Not ‘CAN WE’ but ‘should we’

It’s time to reTHINK the ROI of AI

When most organizations talk about AI, they talk about efficiency. How much time can we save? How many costs can we cut? How many people can we replace?

Efficiency is not a strategy. When AI is adopted only as a shortcut to save time and money, it often leads to degraded products, frustrated customers, and even reputational damage. We’ve already seen what happens when companies chase efficiency above all else: enshittified experiences that extract value rather than create it.

The better question isn’t “How much can we save?” but “How can we make what we offer better?”

Used thoughtfully, AI can augment human creativity, improve products and services, and open new possibilities for customers. It can support employees rather than squeeze them, and it can create experiences that build trust rather than undermine it. That requires asking not just whether we can deploy AI in a given task, but whether we should.

Framing ROI this way shifts the focus from exploitation to innovation. It asks leaders to measure success in terms of improvement, trust, and long-term value, not just efficiency gains. It also invites transparency: telling customers and employees when and why AI is being used, so the technology is seen as a tool to serve them, not as a way to extract from them.

ROI will always matter. But the organizations that thrive will be the ones that broaden their definition beyond cost savings, and use AI to create products, services, and experiences worth investing in.