AI: why you Should Slow Down and Define Your Process

New artificial intelligence (AI) tools are released every week. They offer innovative features and the promise of spectacular gains. In light of how quickly the technology is evolving, you’re probably feeling under pressure to take action quickly. However, while this sense of urgency is understandable, it can result in poor decision-making.

According to a study published by the IBM Institute for Business Value in 2025, 76% of Canadian business leaders stated they were investing in technology without fully assessing the ROI for their organization. In reality, only 27% of their AI initiatives delivered the expected returns.

These results reveal the challenge facing business owners. Investment and experimentation do not guarantee success. An organization could launch several initiatives without making progress on tackling the real business challenges.

A sense of urgency to act does not justify a hasty investment. If you’d like to fully harness the potential of AI, you must adopt an approach that allows you to move forward quickly without losing sight of your business priorities.

Focus on the challenge rather than the tool

Let’s take the example of an organization that would like to improve its customer service. The management team is impressed by the possibilities of AI and rolls out a chatbot to answer client questions.

A few months later, the results are disappointing. The tool provides the correct answers to common questions, but the turnaround time for more complex requests is still long since questions must be forwarded to employees. Why is this? Employees still have to check multiple systems to find the information they need.Essentially, the company automated some of its client interactions without addressing the real problem: information cannot be accessed quickly.

To avoid falling into this trap, start out by assessing your business challenges rather than technological possibilities. What issues are your teams struggling to resolve? What are your client and user priorities? On which levels could you significantly improve your work practices?

Giving this some thought can help you to identify situations where AI could add something that your current approach cannot offer or would have difficulty doing so.

A roadmap can help you avoid dead ends

Having a clear direction doesn’t necessarily require spending months drafting an action plan. Since the sector is evolving so rapidly, waiting until you assess every potential outcome would be counterproductive.

The objective is to define a roadmap that is specific enough to guide your decisions while being flexible enough to adapt as you learn.

Start out by assessing your organization’s situation. Do you have the required data? Can your systems support the planned initiatives? Do your teams have the skills they need?

Then you can move forward and define your priorities based on the expected benefits, available resources and project complexity. Focusing your attention on a few promising initiatives is better than spreading your resources thin as you juggle a multitude of tests.

Your first step should involve determining measurable goals and planning regular assessments. This approach will equip you with specific benchmarks that you can leverage to track progress and guide the next steps.

Stay in control

In many organizations, AI projects are primarily led by the Information Technology (IT) teams. Since these teams have the expertise needed to evaluate and integrate tools, this is understandable. However, they should not be the only individuals tasked with decision-making. The transformation process must be led by management with relevant teams also contributing.

Let’s take the example of the education sector. A teaching establishment could develop an AI application to help teachers prepare their lesson plans and design exercises tailored to their students. However, if teachers are not consulted prior to the implementation stage, certain features may fail to meet their needs. The completed work may have to be reviewed, which would result in additional costs and delays.

As a business leader, you must therefore ensure you leverage the right expertise (operations, finance, human resources or customer service, for example) from the outset. Your role doesn’t require mastering the technical aspects of AI. Instead, you should determine priorities, make decisions and ensure the projects deliver the expected value.

You may also have to make difficult decisions. If a pilot project doesn’t deliver the expected results, you should be prepared to pull the plug even if you’ve already invested time and money. These resources could be put to better use in a more promising initiative.

The five-question test

In order to determine whether your AI approach is on the right track, select an ongoing project and ask yourself the following questions.

• Why did we launch this project? Does it address a clearly identified business challenge?
• What value are we seeking to create? What specific results do we expect for the organization, employees and clients?
• Is this the right priority? Compared to other initiatives, does this project deserve the resources allocated to it?
• Do we have the conditions for success in place? Are the right people on board? Do we have the data, skills and resources that we need?
• What have we learned so far? Do the initial results justify continuing, adjusting or shelving the project?

If you’re struggling to answer that question, it might be time to reevaluate your priorities and how you assess your projects.

You can acquire the best AI tools while maintaining control over where they take you.