Husqvarna Group were seeking new ways to utilize AI in their organization. A successful pilot study of Microsoft’s AI assistant, Copilot, was initiated when they turned to the business side and the management team.
Husqvarna Group, founded in 1689, is a global player manufacturing everything from chainsaws to robotic lawnmowers. The company’s 330+-year history proves that innovation has always been a keyword, and now, it is exploring how AI can further strengthen the business.
Consequently, a new department, Enterprise AI, was formed as part of Husqvarna’s central IT function. Witold Pawlus leads Enterprise AI, which, under the Group Management team’s mandate, has identified five priority areas where AI seems to offer the most value.
“We could have done this from an IT perspective and pushed out new tools without considering business benefits. However, we believe in involving Husqvarna Group’s business side and asking, ‘What are your biggest challenges, and how could we solve them using AI?’”
Witold explains that all of Husqvarna Group’s seven units—three divisions: Forest and Garden, Gardena and Construction, and four corporate functions—have an AI transformation lead.
“The role links the business with the Enterprise AI team. The AI transformation lead contributes a business perspective and ensures the tools are implemented and used in practice.”
Business and Customers at the Forefront
According to Pawlus, having the business and customers at the center of such an initiative is key to achieving long-term value for Husqvarna Group.
“Both IT infrastructure and business processes must work together. The internal goal is to have a better business and more efficient employees, but ultimately, it’s all about a better customer experience,” he says. He continues, “Previous experiences have shown that it’s not enough to have data scientists and AI engineers involved; there also needs to be an understanding of how AI can strengthen business processes. You can have the best tool in the world, but if you don’t know how to use it and for what purpose, the tool has no intrinsic value.”
One of the first tests for Enterprise AI was to conduct a pilot study of Microsoft’s AI assistant, Copilot, when 300 licenses were distributed to everyone from the management team to customer service.
“With Copilot, we saw the potential for higher efficiency among our test pilots. To examine this on a broad front, we wanted to allow many roles to test the tool.”
Copilot a Natural Choice
Pawlus explains several reasons why Copilot was chosen over other similar tools.
“Copilot is integrated with Office apps. This makes it convenient. We don’t have to worry from an IT perspective since we are already part of Microsoft’s ecosystem. We can also feel secure about information protection and training the model with internal data.”
Before the pilot rollout in the fall of 2024, 30 early adopters tested the tool during the summer.
“We didn’t do anything special this time; we just provided them with links to publicly available training materials. That test group answered survey questions we would later compare with the pilot study’s responses.”
Copilot Freed Up 34 Minutes Every Day
The 300 people in the pilot study underwent an official introduction and training program with help from representatives of Microsoft’s partner. This proved to significantly affect the results. The survey comprised 90% of questions from Microsoft, one being, “How much time do you estimate you save each day by using Copilot?”
“On average, the participants in the pilot group reported that Copilot freed up 34 minutes each day—compared with Microsoft’s studies, which mentioned 30 minutes—while the early adopter group saved 12 minutes.”
A significant difference indeed.
“From an IT perspective, it might seem simple: we pay and roll out. But now, we see what happens if we invest in a training program. It creates entirely different values. And now we have proof of that. My takeaway is that a holistic and continuous training experience definitely contributes to increased benefits from using the tool.”
85% Reported Increased Productivity
Pawlus sees the saved time as the biggest benefit. In the survey, 85% of the pilot study’s participants reported becoming more productive compared with Microsoft’s benchmark figure of 70%. He mentions how he uses Copilot in his daily work and how it has improved his efficiency.
“I use it to summarize Teams meetings. I let Copilot summarize longer presentations before meetings, and I get help when creating presentations or writing text. Today, I have examples of tasks that take one hour to complete instead of eight hours.”
Pawlus also remarked that only very few—five out of 300—wanted to give up their license at the end of the test period. He takes this as further proof that both the tool and the onboarding process were successful investments.
“Also, some people reported feeling less stressed and not needing to work overtime. They can do more within the same working hours. How the results can be translated into concrete figures on the bottom line remains to be calculated.”
Further Rollouts and Tests Await in 2025
The goal is to prove the tool’s value by 2025. The next step is to allocate more licenses.
“Here, we will also spread the license allocation widely, with some specific roles in mind where the tool can have a measurable impact on revenues or cost savings. We will carefully measure usage and business effects. Then we can decide on potential broader rollouts.”
Pawlus’ Three Key Factors for a Successful Copilot Implementation
- Start with the business—prioritize the business needs and ensure to keep the customer in the focus. AI must create concrete value for the business and its customers.
- Clarity and transparency build trust. Build trust by clarifying what the AI tool can and cannot do, ensuring we understand how we can use the tech for the benefit of our customers.
- Leadership engagement is crucial. Leadership must commit to resources, training, and usage; therefore, getting the leadership on board early is essential to effectively serve customers.
