for:
- analyzing how Copilot and Microsoft 365 tools are used in the company
- identifying the main risks associated with unmanaged AI usage
- defining usage rules adapted to the teams
- supporting the integration of concrete use cases into daily work
This process generally begins with a Microsoft 365 audit, which evaluates the environment configuration, information organization, and the company’s readiness level to use Copilot safely and productively.
As summarized by a business leader we supported:
“I thought our use of Microsoft Copilot was already well-structured. The audit highlighted several flaws we would never have seen. What I appreciated was the clarity of the recommendations and the ease with which we were able to take action.”
If you are ready to structure Microsoft Copilot usage in your team and transform current experiments into real gains for the company, let’s schedule a 30-minute call to discuss your situation. The conversation is non-binding and can help you identify the next steps to structure AI usage in your Microsoft 365 environment.
