Artificial intelligence has the potential to improve care radically. Think of systems that detect disease sooner, ease administrative work, or monitor patients more precisely. Yet in hospitals and care institutions we regularly see disappointing situations: AI systems that do not do what they promised, care staff who become frustrated by the extra work, and managers at their wits’ end because nobody is using the new system.
At Twentynext we see every day that this is not because the technology falls short. The stumbling block usually lies in forgetting a crucial balance between what we call the hard and the soft side of innovation. The hard side is about clear agreements, sound contracts, and manageable cost and risk. The soft side is about the people who work with it: do they experience the added value, does it fit their way of working, and do they understand the system?
Academic insight helps to understand that balance. On one side stands Transaction Cost Economics (TCE), which holds that organisations must always account for the costs and risks that come with any new technology or partnership. According to economist Oliver Williamson that means thinking carefully about dependence on suppliers, uncertainty around performance, and how frequently a technology is used. The more specific an investment, the greater the dependence and therefore the risk of vendor lock-in.
On the other side stands Innovation Diffusion Theory (IDT), from sociologist Everett Rogers, which explains why some innovations are adopted faster than others. Rogers highlights five factors that are crucial for acceptance by users: perceived advantage, compatibility with existing processes, ease of use, the ability to try it out first, and visibility of positive results.
What turns out to be the case? When AI implementations fail, it is usually because either too much emphasis falls on technology and contracts (so nobody experiences the real benefit), or because attention goes only to the human side, without a clear framework or financial viability.
The art of a successful implementation lies in combining those two perspectives. It means care managers should ask sharp questions from the outset. For example: does this system genuinely save nurses time? Does it connect well to existing systems? And do we have a clear and flexible agreement with the supplier, so we are not stuck if things turn out differently than expected?
A hospital in Utrecht sets a good example here. They introduced an AI system on a small scale first. Care staff could experience what it gave them, while the hospital kept its options open with flexible agreements. Only once users were enthusiastic and the benefits were clear did they move to broader implementation.
At Twentynext we hold the same philosophy. We believe in starting small, learning continuously, and above all paying attention to the professionals who will work with AI every day. We make sure hard and soft factors go hand in hand from day one.
Would you like to know how this could work in your organisation? Curious how to limit contractual risk and get your colleagues enthusiastic about smart innovation at the same time? We are happy to help.
Get in touch at info@twentynext.nl or arrange an introductory conversation with no obligation. Together we will make sure your next AI project genuinely succeeds, not only on paper, but on the ward. Interested in the whitepaper behind this article? Send a request to info@twentynext.nl.
Twentynext: data-driven care, delivered with people in mind.
