ROI vs Cost Cutting With AI Initiatives

In partnership with

At a Glance

• Most AI initiatives can’t prove real ROI

• So companies shift to cost cutting — and layoffs

• But cutting too deep undermines the very systems AI is meant to improve

• The real opportunity: augment teams, don’t replace them

• AI works best when humans get better, not fewer

Hey folks,

By now you’ve probably seen the headline: 95% of AI initiatives fail to return their investment. Depending on who you ask, that’s either a sign of early-day experimentation or a flashing red warning that we’re repeating the tech hype cycle all over again.

Either way, the corporate reaction has been predictable. If AI can’t prove revenue, it now has to prove cost savings. And when cost savings becomes the only success metric, there’s a straight line from “efficiency” to layoffs — a line many companies have already crossed.

But here’s the thing no one wants to admit: these cuts assume the AI systems will magically deliver more value than the people they replaced. That promise hasn’t materialized. The humans left behind are burning out trying to keep everything afloat.

There’s a better path. One that starts with capability, not headcount. One where AI makes teams better — not smaller.

– Jeff

Why AI Isn’t Delivering ROI (Yet)

By now you’ve at least heard about the MIT report stating that 95% of all AI initiatives fail to return on their investment. To be fair, it’s early days and there’s a corporate stampede to make sure every enterprise doesn’t fall behind the AI curve. Organizations are scrambling to figure out ways to justify the oversized investment in internal AI work and, for the moment at least, are struggling to do so.

The net result of this failure has been a refocusing of success criteria for these initiatives away from revenue-based ROI and towards cost cutting. That may sound logical, but at the end of that line of reasoning are the inevitable corporate layoffs that will follow.

AI Is Expensive

The obvious costs of AI integration include licensing the models and paying for their ongoing usage. But on top of that companies need to hire specialists well-versed in artificial intelligence, machine learning and data science. Those folks don’t come cheap. Bidding wars in Silicon Valley are seeing total compensation packages for top talent topping $1MM base salaries and sometimes much more. Even outside of Silicon Valley these experts are few and far between and, once again, expensive. And let’s not forget good old technical debt.

As the rush to power our systems with AI increases in speed, many companies aren’t slowing down to think through whether AI is actually worth the cost in each of those initiatives. Just because we can add it into the system doesn’t mean we should. If we can’t recoup the costs of the technology and the teams to support it perhaps it’s not the right solution in every case.

If Not Profit, Then Cost Reduction

Which leads us to the situation we find ourselves in today, in many companies. The profit from AI work continues to be elusive. So much so that in several of our conversations with clients and folks inside large organizations in recent months we’ve seen a pattern develop.

AI initiatives no longer need to prove a revenue-based return on their investments. Instead, the only way AI initiatives are even considered in these organizations is if they can point to direct reduction in costs.

Of course there are lots of ways to reduce costs. For example:

• We can make our staff more efficient thus able to handle a bigger workload with the same headcount.

• We can remove tedious tasks completely out of the hands of human staff allowing them to focus on bigger, more challenging tasks.

But ultimately where this trend is pointing is layoffs. The promise of AI, at least at the moment, seems to be the ability to cut staff entirely. The humans who have been building these systems for years if not decades are now tasked with automating themselves out of a job.

The Cuts We’ve Seen So Far Have Been Deep

Tens of thousands of layoffs have been announced at marquee companies like Amazon, Microsoft and many others. The staff left behind is now tasked with maintaining these “autonomous” systems to ensure they can backfill the work the laid off humans used to do.

What Happens When the Systems Don’t Deliver?

There’s an expectation that not only will the AI systems replace the laid off humans but they’ll also be more productive in the long run. That promise has yet to be proven. In fact, many organizations are finding out now that they cut too deep. The staff left behind is overworked and overwhelmed. Even the cost reduction promise was exaggerated or, at best, poorly estimated.

The sad part is we’ve been here before. New technology is enticing. AI is probably the most enticing new tech we’ve ever seen since the internet. The promise of “the machines doing the tedious work” isn’t fully here yet. What if, instead of looking for people to fire to justify the spend on this new technology, we looked at ways to make the existing staff more productive? What if we taught them new ways of working that took advantage of this new tech in ways that do deliver on its promise?

We Don’t Have To Fire People To Be Profitable

It can be done. China, for example, is rewiring their entire society around AI. In the process they’re not discarding a population of workers. On the contrary, they’re training them to use these new tools to be even more productive and successful. If this can be done at the societal level, it can certainly be done at the corporate level. There just has to be a desire to do so and a realization that an AI-empowered staff is likely to be far more profitable in the long run than a laid off one.

Bottom Line

The rush to justify AI spend by cutting headcount is shortsighted. Yes, AI is expensive. Yes, ROI is elusive. But replacing people before the tech is ready — or before teams know how to use it — only compounds the problem.

The companies that will actually win with AI aren’t the ones firing the fastest. They’re the ones investing in their people, teaching them new ways of working and using AI to amplify human capability rather than eliminate it.

If you want durable ROI, don’t start with layoffs. Start with learning. AI-powered teams create far more value than AI alone ever will.

Upcoming Workshops from Sense & Respond Learning

Free Webinar: OKRs and Portfolio Management

If your OKRs feel “right” on paper but messy in practice… You don’t have to figure it out by trial and error. That’s exactly what we’ll unpack in our next Sense & Respond Learning webinar with Josh Seiden and Certified Training Partner Jim Sammons.

📅 December 18, 2025
🕛 12:00 p.m. ET
Register here

To view all upcoming workshops about Objectives & Key Results, Lean Product Management and Product Discovery for Agile Teams from Sense & Respond Learning, click here.

The Tech newsletter for Engineers who want to stay ahead

Tech moves fast, but you're still playing catch-up?

That's exactly why 100K+ engineers working at Google, Meta, and Apple read The Code twice a week.

Here's what you get:

• Curated tech news that shapes your career - Filtered from thousands of sources so you know what's coming 6 months early.

• Practical resources you can use immediately - Real tutorials and tools that solve actual engineering problems.

• Research papers and insights decoded - We break down complex tech so you understand what matters.

All delivered twice a week in just 2 short emails.

Join 100K+ engineers