Most Companies Aren't Getting a Return on Their AI Investment. Let Me Tell You Why.

I've led enterprise level change management and AI integration at multiple companies. Please learn from me.

here is an actual visual representation of integrating AI

12 months into a large scale AI integration at a Fortune 500 SaaS company and employees still weren't utilizing AI and it wasn't returning on the large financial investment executive staff were hoping for.

We are talking massive investment - huge branding / internal communications campaign, this was not going away.

So what did they do?

They made it mandatory.

Rolled out KPIs that required AI use.

When employees pushed back and shared that the AI tools weren't providing reliable data, that it was creating more work to double check they outcomes AI provided, they were told they just needed to use it more. They needed to spend their time training it.

When employees asked if they were training their replacements, management laughed it off.

I talked to one of the comployees who were intended to be the ones benefiting the most from this new technology. She wasn't anti-technology. She wasn't afraid. She was completely reasonable, she'd watched two previous technology rollouts at this company go nowhere, she'd done the training both times, and she'd learned that the smart play was to look compliant and keep doing her job the way she knew how to do it.

She wasn't resisting the AI. She was managing her own career risk in an organization that had never given her a reason to believe this time would be different.

Our adoption numbers were going up.

I built a measurement system that was excellent at counting the wrong things.

First of all,

I don't like wasting my time.

Secondly,

I like having big goals and I liked achieving big goals.

Third,

I don't like feeling like I've been misled.

The number that should be on every leadership team's agenda right now.

MIT's Media Lab released a finding in 2025 that got a lot of coverage and blew up LinkedIn.

Despite $30 to $40 billion in enterprise investment into generative AI, 95% of organizations are getting zero return.

When I saw that number, I didn't feel vindicated. I felt implicated. Because I know how organizations end up in that 95%, and it's not because the technology failed. It's because we told ourselves a story about transformation while doing remarkably little to actually transform anything at the organizational level.

• PwC's survey of 4,450 CEOs across 95 countries found that only about one-third of enterprises have seen any tangible benefits from AI in the last twelve months

  
  

• Deloitte found that most organizations aren't seeing satisfactory ROI until two to four years in. That's two to four times longer than the standard payback period for a technology investment. Only 6% saw meaningful returns within the first year. Deloitte

Nobody put that timeline in the board presentation when the budget got approved. I know because I've seen a lot of those presentations, and they tend to be long on vision and short on the part where it takes three years before the numbers move.

Here's what I find more interesting than the ROI data, though. Despite all of this, 85% of organizations increased their AI investment in the past twelve months, and 91% plan to increase it again this year. Deloitte

So we are collectively spending more, faster, on something that mostly isn't working yet, while the people driving those decisions are largely shielded from the reality of what's happening at the ground level.

Then executives are shocked when we see employee engagement metrics dropping and our best performers leaving for compettitors.

The thing nobody says in the post-mortem.

I have sat through more AI program retrospectives than I can count across tech and financial services companies. Big ones. The kind where real money was spent and real careers were staked on the outcome.

I have never once been in a room where the conclusion was: leadership didn't lead this well enough.

It's always the data. The legacy infrastructure. The vendor. The employees who "weren't ready." Never the leadership team that approved a three-slide strategy and then assumed execution would handle itself.

Harvard Business Review published a piece in 2025 that said what most post-mortems won't: most AI initiatives fail not because the models are weak, but because organizations aren't built to sustain them - and scaling AI is less about technology and more about creating the organizational backbone that turns experiments into measurable business results. Harvard Business Review

The organizational backbone. That's the work. Not the tools.

A separate piece in HBR that I've quoted in probably a dozen executive briefings since put it even more directly: many AI projects fail because leaders treat adoption as a tech purchase instead of a behavioral change problem. People resist tools that disrupt routines, overreact to visible AI errors, and prefer familiar human judgment - and as a result, even good systems fail to gain traction. Harvard Business Review

Tech purchase versus behavioral change problem.

Change is hard.

I complained when my toll road switched to tap pay instead of the automatic liscense plate scan. I use an email address from high school. I still listen to CDs in the car.

There are significant advantages to changing all of these things - it doesn't matter, I don't do it because it's difficult. It requires an extra step instead of an automatic process.

How much more difficult is it to absorb change when it comes to our careers? Our jobs? Our livelihood? Why would I trust a machine to do what I've been doing for years? Especially if I don't entirely trust it's output. If it messes something up, that is my name on the line and my career at stake.

In SaaS - there is a move fast and break things approach - In financial regulation, there is a long term data preservation approach. 2 completely different industries, both rolled out AI integrations like a communications campaign instead of a years-long culture shift. We thought the day we finally hit full production was the transformation - it was only the beginning. We hadn't built anything to sustain what came after the launch.

What's actually happening while leadership watches the dashboard.

Here's a cycle I have watched play out at multiple companies, across very different cultures and very different budgets.

