Tech analyst flags S&P 1500 Software revenue-per-employee spike as AI impact evidence
Carl Quintanilla's chart showing parabolic growth in software sector productivity draws pushback from critics who attribute gains to mass layoffs, not AI value creation.
What to know
- Quintanilla's chart shows S&P 1500 Software revenue per employee spiking sharply upward, which he frames as direct evidence that AI is beginning to produce real economic impact.
- Critics argue the metric reflects mechanical inflation from tech sector mass layoffs (2022–present) rather than AI-driven productivity gains—firing workers inflates the ratio without proving value creation.
- Commenters question the timing, noting the trend predates widespread production AI adoption (mid-2025) and that the chart conflates headcount reduction with innovation.
The dispute Whether the chart demonstrates AI creating economic value (Quintanilla's claim) or merely reflects layoff-driven ratio inflation with misaligned timing relative to actual AI adoption. · positions read across 19 posts and comments
The chart reflects tech layoffs inflating the ratio, not genuine AI-driven productivity gains.
-
“There was *a lot* of over-hiring in the software sector for decades, and the waves of layoffs since 2022 offset AI buildout costs and the net effect is an artificially inflated revenue/employee metric”
datadrivenmd.social · Bluesky ↗
The chart's upward trend predates practical AI adoption, misaligning the narrative with real implementation timelines.
-
“Model harnesses were not a widely adopted thing until mid-2025, which is where 'useful' vibecoding took off. The idea that the linear trend would be the same before and after that point doesn't scan.”
svlch.bsky.social · Bluesky ↗
AI could have genuine productivity potential, but current market dynamics prioritize deals and user bases over product innovation.
-
“I am solo-building a video game in UE5 and I will say, AI in this particular regard has a lot of potential. But, it changes everything. For example, I built my own version of Teams/Slack/Discord. People will realize it is the deals and…”
phoenixsoap.bsky.social · Bluesky ↗
Carl Quintanilla CNBC financial analyst
How it unfolded 4 developments, newest first · click a bar or a number to jump postscomments
-
4
Critic argues layoffs serve as pretext for AI narrative rather than proof
A commenter frames the revenue-per-employee gains as mechanically inflated by headcount reduction, contending that AI is being used as justification for layoffs rather than the chart demonstrating genuine AI-driven value creation.
“If you fire a bunch of people you temporarily make more money per employee without actually improving your product at all. Has nothing to do with AI except using it as a pretext for layoffs.”
— brandongoer.bsky.social -
Right. Turns out if you fire a bunch of people you temporarily make more money per employee without actually improving your product at all. Has nothing to do with AI except using it as a pretext for layoffs.
2 more of the top 3 · 12 posts in this stretch
-
Chart arrow escapes reality. This is a numerator/denominator ratio, and 2023 is exactly when Tech Bros did mass layoffs. Revenue-per-employee spike / headcount plunge = warped chart. That's not "AI creating value," that's "AI narrative justifying layoffs", and the distorted ratio mechanically jumps.
-
They used to think people drove growth, now they think it's GPUs. It might just be the something else.
-
-
3
Skeptics question timing of trend relative to actual AI adoption
Commenters note that the chart's upward trend predates widespread AI implementation in production systems. One notes that model harnesses and practical AI tools did not see adoption until mid-2025, creating a mismatch with the chart's narrative.
“Model harnesses were not a widely adopted thing until mid-2025, which is where 'useful' vibecoding took off. The idea that the linear trend would be the same before and after that point doesn't scan.”
— svlch.bsky.social -
Yeah, I don't buy this. Model harnesses were not a widely adopted thing until mid-2025, which is where "useful" vibecoding took off. The idea that the linear trend would be the same before and after that point doesn't scan.
2 more of the top 3 · 4 posts in this stretch
-
This chart uptic starts prior to any ai implementation in jobs. You nailed it
-
This chart doesn't prove that at all lol
-
-
2
Commenters contest the metric, pointing to tech sector layoffs as driver
Multiple Bluesky users challenge Quintanilla's interpretation, arguing that the spike reflects mass layoffs in the software sector since 2022 rather than AI-driven productivity gains. The revenue-per-employee ratio rises mechanically when headcount drops while revenue stays flat or grows modestly.
“There was *a lot* of over-hiring in the software sector for decades, and the waves of layoffs since 2022 offset AI buildout costs and the net effect is an artificially inflated revenue/employee metric…”
— datadrivenmd.social -
Alternative analysis: there was *a lot* of over-hiring in the software sector for decades, and the waves of layoffs since 2022 offset AI buildout costs and the net effect is an artificially inflated revenue/employee metric
1 more of the top 2 · 2 posts in this stretch
-
Would like another line for total employees...
-
-
1
Quintanilla cites S&P 1500 Software revenue-per-employee spike as AI impact
Carl Quintanilla posts a chart showing S&P 1500 Software revenue per employee trending sharply upward, framing the parabolic rise as direct evidence that AI is beginning to influence the real economy.
“S&P 1500 Software revenue per employee has gone parabolic. If you are looking for evidence that AI is starting to impact the real economy, this is exhibit A.”
— Carl Quintanilla -
C
“.. S&P 1500 Software revenue per employee has gone parabolic. If you are looking for evidence that AI is starting to impact the real economy, this is exhibit A.” (via OpCo desk)
-
What people are saying 8 voices from 1 site · best of 19 · verbatim
- Does the chart show actual total employee headcount decline, or only relative changes?
- How much of the S&P 1500 Software companies' revenue comes from AI products or AI-assisted workflows versus pre-AI business?
- What is the breakdown of revenue-per-employee across the index—is the gain concentrated in a few firms or broad?
- Sep 15
-
Rev/Employee chart where you completely ignore the denominator, great job Carl, you can mislead w/ the best of them. You really have them convinced you're one of us.
-
I am solo-building a video game in UE5 and I will say, AI in this particular regard has a lot of potential. But, it changes everything. For example, I built my own version of Teams/Slack/Discord. People will realize it is the deals and user base, not the software itself that matters for companies
-
I want to see the subtabs. I suspect we are not going to see a normal distribution.
-
Long b4 AI can kill us all, it will replace enough workers & massively reduce aggregate consumer demand so that 1st western digital economies will collapse, then the world economy. Then civil wars, regional wars & police state coercion will produce intensifying & enduring humanitarian catastrophes.
-
Do they mean parabolic in the sense that revenue is now O (number of employees)^2 for a given firm’s head count? Or do they mean anything by it at all, other than “increase; new conditions”? Chart looks like a linear fit to me
-
Turns out a lot of people got paid a lot and produced very little.
-
they're firing employees and replacing eith "ai", they mention how much they're spending on that?
-
Good for tech-heavy retirement portfolios and bad for anyone hoping AI-era tech hiring will look like the hiring booms of the past two decades…