The path to enterprise AI adoption can be daunting. Many businesses are hesitant to take the first step thanks to concerns about implementation risks. Yet, once in place, AI systems can ‘hallucinate’ and produce inaccurate outputs that lead to errors, financial impacts and potential reputational damage.
Yet, AI can transform organisations, and the technology’s ability to handle subtle and complex tasks is only growing more refined. Business leaders must not let concerns prevent them from taking the crucial first steps towards AI integration.
Success depends heavily on aligning AI initiatives with business goals. The question isn’t whether to adopt AI, but how to mitigate risks while moving forward strategically.
The first step is experimentation, in which businesses explore different AI use cases across the organisation to discover not only the technology’s potential but also its limitations, from both a technical and cultural standpoint.
Next is the impact phase, where leaders showcase the tangible benefits AI experiments can bring to an organisation. Finally, it is essential that leaders focus on long-term innovation and develop sustainable strategies around AI that balance the needs of today with the demands of tomorrow.
Experimentation
When it comes to experimenting with new tools, the thought of trial and error can be off-putting for many businesses, as it suggests a drain on time and resources without any promise of a solution.
However, this is a misperception, according to experts at the global design and innovation company IDEO, who argue that continuous experimentation is one of the hallmarks of successful businesses competing in an evolving landscape.
“A lot of the time the technologies that we are hearing about are new, so we don’t always know how they are going to shape our businesses, products or processes. Encouraging experimentation is an easy, low-barrier-to-entry practice that helps us explore how a technology might make a bigger impact before we make massive investments,” says Savannah Kunovsky, managing director of emerging technology at IDEO.
“Experimentation also creates familiarity in an organisation, helping to uncover use-cases that maybe the technology wasn’t built for in the first place and that can provide a competitive edge,” she says.
While it’s a good idea to have processes and guidelines to encourage and direct experimentation, Kunovsky says many workers are likely experimenting with different tools already, without telling their managers.
“A good place to start is asking people inside your organisation what they’re already doing,” says Kunovsky. “If you want to set up a process to understand how a technology like AI might impact a specific part of your business, start by determining whether it’s already being used.”
She continues: “If your people do not understand how these tools and technologies might affect their careers, then it can be an overnight surprise once something bigger rolls out. It’s often in people’s best interest to do that experimentation themselves.”
Skipping the experimentation phase can be risky and potentially very costly, according to IDEO’s executive director, Sergio Fregoni, who observes that there’s often no other way to determine whether a specific tech deployment will be the perfect solution for the business.
“In a world that’s increasingly unsettled, there’s little chance that you can actually be certain about what you’re going to put out into the world,” he says. “More strategising, more thinking in a vacuum, will not give you more confidence about your product and its product market fit.”
Impact
So how can leaders decide whether a particular use case is appropriate for a company-wide rollout? According to Kunovsky, listening to employee feedback and making sure their questions are answered is vital.
“People don’t want to adopt something that’s not useful to them,” she says. “Find ways to figure out what will help that individual, because that may be different to the person sitting next to them.”
Rather than simply dictating when and how to use a new tool, Kunovsky recommends running training sessions and workshops for people to figure out how different technologies might help them in their specific roles.
“Systematic experiments help you to figure out more effectively what’s working and what isn’t, and then establish best practices and figure out how to adopt them in different parts of an organisation.”
For example, one of IDEO’s teams focuses specifically on emerging technology and understanding the implications for the company before tools are rolled out to the wider business.
“It’s not every person’s job to stay at the forefront. That’s just exhausting,” says Kunovsky. “We identify things that we believe can be valuable, present them to the team to see how people respond to it and then let them run with it. This helps us to roll out tools we know are going to be highly impactful. We prototype and evolve them before launching to a broader group of people. It’s a little bit like product incubation.”
For large businesses, collaboration with other departments is also helpful, since it can build buy-in from across the business, while effective communication can drive new ideas internally, says Fregoni.
“When you want to build a new belief around a new strategic direction, we don’t recommend overhauling the whole organisation all at once,” he says. “To bring different departments along on the journey, you have to understand them. You have to make them feel understood and heard. Then you have to design for them. It is essential to have good storytelling to build belief.”
Kunovsky recalls a recent AI adoption project IDEO ran with a large conglomerate. The parent company was made up of more than 20 separate businesses, so IDEO set up a boot camp that worked vertically across the organisation, bringing in selected employees at different levels – from individual contributors and middle managers all the way up to the CEO – to introduce them to new technologies.
“We gave them insight into what the technology was and how it worked, and took them through an accelerator-style experience where we really got into the problems the tool might solve for them. We looked at potential processes and solutions, and how to implement them in day-to-day work. Then we took whatever was working and spread it out across different parts of the conglomerate,” she says.
Innovation
This focus on communication is also key to convincing decision-makers to get on board. For this group, setting a realistic timeframe is crucial, since conversations about new technologies often come down to ROI.
“It’s very hard to have constructive conversations about culture, change and transformation when the ROI horizon is, say, three months or the next two quarters. Those conversations tend to be more about short-term operational gains and less about innovation,” says Fregoni.
Conversations about innovation require a longer time horizon, he says. “You need to start unpacking the problem and ask: ‘What is the North Star? What is the vision? Is this technology going to transform our offer or the way we interact with customers?’ Being able to articulate a set of preferable futures for your organisation, then working backwards from them, will help you take the first steps.”
Above all, however, ensuring a strategy stays in motion and is flexible to change is a defining factor in a business’s long-term success.
Implementing a fixed strategy, given the current pace of innovation in AI, is a “recipe for disaster”, according to Fregoni. “If you think back to two years ago, nobody was talking about ChatGPT or DeepSeek, and now it seems like nobody can live without it,” he says.
Make sure you can fail fast, talk about the failure and not sweep it under the carpet
He continues: “Strategy is a living thing – it’s what you do, not what you say you’re going to do. If you’re responding to market forces – like a changing landscape, changing competitors or changing customer expectations – it means your business strategy is evolving and your AI strategy should change accordingly.”
The most successful organisations are those that are open to trying new things, failing and iterating as they go along. This flexibility should permeate the organisation’s operational activities and strategic decision-making.
“The best learning organisations are those that create the conditions for psychological safety – failing at pace and making sure that you can learn from failure. As a leader, it is your job to make sure feedback loops and learnings are effective,” says Fregoni.
“If the failure loop is short, it’s fairly inexpensive to fail - and if it’s inexpensive to fail, then it isn’t a problem to fail,” he says. “Make sure you can fail fast, talk about the failure and not sweep it under the carpet. Those conditions create psychological safety, which helps to foster a culture of continuous experimentation.”
The message is clear: businesses can’t let fear of failure stall their transformation efforts. By embracing rapid experimentation and open dialogue on setbacks, organisations can make failure less costly and less threatening. In today’s competitive landscape, the biggest risk isn’t getting AI wrong - it’s not getting started at all.
For more information, please visit: ideo.com
The path to enterprise AI adoption can be daunting. Many businesses are hesitant to take the first step thanks to concerns about implementation risks. Yet, once in place, AI systems can ‘hallucinate’ and produce inaccurate outputs that lead to errors, financial impacts and potential reputational damage.
Yet, AI can transform organisations, and the technology's ability to handle subtle and complex tasks is only growing more refined. Business leaders must not let concerns prevent them from taking the crucial first steps towards AI integration.
Success depends heavily on aligning AI initiatives with business goals. The question isn't whether to adopt AI, but how to mitigate risks while moving forward strategically.