The executive team announces the initiative. The energy is real and genuine. These aren't cynical programs. The leaders I've worked with genuinely believed in what they were building. This really can help make employees more productive. Save time crunching data. The communications go out. The town halls run. The training completion numbers climb.

And underneath all of that, something else is happening.

A 2025 survey of 800 executives and 800 employees at large enterprises found that half of executives said AI is actively "tearing their company apart."

Around half of employees reported that AI-generated information at their organization is inaccurate, confusing, or biased.

And 41% of Millennial and Gen Z employees admitted to sabotaging their company's AI strategy - refusing to use the tools or the outputs. Axios

Forty-one percent. Not skeptical. Not slow to adopt. Actively working against it.

Have any of your senior leaders actually tried the AI tools they are encouraging their staff to use?

In my experience, a big chunk of that isn't because the tools lack value.

It's because employees feel their work is so de-valued that a robot could be trained to do it.

AI Integration isn't adding another tool to your Sharepoint tech stack - is is an organizational culture change

The biggest barrier to enterprise AI scaling isn't employee resistance. Employees are largely ready. It's a lack of leadership drive and vision - leaders aren't steering fast enough, clearly enough, or consistently enough.

The employees were never the primary problem.

I'll say that again because it runs counter to how most AI change management is designed: the employees were not the primary problem.

We built elaborate frameworks to manage their resistance, designed communication campaigns to overcome their skepticism, ran listening tours to get ahead of their fear. And the data says we were solving for the wrong thing.

AI adoption isn't a launch event; it's a journey that requires continuous conversation. Two-way dialogue not only surfaces adoption barriers early - it demonstrates that leadership values employee input.

The primary problem was - and in most organizations still is - leadership that approved a transformation without fully understanding what leading one actually requires.

Not announcing one. Not funding one. Leading one. Every day, for years, in ways that are visible to the people being asked to change.

That's an uncomfortable thing to bring to a C-suite. I know because I've done it, and the reception has ranged from genuinely receptive to politely hostile. But the communications leaders I've seen do their most effective work in AI integrations are the ones who found a way to deliver that message early — before the program was in trouble, when there was still time to course-correct.

What this series is actually for.

This is the first of six posts. Each one goes deep on a question I get asked constantly by communications and change leaders who are in the middle of these programs right now - the questions they ask me directly, not the ones they put in a webinar chat.

• How do you actually get employee buy-in when the tools don't always work and the fear is real? 

  Not the theory. What I've seen work, and what I've watched backfire badly including on myself.

• What does a multi-year AI communications strategy actually look like?

  Not a launch campaign. A discipline built to survive leadership transitions, budget cycles, and the energy collapse that hits almost every program around the eighteen-month mark.

• Why do these programs fail, and what can a communications leader realistically do about it? 

  Because naming the problem and having the standing to fix it are two different things, and I want to talk about how to build that standing before you need it. Also the one thing I did that made the suprising biggest impact.

• How do you measure adoption in a way that actually reflects reality? 

  Including the specific mistakes I made with metrics, what they cost, and what I rebuilt after that conversation on the floor.

• And what is the real job of an internal communications leader in an AI transformation? 

  Because there's a version of this role where you're handed the messaging after every major decision has already been made. And there's a version where you're shaping the decisions. The outcomes are not remotely comparable — and the difference between those two versions is almost entirely about how you position yourself before the program starts.

Here's where I land on all of this.

42% of companies abandoned most of their AI projects in 2025.

The average organization scraps 46% of AI proofs-of-concept before production.

The gap between what's being invested and what's being returned is real, it's large, and it's not closing as fast as anyone projected.

But here's what I also know after doing this work at multiple companies over multiple years: the failures are not mysterious.

They're not the result of technology that doesn't work or employees who are fundamentally unwilling to change. They're the result of predictable, repeatable organizational mistakes that can be anticipated, named, and in many cases prevented - if the right people are in the right conversations at the right time.

McKinsey found that companies with mature change management practices are 3.5 times more likely to outperform their peers in technology adoption. Medium 

Not AI strategy. Not data architecture. Change management. The discipline that lives closest to internal communications of anything in the organization. Stop treating culture and organizational growth and employee engagement and intenral communications like woo-woo nice-to-have afterthoughts. They literally run your company. The more you invest in them the more productive your employees will be. The more your company is valued.

The path forward is not complicated. It's hard, and it's long, and it requires a kind of organizational honesty that most leadership teams are uncomfortable with. But it exists. I've seen it work. And the communications leaders who understand what they're actually being asked to do in these programs > not messaging, not campaigns, but building the conditions under which real behavior change becomes possible < those are the people who end up doing the most consequential work of their careers.

That's what this series is built on. Let's get into it.

Post 2 of 6 → The Real Reason Employees Aren't Using Your AI Tools (And Why Your Communication Strategy Is Probably Making It Worse)

Why employee buy-in is failing and what leadership is doing wrong - the assumptions they're making about employee enthusiasm that don't match what employees are actually experiencing.