Many organizations report struggles with their big AI implementations
A few weeks ago, I spoke at an event on AI and the future of services businesses with Salesforce and Certinia. Just before the panel, MIT had released a study showing that 95 percent of AI implementations don’t produce measurable value. This data grabbed everyone’s attention. We’d all seen what AI is capable of. Where is this value gap coming from?
It’s easy to think of possible reasons why 95 percent of implementations don’t deliver the expected value. Some happen too quickly without enough planning, and many may suffer from poor data quality. Some use subpar tools. But these challenges can happen with any new technology in a business, and at least at Spaulding Ridge, our hit rate is much higher than that. The million-dollar question must be “What is working in the five percent of implementations that do deliver value?” My opinion: Trust.
Trust in AI comes from managing the change
One of the biggest challenges in delivering any technology solution is ensuring that it sticks. Staff members need to trust it enough to use it at launch, and then see enough value to continue using it throughout its lifespan. AI adds new layers to this challenge. We’re already seeing independent adopters within organizations experimenting with standard large language models like ChatGPT or Claude. These early adopters often find faster ways of doing things and create “shadow systems” that are surprisingly sticky, sometimes even more so than larger, purpose-built implementations. Every company has these innovators, but the harder task is bringing the next wave of employees on board.
Produce results and set expectations
Convincing the doubters is rarely about the technology itself. Usually, it’s about how the organization manages the transition. To build trust, leaders need to show early wins. Demonstrating that AI can deliver a 90 percent correct answer in minutes—compared to days of manual effort—helps employees see immediate value. Trust also requires clarity and transparency. Setting expectations that AI is iterative, not perfect, makes employees more comfortable experimenting. Training and communication play a critical role here. Staff need to know not just what tools are available, but how they fit into workflows and why they matter. Just as important, leaders must make it clear that trying, iterating, and even making mistakes are part of the process. When employees feel supported, not judged, they are more willing to experiment. This strengthens adoption and builds confidence in AI over time.
Of course, speed matters, but time to value isn’t the only definition of success in AI. Rushing too quickly can actually harm trust in the long run. We’re still approaching AI’s tipping point. The capabilities are exciting, but the real transformation is still ahead, and we don’t yet know exactly how it will reshape business. The true goal is to prepare organizations and their people to keep delivering value after that tipping point arrives. Getting hands-on, building comfort, and learning along the way is the only way to ensure that when AI reaches maturity, your business is ready to take full advantage.
Trusted AI allows for deeper results
One of the most important shifts that comes with trusted AI is how organizations evaluate results. Traditional technology rollouts were often measured by whether a system worked or not. But AI tools are far more flexible, and with that flexibility comes a need to rethink evaluation. Rather than focusing on adopting a system or product, the question becomes: what outcome are you trying to achieve? With AI, you can directly measure whether your technology investments are delivering the results you expected. That requires educating leadership and boards. It’s no longer about spending on tools, services, or capabilities—it’s about spending on outcomes.
Another key change is that AI models don’t stop at deployment; they continue learning and adapting. This is a very different rhythm from the binary “it works or it doesn’t” mindset that many customers are accustomed to. With AI, value often comes in stages—an initial version may deliver 90–95% accuracy, but the model keeps improving as it learns from new data and interactions.
AI changes how value is assessed
This shift changes how advisors, vendors, and clients work together. Instead of seeing value as a one-time deliverable, organizations need to embrace an iterative approach where progress compounds over time. Far from being a limitation, this adaptability is one of the most exciting aspects of AI.
Finally, trusted AI depends on trusted data. For AI adoption to scale, organizations must take responsibility for the quality, consistency, and governance of their data. This is not just about standing up a single source of truth or keeping the data platform up-to-date, though those are essential. It’s about building a culture of data responsibility across the organization. The more accurate, consistent, and well-governed the data, the more powerful the AI models become. That shared responsibility will unlock deeper results, ensuring that AI supports and elevates the outcomes businesses are striving toward.
AI makes a human advisor even more critical
When I think about the explosion of AI agents and tools in the market right now, my head starts to spin. Every week, there seems to be a new entrant promising revolutionary capabilities. In a technology landscape like this, what organizations really need is not just more tools, but an air traffic controller—someone who can coordinate the systems, projects, and strategies into a cohesive whole. The architecture has to work seamlessly, and that requires a partner who not only understands the technical complexity but also grasps the broader strategic goals of the business.
In an AI-powered world, the role of an advisor has shifted. It’s no longer about installing point solutions or deploying a single system in isolation. Instead, it’s about weaving these different systems and agents into workflows that actually drive business outcomes. Leading firms like Spaulding Ridge focus on this integration—making sure the technology isn’t just impressive on paper, but operationally valuable in practice.
The payoff of this approach is a better value exchange for organizations. Just like technology tools are expected to deliver measurable outcomes, advisors today must also be outcome-oriented, not just service providers. The real value of an advisor doesn’t lie in the behind-the-scenes activity of consultants, but in how people and AI agents can work together to solve real business problems. By combining human judgment with the speed and scalability of AI, advisors help companies unlock results that neither could achieve alone.
Ready to begin your AI transformation? We can help.
Even as the reality of AI adoption becomes clearer—and in many cases, harder than organizations initially expected—the urgency hasn’t diminished. Businesses know they need to embrace AI to stay competitive, but the path forward can feel daunting. From pilot programs to enterprise-level strategies, trusted advisors can help chart that path to value, ensuring organizations not only adopt AI but actually gain from it.
That’s why conversations about AI are more important than ever. Whether you’re just starting to explore these tools or already deep into your journey, advisors like Spaulding Ridge can help you define the right approach for your business. We’d love to continue the conversation—reach out to us to see how human expertise and AI together can move your business forward.