AI Readiness Means Pointing AI at the Right Problem

Amy has handled the same call hundreds of times. The experience of handling all of those calls has given her a unique insight not only into how the company's operation works, but more interestingly, how it breaks... and then gets repaired. When a customer reaches out about an issue with their order or an error in their account, she's reaching for the fix before the caller finishes their sentence. That valuable first-hand knowledge makes Amy an important part of the company's ongoing success and an authority on the company's practical operational realities.

Three floors up, Melissa, the founder, has already done what every leader was told to do. Eighteen months ago she authorized the IT team to roll out AI. She gave it a budget, a champion, a launch announcement. If you ask her today whether it worked, her honest answer is probably “I’m not sure”. Something is being used somewhere. A number moved, or it didn't, for reasons nobody can clearly specify. She was told adoption would be the hard part. Nobody told her that adoption would not necessarily equate to impact.

She's not alone. The first wave of AI adoption has run its course. Companies bought in and gave it real time, and what's left is the harder question: can anyone actually convey what resulted?

The First Wave of AI Adoption Has Run Its Course

MIT researchers looked across corporate AI pilots and found 95% producing no measurable financial impact. They trace that result not to any shortfall in the technology, but rather to where many companies aimed their AI initiatives.

And as we know, these initiatives don’t come at a real financial cost, with Uber’s story serving as a most notable example. It pushed staff to use AI as much as possible, ranked them on a usage leaderboard, and burned through its annual budget in only a few months.

That's the shape of the story most companies are living. Money went out. Usage went up. Months later leadership can't tie any of it to a result that truly matters. Companies are spending AI dollars on an array of subscriptions and software with the intention of creating positive change both externally and internally, but are left discouraged, or even uncertain as to what progress they've made as a result. Where is the impact?

AI Impact Gets Lost Between the Front Line and the Top Floor

The impact may well be there. Two things happen at once that keep leadership from seeing it, and either one alone would be enough.

The first is that a leader's view of her own company is assembled by other people, and fear edits that view before it comes into focus. More than a third of workers worry AI will make part of their job obsolete, which gives nearly everyone below the top floor a reason to shade what they report. Close to a third already use AI without telling anyone. Put those together and what our CEO Melissa is actually measuring is how comfortable her people are admitting to something, which is a very different number than actual adoption. The managers meant to relay reality upward are the most depleted layer in the company, with engagement among them falling from 27% to 22% in a year. An exhausted relay drops the signal.

There's a second problem underneath the reporting, and it's a matter of design. The decisions about where to deploy AI were made without the people best positioned to know where it would genuinely help. The plan itself was built without an Amy in the room, so even a flawless rollout was aimed at the wrong problem from the beginning.

Most corporate AI budget flows to sales and marketing, to converting and growing, while the larger returns (ie. Impact) sit in service and operations, in the work of serving people well and keeping them. (Fortune) In the case of Melissa’s company, optimizing Amy’s the call sits on the wrong side of that line. It is a service moment carrying a retention outcome, even though AI investment has largely been directed elsewhere

So leadership ends up with a clouded readout of activity, built on a plan that never had the full picture to begin with. Money and effort were never the missing ingredient. Knowing where to point them was, and that starts with who's in the room when those decisions get made.

Underneath both sits an important distinction. Reporting tells a company what already happened. Intelligence tells it where a customer is heading and what reaching them will cost. Melissa has a wall of the first and almost none of the second, which is why her instruments can show her the spend and never the opportunity.

AI Readiness is a Question of Proximity

When the ROI number comes back flat, the instinct is to ask whether the company chose the right tool. That treats readiness as a shopping problem, as if the right platform, purchased and swapped in for the last one, will eventually locate its own use case.
The second instinct is to aim for cheaper. Cost and speed are the only outcomes to measure, which makes them tempting stand-ins for impact. But making the wrong work faster or less expensive does not make it the right work.

McKinsey points somewhere more specific. Employees are already ready, and leadership is the largest barrier to return. Leaders underestimate how much their own people use AI by roughly a factor of three, and firms seeing real return are three times likelier to have leaders who use the tools themselves. The variable that predicts payoff is proximity.

Find AI Use Cases the Way You Fix a Customer Journey

This is the same problem UX and CX practice has been solving for years. Nobody repairs a broken customer journey by polling the executive team and buying software. You go watch the work. You sit with the person on the phone whose actual experience provides the clearest model in the building of where things break and what a real fix requires.

Amy has that model. The purpose of sitting with her is to find where the legacy process wastes her time, doubles her effort, or forces her to manually work around a poorly designed or broken process. Those are the opportunities where AI can close the gap and give her more room for the part of the job only she can do. Her value has always been the pattern recognition underneath the repetitive motion, and the frictions that entails. Readiness aims AI at reducing the friction in front of Amy.

What AI Readiness Gives Back to the People Doing the Work

When this works, the same wrong order arrives and Amy no longer has to untangle the problem from scratch. Because someone finally asked what kept slowing her down, AI can surface the relevant context, recognize the likely issue, and put the right next step in front of her. Amy is still on the call, but she can spend less time navigating systems and reconstructing what happened and more time actually helping the customer. The freed time and attention go to the moments that need her judgment, including the exceptions and relationships that cannot be reduced to a workflow.

AI Readiness Starts With One Conversation

Readiness can start with one conversation. Sit with the person on your team who lives inside the process every day and ask where it slows down, where they lose time, and where they end up working around the system instead of through it. Then ask which parts of the job they believe should stay theirs.

You will leave with a short list of places where AI can take real work off their plate and a clear line marking where its support has to stop, and both of those came from the one person no rollout plan thought to consult. The software has been ready for a while. This is the part that makes the people ready too.