The Road to AI: Turning Concepts into Concrete Results
While boardrooms buzz with AI strategies and venture capital flows toward AI startups at unprecedented rates, a fundamental question persists: How do companies really move beyond the hype to deliver tangible value?
The gap between AI's theoretical potential and its practical implementation continues to challenge even the most forward-thinking organizations. A study released in July by MIT claiming that 95% of corporate GenAI pilots fail spooked markets and sparked new debates about an AI bubble.
At Revaia's recent Portfolio Days, three industry leaders gathered to dissect this challenge, offering candid insights from the trenches of AI adoption. The conversation revealed that successful AI implementation isn't about grand transformational visions or following competitor moves. Instead, it's about incremental progress, deep customer understanding, and maintaining focus amid the noise. The discussion highlighted how companies across different stages of AI maturity are navigating regulatory complexities, pricing challenges, and the delicate balance between innovation and execution.
Guillaume Vitrich, a partner at White & Case, moderated the discussion between Roger Dudler, founder and CEO of brand management platform Frontify, and Nnamdi Emelifeonwu, co-founder and CEO of legal productivity suite Definely — both Revaia portfolio companies. Together, they represented the spectrum of AI integration: from AI-native companies where artificial intelligence forms the core foundation, to established businesses retrofitting AI capabilities into existing products and workflows.
“The big challenge is to create excitement for something, and then find the sweet spots beyond the hype,” Roger said. “Every customer sees a fancy AI image generation thing, but no one really gets value out of it on a daily basis. You get really fast to something exciting and something impressive, but getting to a place where it truly delivers value, that is very hard to make it scalable.”
Small Steps Over Grand Plans
The panelists converged on a counterintuitive approach to AI strategy: start small and build incrementally rather than pursuing comprehensive transformation plans. This philosophy challenges the conventional wisdom that successful AI adoption requires sweeping organizational change and massive upfront investments.
Roger, whose company serves major brands like American Express and Lufthansa, emphasized the cultural dimension of AI implementation. "You need to first work on the culture to make sure you get everyone on board for the journey ahead," he explained. "For me, it was about convincing people of the exciting opportunities while acknowledging their fears and challenges, and then gradually making small changes in the product to deliver quick wins."
This approach proved particularly effective at Frontify, where early wins came from automating metadata management for images, a seemingly mundane but valuable capability that demonstrated AI's practical benefits without requiring months of development. "I'm slightly allergic to big projects and plans, because I believe that in fast-changing areas, they often end up obsolete by the time they’re delivered," Roger noted.
Nnamdi echoed this sentiment from the perspective of an AI-native company operating in the traditionally conservative legal sector. "Our strategy has been to be very intentional about what we do - and do it exceptionally well," he said. Rather than attempting to cover every legal workflow, Definely focused specifically on contract drafting and review, building deep expertise in that narrow domain.
The incremental approach extends beyond product development to internal operations. Both leaders emphasized using AI to enhance existing capabilities rather than replace entire teams or processes overnight.
Measuring Success Beyond the Hype
One of the discussion's most revealing moments came when addressing ROI measurement, a topic that many AI vendors and adopters struggle to quantify effectively. The panelists offered honest perspectives on the challenges of demonstrating concrete value from AI investments.
Roger challenged a fundamental assumption about AI's financial impact: "There's a big misconception in the market that applying AI for productivity and efficiency means companies like ours will need fewer people. It's not really about saying, ‘ we need less staff for x, y, z.’ Maybe it means we don’t have to hire as quickly. But the real point is: if you don’t adopt AI, you risk losing business — because everyone else will.”
This perspective reframes AI adoption as a competitive necessity rather than a direct profit driver, at least in the near term. The real value lies in maintaining market position and meeting evolving customer expectations rather than achieving dramatic cost reductions.
Nnamdi provided a concrete example of how AI creates measurable business value beyond traditional metrics. "Last year, before we started rolling out our AI empowered products, there were a lot of companies or firms that we went to where they would actually just have the budget solely if your technology has AI in it. And if you didn't have it, then you were put to the back of the list."
This "AI table stakes" phenomenon illustrates how artificial intelligence has become a qualification criterion rather than just a differentiator. Companies without AI capabilities find themselves excluded from procurement processes, regardless of their other strengths.
Guillaume, drawing from his experience advising corporate clients, reinforced this dynamic: "Many people think that lawyers are incentivized to spend as much time as possible on a matter because they do bill by the hour. The very reason for that is there is too much work for us. So productivity is key for us, and AI will help us a lot on the tasks where it's needed."
Staying Grounded While Monitoring Competition
The conversation revealed a sophisticated approach to competitive intelligence that balances awareness with strategic focus. Both companies maintain dedicated channels for monitoring competitor activities, but they've learned to filter noise from signals.
"We also have this famous competition Slack channel, which is fully packed with everyone posting in there," Roger said. "But it can be misleading to always look at what the competition is doing. I think if you believe in something that’s going to make the difference in two years, you kind of need to have a certain belief that something specific will work and something else will not."
Roger Dudler, founder and CEO of Frontify - Nnamdi Emelifeonwu, co-founder and CEO of Definely
Nnamdi shared a particularly instructive example of how media attention doesn't always correlate with customer success. "I was at a conference last week, and one of the actors C-level to a top white shoe law firm in the US came up to me randomly and just say, hey, I really love what you guys are building in our company, in our firm, your product is used twice as much as this other company, and they're the ones who have all the limelight."
This disconnect between publicity and actual usage metrics underscores the importance of customer-focused development over attention-seeking features. Both leaders emphasized the value of deep customer relationships and regular feedback collection as more reliable guides than competitor analysis.
"It's really important to monitor, know what your competitors are doing, because that gives you an advantage," Nnamdi noted. "But it's also equally important not to lose track of what your vision is. You have to just believe in what you're doing. You can't get distracted by noise."
The Regulatory Reality Check
Perhaps no aspect of AI implementation proves more complex than navigating the evolving regulatory landscape, particularly for companies serving enterprise clients. The discussion revealed how regulatory uncertainty creates both challenges and opportunities for AI vendors.
Nnamdi, whose legal tech company works with major law firms and corporations globally, described the delicate balance required: "You have to build your products in a way where they can scale quickly as the regulation updates itself. You have to build that trust in your product and also your process with your customers, from day one."
The regulatory challenge extends beyond compliance to customer confidence. Law firms, inherently risk-averse, require extensive security reviews and often prefer on-premise deployments over cloud-based solutions when AI is involved.
Roger identified a practical approach to managing regulatory risk and building trust with customers: focus on transactional AI applications rather than open-ended generative capabilities. "Finding those elements that are more transactional, like detecting objects in an image: how many glasses are in this picture? How many people, how many women versus men, or stuff like that. Is it safe for work or not? This is much easier to put in place first, to get confidence in the system."
This strategy allows companies to build internal confidence and regulatory comfort with AI while avoiding the higher-risk scenarios associated with general-purpose language models.
The discussion concluded with a forward-looking perspective on talent and organizational adaptation. Rather than dramatically restructuring hiring practices around AI skills, both leaders emphasized curiosity and adaptability as enduring qualities that will serve organizations well regardless of technological changes.
As AI continues evolving at breakneck speed, this emphasis on fundamental human qualities—curiosity, adaptability, customer focus—may prove more valuable than any specific technical expertise.
The road to AI ROI, it seems, is paved with incremental progress, deep customer understanding, and the wisdom to stay grounded amid the hype.
"You need adaptable people. You need smart people, people who can learn fast and jump into something new at any time," Roger said. "I just believe in those people more than in someone who can simply prove that they know something really well."