AI Won’t Fix Your Salesforce. It Will Accelerate What’s Already Broken.

AI Won’t Fix Your Salesforce. It Will Accelerate What’s Already Broken.

A conversation with Phil Volltrauer, 25-year technology executive — Heartbeat: by Mike Sommer

Here’s the thing nobody wants to say out loud in the room where the AI budget is being approved:

AI doesn’t fix broken processes. It accelerates them.

That line came from Phil Volltrauer about twelve minutes into our first Heartbeat conversation — and it landed the way true things do. Not as a warning. As a verdict on what’s already happening inside most mid-market organizations right now.

Phil has spent 25 years leading technology transformation across energy, oil and gas, pharma, and life safety. He has been the CIO who inherited the mess, the one who built the strategy, and the one who had to go get the funding from a CFO who didn’t fully understand what was broken. He has also been the person in the room when the Salesforce implementation blew up — and when it didn’t.

What follows are the insights from that conversation that I believe every IT leader needs to read before they sign another statement of work.

The Problem No One Is Naming

Every CIO I talk to right now is navigating some version of the same situation. The business has been handed an AI mandate. The platform — Salesforce, in most cases — was built for a version of the company that no longer exists. And the people in the room who have the authority to approve the investment are the same people who have never looked at the actual state of the system.

Phil put it plainly: “If you don’t fix the fundamentals, you just merely accelerate chaos.”

That is not a technology problem. That is a decision problem. And the decisions are being made right now, in organizations just like yours, by people who are getting their information from parties with financial skin in the game.

Your system integrator built the platform. They profit from the next phase. Your Salesforce AE needs to hit quota. Your implementation partner is designing for a long-term managed services agreement — which means complexity is a feature, not a bug, from their perspective.

No one in that room is structurally positioned to tell you the truth.

What Phil Learned About Third Parties — And What He Got Wrong

Phil is skeptical of outside consultants. He said so directly, and he said it with the earned cynicism of someone who has paid for expensive PowerPoint decks that parroted back what he already knew.

His initial position: a strong CIO shouldn’t need a third party to tell the story they can already tell.

I pushed back. And he did something rare — he revised his thinking in real time.

The distinction that emerged from that exchange matters: there is a difference between a third party that comes in to tell your story for you, and one that surfaces platform-specific truth you don’t have the time or structural independence to surface yourself.

Phil’s words: “If it’s in partnership with the CIO, then all it is is more ammunition to drive the strategy forward.”

That is exactly right. The CIO drives the strategy. What they need is an independent, conflict-free source of the data that validates — or challenges — their read of current state. Not another vendor. Not another partner. A witness with no stake in the outcome.

Clarity Before Everything

We spent time in the conversation on what I see as the first failure point in most AI initiatives: the absence of a shared picture of reality before a single dollar is committed.

Not strategic vision. Not a roadmap. A one-page, non-technical diagram that answers three questions simultaneously for the sales team, the IT function, and the executive team: What are we actually trying to do? What does the platform currently support? Where do those two things not match?

Phil’s framing of the gap was precise. He described organizations where the sales team believes their process has five steps — that’s the context they live in — and the Salesforce metadata tells a completely different story. Twenty-five steps. Built over years of accumulated workarounds, custom objects, and process debt that no one has ever formally documented.

When that gap exists, you don’t have an AI problem. You have a clarity problem. And no AI tool — not Agentforce, not Einstein, not anything Salesforce demos at Dreamforce — can close that gap for you.

The Customization Trap

Phil’s rule of thumb on Salesforce customization: use the system as designed from zero to eighty-five percent. Customization is the path from eighty-five to one hundred. It should never be the path from forty to one hundred.

He identified exactly why organizations keep falling into the trap. Functional leaders who rose to their current role based on how things have always been done experience out-of-the-box processes as a threat to their identity, not just a change to their workflow. The “ick factor,” as Phil called it. The sense that vanilla best practices somehow erase what made them valuable.

I shared a story from my time at IBM — a Salesforce implementation that collapsed because a group of very smart engineers tried to force the platform to become a CPQ-first tool. Salesforce said don’t do it. The project blew up. Cost a lot of money. The lesson that emerged when we came in for the retrospective: define what actually differentiates your business. Customize ruthlessly there. Leave everything else alone.

Phil’s response was honest: “I don’t always win that one.”

Neither does anyone else. But you can’t afford to stop having the conversation. The moment you let customization creep from the boundaries into the core, you have guaranteed that the platform will not support what comes next — including AI.

How to Get the Funding to Fix What’s Actually Broken

You know what’s broken. You know it needs to be fixed before you can deliver the AI outcomes the business is expecting. And you know that “we need to remediate technical debt” is not a funding conversation that ends well in most boardrooms.

Phil’s advice: attach it to what already has momentum. If there is an AI initiative with board-level enthusiasm, the remediation is not a separate project. It is the foundation of the AI initiative. The cleanup is the launchpad.

He was careful to note that slapping “AI” on a project is not a magic funding mechanism — at least not with leadership teams sophisticated enough to look past the label. But if you can show that the foundational work makes the headline initiative more likely to succeed, and more likely to succeed faster, you have a real conversation.

His personal preference: stay close to the CFO. Regardless of where the funding comes from, know the financial state of the business. Know what discretionary budget exists. And when you can’t find it anywhere else — he said this plainly — fund it from within IT. Keep a portion of the budget reserved for exactly these situations.

The Question Every CIO Should Ask Right Now

Phil described his role as Chief Information Officer also becoming Chief Education Officer during the AI wave. The reason: the market is full of noise. Leaders are getting their information from vendors selling the dream and consultants selling the engagement. Nobody is coming in as the disinterested party whose only job is to tell the truth.

That is the gap Business MRI was built to close.

Before you commit to an Agentforce rollout, before you sign the next phase of a Salesforce implementation, before you approve the AI budget your team has been building toward — someone needs to answer one question without a financial stake in the answer:

Is this platform actually ready?

Not “can we make it work.” Not “what would it take.” Is it ready — cleared — for what you are about to bet on it.

That is the Clearance. And the only people who can give it to you are the ones with nothing to sell on the other side.

Listen to the Full Conversation

Phil Volltrauer brings 25 years of hard-won perspective to every question in this episode. We covered the engineering mindset gap in IT leadership, how to navigate the politics of transformation, where AI governance needs to sit inside the organization, and what it actually looks like when a CIO becomes the connective tissue across every function.

Watch the full episode on YouTube or listen wherever you get your podcasts.

Connect with Phil on LinkedIn: https://www.linkedin.com/in/phillipvolltrauer

Where Does Your Platform Actually Stand?

Business MRI delivers pre-investment diagnostics to mid-market companies preparing to bet on AI and automation. Independent. Conflict-free. Binary outcome: Cleared or Not Cleared.

If you are a CIO or IT leader being asked to deliver an AI mandate on a platform you are not sure can support it — start with a three-minute Clearance Scan.

Take the Clearance Scan: https://scan.businessmri.com

Heartbeat is the Business MRI podcast — conversations with practitioners who have navigated the gap between what the platform promised and what it actually delivered. New episodes bi-weekly.