Wren Okafor-Tully is not, technically speaking, a real person — though if you work in enterprise analytics, you probably sit three desks away from them. They are a composite: a senior data analyst, five years in, whose team's output has improved by every metric that matters since agent deployment eight months ago. They are also, as of six weeks ago, quietly updating their LinkedIn on their phone during standups. We spoke over video on a Saturday morning, which Wren chose specifically because "nobody from work is going to accidentally see this on my calendar." They were drinking coffee from a mug that read WORLD'S OKAYEST ANALYST, which they held up to the camera unprompted. "This used to be ironic."

You told me you wanted to talk about this. Why?

Wren: Because I keep having this argument with myself and I'm losing both sides? I don't know. I read something, the HBS piece about the satisfaction gap,¹ and it was like, oh, someone gave this a name. That was weirdly comforting. Like when you finally Google a symptom and it turns out to be an actual condition and not just you being dramatic.

What's the argument?

Wren: Okay, so. My team is objectively better. I can see it. Our turnaround on quarterly reporting went from two weeks to three days. The anomaly detection alone: we had a routing cost issue last quarter that the agent flagged in forty minutes. That would have taken me a full day, maybe two, and I might have missed it anyway because the pattern was across three datasets I wouldn't have thought to join. So I'm not delusional. The work is better.

But I don't do anything anymore. Not in the way I used to. I used to build the analysis. Now I approve it. I check it. I look at the agent's output and go, yeah, that looks right. And it usually does look right. And then I move on to the next one.

That sounds efficient.

Wren: It's extremely efficient. I'm very efficient at agreeing with a machine.

Is that unfair to what you actually do?

Wren: Probably. A little. I still catch things. Last month I caught a unit conversion error that would've made it into the board deck. That felt great, genuinely great. I told my partner about it at dinner. "I caught an error today." Five years of training for the thrill of catching one error per month.

But here's what I keep coming back to. When I started this job, I was cleaning CSVs. Absolute grunt work. And everyone says that was the bad part, the drudgery, the manual stuff. They're right, it was tedious. But it was also how I learned what the data felt like. You develop this sense for when numbers are wrong before you can articulate why. You build that by handling the raw material over and over.² I'm not sure you can build it by reviewing outputs.

Do you think you're losing that sense?

Wren: I think I'm not exercising it. Which might be the same thing over time. There's this moment every day where I'll be looking at an agent-generated report and I'll think, I should dig into this section more, and then I look at the clock and there are fourteen more reports in the queue because we produce so much more now,³ and I just... don't dig. I approve it. And it's probably fine. It's almost always fine.

Almost always.

Wren: Yeah. Almost always.

Your team's metrics are strong. Your manager is happy?

Wren: My manager thinks this is the best our team has ever performed. She's not wrong. She put me up for a spot bonus last quarter.

And you're job-searching.

Wren: (long pause) Yeah. I mean, I'm looking. I'm not flying to interviews. I'm in that zone where you scroll job boards at 11 p.m. and save things and don't apply to most of them.⁴ It's not urgent. The pay is good, actually better than it was two years ago, which is its own weird thing.⁵ I'm not miserable. I just...

What?

Have you talked to your manager about this?

Wren: And said what? "The team is performing too well and I'm sad about it"? In this economy?⁶ No. The narrative inside the company is that this is a success story. And it is a success story. I'm not going to be the person who stands up at the all-hands and says actually I miss writing SQL queries. That sounds like nostalgia. It sounds like resistance.

Is it?

Wren: (thinks) No. I think it's something else. I was hired to be an analyst and now I'm a reviewer, and those are different jobs, and nobody asked me about the switch. It just happened. The title didn't change. The salary went up. The work got hollowed out.

What are you looking for, when you look?

Wren: Honestly? I don't know. That's the worst part. Every job posting I see either wants someone to manage agents, which is what I already do, or wants skills I'm not sure I still have at the level I used to. There was a posting last week for a senior analyst role at a smaller company, no agent stack yet, and I almost applied. And then I thought: am I really going to go backward? Is that even possible?

Would you take it?

Wren: Ask me on a Tuesday when I've approved forty reports I barely read. Yes. Ask me on a Thursday when the agent catches something brilliant and I get to explain it to leadership. No.

That mug. "World's Okayest Analyst." When did it stop being ironic?

Wren: (laughs) About four months ago. When I realized the agent is the analyst now and I'm just okay at watching it work.

Gallup's 2026 global workplace report found engagement at its lowest level since 2020, with only 20% of employees engaged worldwide.⁷ HBR research from February 2026 found that AI tools don't reduce work so much as intensify it, expanding pace and scope in ways that lead to cognitive fatigue and burnout.³ The satisfaction paradox Wren describes, better outcomes paired with diminished meaning, may turn out to be the defining management challenge of the agent era. Or it may be the kind of thing that only matters until the next cohort of analysts arrives having never known anything different.

Wren, for the record, finds that possibility the most unsettling of all.

Footnotes

1. Harvard Business School Institute for Business in Global Society, "What to Watch in 2026," December 2025. https://www.hbs.edu/bigs/bigs-what-to-watch-in-2026 ↩

2. PMC, "Examining the Double-Edged Sword Effect of AI Usage on Work Engagement," February 2025. Research on how AI segmentation of workflows reduces perceived task identity and integrity. https://pmc.ncbi.nlm.nih.gov/articles/PMC11852299/ ↩

3. Harvard Business Review, "AI Doesn't Reduce Work — It Intensifies It," February 2026. https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it ↩ ↩²

4. Gallup State of the Global Workplace 2026 found that employees with reduced autonomy and choice are significantly more likely to be job-searching. https://www.gallup.com/workplace/349484/state-of-the-global-workplace.aspx ↩

5. 365 Data Science reports average U.S. data analyst salary rose to $111,000 in 2026, up from $90,000 in early 2025. https://365datascience.com/career-advice/data-analyst-job-outlook-2025/ ↩

6. Fortune/ILO reporting indicates roughly 40% of unhappy American workers stay in their roles due to economic anxiety. Cited via Locus Robotics workforce analysis, November 2025. https://locusrobotics.com/blog/workforce-predictions-ai ↩

7. Gallup, State of the Global Workplace 2026 Report, April 2026. https://www.gallup.com/workplace/349484/state-of-the-global-workplace.aspx ↩