Most Leaders Aren’t Trying to Win with AI—They’re Trying Not to Lose
AI has become the new “digital transformation”: overhyped and underdelivered.
I’ve worked with executives across industries—from financial services to IT providers to SaaS companies—and I keep seeing the same pattern: most AI initiatives never make it past the “copilot” phase. Not because the tech isn’t ready, but because the culture isn’t.
Leaders talk about data quality, model accuracy, or lack of AI talent. But those are just symptoms. The root cause is deeper—and human.
The Real Reason AI Initiatives Fail: Fear
You don’t get transformation when your top priority is avoiding blame.
Research suggests that more than 75% of transformations fail. AI is no different. We’ve been burned before—by failed ERP rollouts, cloud migrations that made IT more expensive, and analytics platforms that delivered dashboards but no decisions.
So when AI enters the conversation, we default to risk management. The instinct isn’t to win. It’s to not lose.
That creates a culture of hedging: small experiments, lab environments, and science projects with no connection to business goals. When I see an “AI task force” or CDAO reporting to a VP of Innovation with no P&L responsibility, I already know how the story ends. That’s not innovation. It’s marketing.
Start with the Business Outcome
I often get asked, “Do you do AI?” Wrong question.
What I want to know is: What’s the one business constraint you’d break if you could?
For example, one client in financial services wanted to grow revenue without hiring more reps. We focused the AI initiative on increasing profit per employee—from $200K to $2M. That gave the project teeth. When resistance came, we had a hard number to anchor the strategy.
Don’t mistake goals for objectives, though. There are too many unknowns early on. Set the goals to guide strategy. Create quantitative objectives only once a cadence is established.
I also met with the COO of a $150M manufacturing company with a 60-day quote-to-cash cycle. They knew it was killing margin and blocking expansion. We structured the AI initiative around compressing that to under 7 days—unlocking both EBITDA growth and a new mid-market playbook.
When you lead with business outcomes, AI becomes a means to an end—not a shiny object. It also becomes a rallying cry for why we must change the way we do things.
The Curse of Past Success
The leaders most resistant to AI are often the ones who succeeded under the old model.
They built businesses on process, efficiency, and scale. AI breaks that mindset. It’s not about more process—it’s about adaptability. It’s not about scaling what worked—it’s about discovering what works next.
I’ve had to rewire that thinking. We begin with why: What is the business imperative for exponential vs. incremental improvement? Then we move on to metrics leaders already track—customer acquisition cost, time to quote, revenue per head—and model what happens if those numbers shift by 10x instead of 10%. That’s when the lightbulbs go off (sometimes).
That shift is hard to absorb when your performance model is rooted in Excel forecasts, Six Sigma, and ITIL. The worst place to start an AI initiative is in IT, where risk mitigation is more important than innovation. Sorry folks—successful ERP migrations don’t count as innovation (although they do count as execution).
Scenario: Financial Services — Precision Growth with AI + Execution
A financial services firm wanted to increase wallet share without spamming clients with generic offers. Their data team had predictive models, but legacy systems and compliance overhead made it nearly impossible to act in time.
We used the PrivOps Matrix data fabric’s speed-to-data to efficiently unify entitlements, client context, and marketing policy—making them executable across systems. AI predicted the right moment to act, created the right offers, and triggered the workflow within PrivOps, which sent the offer to the customer, enforced policy, and logged the execution.
The real challenge was getting buy-in from sales, marketing, data, and IT operations teams. (Compliance was easy.) Without the business case, it would have been impossible.
Result:
– Cross-sell conversions increased by 4x
– Compliance risk dropped
– Campaigns became continuously adaptive, not quarterly exercises
Scenario: IT Services — Real-Time Sales + Delivery Orchestration
A managed service provider I talked to was struggling with integrations that prevented them from using AI and automation to streamline operations. They had been working on it for over 3 years.
We layered the PrivOps Matrix Data Fabric across CRM, service desk, SIEM, quoting, provisioning, and billing—in less than 2 months. The entire process became programmable. Using AI and the data fabric, we automated over 90% of onboarding: quote, approvals, and provisioning.
The client had to fire the data engineering lead who kept trying to sabotage the project—but the rest of the team got on board once they saw success was possible.
Result:
– Onboarding dropped from 60 days to under 3 days
– Cost of sales fell by 35%
– New product/service launches dropped from 6 months to 3 weeks
Culture Is the Hardest Part
Even when the strategy is solid and the tooling works, resistance remains.
AI threatens workflows. It exposes inefficiencies. It shifts power. And it demands a faster, looser, more exploratory cadence. That’s deeply uncomfortable for organizations designed to optimize, not adapt. Design to fail and iterate.
But here’s the truth:
Managing culture is harder than managing models.
And it’s the part most leaders overlook.
How to Fix the Culture
If you want AI to transform your business, the culture has to change. Here’s what I’ve seen work:
Tie AI wins to personal wins.
If the COO’s bonus is tied to quote-to-cash, frame the AI initiative as their best shot at hitting the number.Make early results visible.
Deploy fast, narrow wins—30–60 day initiatives with measurable impact. Agile thinking, not waterfall.Reduce execution risk.
Use infrastructure like the PrivOps Matrix to ensure that once AI makes a decision, the system can follow through—compliantly and programmatically.Shift incentives.
If your org rewards process adherence over adaptability, it’s set up to fail. AI rewards iteration and feedback loops.Normalize iteration.
Frame AI not as a one-time project but as a new operating model. If digital transformation failed before, your AI initiatives will too—unless you change how things are done.