How a French Spirits Company Created Employee Buy-In for AI

This article was produced by Harvard Business School Working Knowledge and features the insights of faculty members Iavor Bojinov and Edward McFowland III. For many businesses, digital transformation fizzles when employees resist, but at Pernod Ricard, buy-in bubbled up as the company carefully nudged employees to embrace new tools powered by artificial intelligence. Its success in gaining staff acceptance surprised Harvard Business School professors Iavor Bojinov and Edward McFowland III, who expected the 200-year-old French spirits company to encounter significant pushback from employees accustomed to selling and marketing products the traditional way—without the aid of AI. “I went in with the hypothesis that this was going to be very hard, if not impossible, to do well,” says McFowland, an assistant professor. “I was very pleasantly surprised by the fact that they had done a fairly good job—better than some tech companies I’ve seen.” Subscribe to the Working Knowledge newsletter to get their research coverage in your inbox every week. The digital journey of Pernod Ricard, the world’s second-largest spirits company, offers guidance in navigating a common challenge organizations face today: how to get employees on board with employing the advanced technology that executives are investing in as they look to streamline operations and boost profitability. “The value only comes when people actually use it,” says Bojinov, an associate professor. “It could be the best tool, but if no one uses it, who cares?” McFowland and Bojinov say the key to smoother adoption is proactively addressing resistance as well as training and supporting employees who are hesitant to give new digital tools a try. The researchers wrote the May 2024 case, “Pernod Ricard: Uncorking Digital Transformation,” with Francois Candelon of the BCG Henderson Institute, and Nikolina Jonsson and Emer Moloney of the HBS Europe Research Center. Employees Are Initially Skeptical As a global company operating in more than 70 countries across 90 production facilities, the drinks giant had built its success based on local market expertise and traditional relationship-driven sales and marketing, with staff often operating on intuition and gut instinct to determine the profitable path forward. But by 2020, managing an increasingly complex brand portfolio with analog processes felt unsustainable for company leaders. Plus, company leaders believed a digital transformation would help its sales and marketing teams become more data-driven and efficient in their spending. Pernod Ricard developed an ambitious solution: create four key digital programs powered by AI. The two flagship tools were: D-STAR, a system that used machine learning to optimize sales representatives’ store visits and product recommendations. Matrix, which used AI to allocate marketing spending across brands and channels. The real test came when they introduced the tools to employees. Initial rollouts revealed the depth of employee skepticism. In France, some marketing managers resisted Matrix’s recommendations to scale back on some marketing initiatives because they challenged managers’ emotional attachment to certain brands. In Germany, sales representatives worried that D-STAR might diminish their strategic efforts—the part of the job many valued most. Some resistance stemmed from legitimate concerns about job security and professional identity. But much of it reflected a deeper organizational challenge: Employees had little reason to trust or invest in tools they hadn’t asked for and didn’t understand. How the Company Gained Buy-In To convince employees of the tools’ value, Pernod Ricard developed a careful approach built on four key pillars: Demonstrate real value through testing. The company ran A/B tests to prove the tools delivered measurable improvements. When it deployed D-STAR in France, stores where sales reps followed the AI recommendations saw better net market share growth and net sales growth compared to control groups. At that point, the benefits weren’t theoretical; they were tangible improvements that sales teams could see in their results. Take the risk out of adoption. Pernod Ricard restructured performance evaluations to remove fear from the equation. Sales representatives who followed AI recommendations but missed their targets weren’t penalized, while those who ignored the tools and missed targets faced scrutiny. “If you follow the recommendations and you don’t quite meet the quota, that’s OK,” Bojinov says. “But if you don’t follow the recommendations and you don’t meet the quota, that’s not OK.” Invest in education and support. The company created dedicated deployment teams for each market rollout, including local change management specialists, data analysts, and trainers. Employees received extensive education on how to use the tools and why they were valuable. Plus, the company created hotlines so employees could immediately report problems or get help. Leverage internal champions. Perhaps most importantly, Pernod Ricard named respected employees in each market who could serve as technology ambassadors. “They identified the people who were listened to,” McFowland says. “These are people who have been around for a while, and who have a really good reputation.” Once these influencers embraced the tools, peer adoption typically followed. The strategy turned digital transformation from something management imposed to something employees actively wanted. “That was really the genius—you can get employees pulling instead of management pushing: ‘I want this, I need this,’ versus ‘You have to take this,’” McFowland says. By 2023, D-STAR achieved 85% adoption rates across deployed markets, while Matrix reached 60% to 70% adoption despite being more disruptive to traditional marketing workflows. The tools delivered sales increases between 1.5% and 4.5% depending on the market, plus up to 15% improvements in marketing efficiency. Balancing Control and Accountability The researchers say a key insight from the study revealed the need for companies to strike a balance between the “span of control” and “span of accountability.” When AI tools reduce employees’ control over their work processes, organizations must also adjust how they hold employees accountable. Pernod Ricard recognized this tension and addressed it, for example, by telling sales reps using D-STAR that they wouldn’t be blamed if they followed AI recommendations that didn’t pan out. “Very often, deploying AI reduces the span of control, but rarely do organizations match that with a change in the span of accountability,” Bojinov says. “That creates a negative incentive to adopt tools because you have less control, and yet you’re still accountable for the same things.” Advice for Other Organizations Based on the study, the researchers offer the following guidance to companies trying to encourage employees to accept new tools: Start with pilot programs that prove value. Don’t roll out tools companywide until you can demonstrate concrete benefits in controlled settings. Pernod Ricard’s approach of testing in select markets first allowed them to refine both the technology and the adoption strategy. Build dedicated deployment teams. Successful adoption requires people whose full-time job involves managing the transition. These teams should include change management specialists, not just technical staff. Address the human side early and often. Most digital transformation failures stem from people issues, not technology problems, the researchers say. Invest heavily in training, communication, and support systems. Adjust organizational processes, not just technology. Don’t just drop new tools into existing workflows. Redesign processes, incentive structures, and accountability measures to align with new capabilities. For instance, Pernod Ricard provided additional bonuses to employees who used the tools successfully. Aim for 85% adoption before expanding. Pernod Ricard wouldn’t deploy tools to new markets until existing markets achieved 85% adoption rates—a threshold intended to ensure the technology and processes were truly working. As AI becomes increasingly central to business operations, the Pernod Ricard case offers hope for traditional companies worried about bringing employees up to speed. Success isn’t about having the most advanced technology, the researchers say. It’s about having the organizational capabilities to deploy that technology effectively. “It’s not a matter of if you should do this; it’s a matter of when you should do this,” Bojinov argues. “And the answer is yesterday.”

How a French Spirits Company Created Employee Buy-In for AI

Summary.

Partner Center