Skip to content
Work Tech Weekly
Jason Bodin

Why Decision Fatigue Is an Organizational Performance Problem

“How many decisions do you think a person makes per day?”

I guessed 8,000. Jason Bodin didn't blink: “I'm gonna go with thirty-five thousand.”

We talk a lot in this industry about AI saving time. What we don't talk about enough is what's actually eating that time in the first place — the small stuff. The approvals, the exceptions, the "hey, can I take Friday off" requests that shouldn't require a human decision at all, but somehow always do.

My guest is Jason Bodin, EVP of Marketing and Communications at Paycom, the single-database HR and payroll platform. Jason's spent 13-plus years building Paycom's brand — he came up through sports journalism, same as me, and that background gives him a sharp instinct for holding a position while the market tries to copy your homework.

We get into decision fatigue as an org-level design flaw, why automation on messy data doesn't reduce cognitive load, and why trust is the real bottleneck in realizing the benefits of Agentic AI.

Why Decision Fatigue Is an Organizational Performance Problem
  43 min
Why Decision Fatigue Is an Organizational Performance Problem
Work Tech Weekly
Play

Decision Fatigue Is a Systems Problem, Not a You Problem

Thirty-five thousand decisions a day. Most of them invisible — what shoes to wear, whether to accept that LinkedIn connection request that’s obviously a step in a sales cadence, whether to approve the fifth PTO request for the same Friday.

Jason's point is that organizations have been treating this like a personal bandwidth issue. Work harder. Manage your time better. Get a system for your inbox.

Wrong answer.

“If technology can peel these off, think about the bandwidth that allows within the organization for all of the employees,” Jason says. It's not a discipline problem. It's an architecture problem that most companies are stacking AI on top of instead of fixing.

A better answer? Build the policy into the process and the tech. Managers shouldn’t need to approve or deny time-off requests one by one, absorbing the awkwardness personally. Let the policy decide. "The manager doesn't become the bad guy," Jason explains. "It becomes the technology that has the process and the policy within it."

This is the whole thesis in miniature: Remove the decision, remove the fatigue.

Fragmented Data Creates Its Own Decision Fatigue

Here's the uncomfortable truth for anyone selling AI right now: it's only as good as the data underneath it.

“AI is a multiplier. It's not a miracle,” Jason says, which is a line that should probably be stitched onto a throw pillow in every enterprise software office. He's blunt about the current state of the models too: they hallucinate, they need fact-checking, and “you've still gotta go back and review the data.”

Paycom's answer, dating back to founder Chad Richison putting payroll on the internet in 1998, was to refuse the integration-and-API approach everyone else took. One record base. One source of truth. Ninety-one percent of HR pros, Jason notes, say they want exactly that, and yet the average employer is juggling six different HR and payroll systems. Is this a sales pitch? Yes. That doesn’t mean there isn’t some truth in there.

The agentic AI wrinkle makes the fragmented data problem even worse. “You may have agents that are watching agents that are watching agents,” Jason says, only half-joking. Without a clean data foundation and an actual process, agentic AI doesn't accelerate anything.

“An organization without a process going into agentic AI is set up to fail,” he says.

Trust Is the Real Bottleneck That Agents Won’t Solve

There’s a crucial element that gets lost in conversations about AI at work: Can I trust the output?

If employees don't trust what the system tells them — about their paycheck, their PTO, their eligibility — the questions just move somewhere else in the org. Whether it goes to HR or your boss’ boss, it’s going upstream. That’s an additional, far more expensive decision tax.

Paycom's payroll transparency tool, Beti, is the case study here. The tagline "now employees do their own payroll" landed like a grenade when it launched. People assumed employees would overpay themselves or break tax law. Instead, Jason says, once workers got visibility before payday, “voids, corrections, reversals — those things are gone.” Transparency created trust. Decisions were eliminated or moved downstream rather than up. That’s the lesson.

That's also the catch in the leadership dividend that AI vendors are promising right now. Automation is supposed to hand managers back their time. They can stop refereeing PTO requests and start actually leading. But that trade only clears if people trust the system enough to stop relitigating its calls.

Trust isn't a feature you ship. And, if you want the AI leadership dividend, it’s a step you definitely can’t skip.

 

Let's Talk