Treat your 2027 entry-level hires as their own prosperity row and not part of general headcount. The finest current payroll data shows that occupation in AI-exposed jobs is shifting, alongside small younger hires fairly than layoffs. Most promotion prosperity discussions are looking location else.
In July, I asserted that chief promotion officers should prosperity five functions alternatively of bequest conduit buckets. In September, I asserted that SEO and paid media managers rotate into board by proving AI pays for itself during keeping their teams intact. Stanford’s Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen, and Kevin Indig’s October preparedness memo, assistance parts of the two arguments.
They additionally disclose a gap.
What Stanford Measured, And What It Can’t Tell You
The Stanford Digital Economy Lab operating paper, revised Aug. 12, 2026, draws on anonymized ADP payroll records. The inspection example is a stable panel of firms alongside between 3.5 and 5 myriad workforce per month, operating through June 2026.
The header finding is that occupation of workers old 22 to 25 in AI-exposed occupations stands 19% below anywhere it would be if it had kept gait alongside their less-exposed peers. In raw terms, the two most exposed quintiles of young workers cut concerning 11% since November 2022, during the three least-exposed quintiles grew concerning 10%.
Experienced workers display no comparable gap. Every older age collection grew complete the identical period, by as much as 11% for workers old 35 to 40.
The scheme matters additional than the headline. The authors say the divergence plant chiefly through reduced hiring of young workers, and separations among young workers cut at smallest as much in the most exposed occupations as in the smallest exposed. That is the contrary of what relocation would imply. The declines concentration anywhere AI substitutes for individual tasks. Where AI complements workers, occupation is flat or rising, particularly for informed workers, and the adaptation shows up in employment, not in pay.
The authors of the Stanford document call the facts “early, descriptive indicators,” and say the activity does not evaluation a causal effect of AI. The form shrinks whenever they authority for education, from concerning -0.18 to -0.09 in their chief sample, and the smaller fig is important lone at the 10% level. Earlier versions of the document reported regression estimates of 13% and 16%, and the header measure has since changed, so the 19% should not be lined up against those numbers.
The document additionally measures occupations broadly. It reports no SEO or promotion figures.
My July Column Holds Up, With 1 Missing Line
In my former article, 2027 Marketing Budgets, I told CMOs to rebuild the prosperity about five functions, one of them individual judgement and editorial oversight. I additionally cited Gartner data showing labor rising from a average 21.9% to 24.5% of promotion budgets, equal although 43% of CMOs expected to cut labor spending.
Stanford’s document supports the editorial oversight prosperity line. Where AI complements workers, occupation holds or grows, particularly for informed workers. That describes the editors and strategists I asserted CMOs have to be capable to defend.
What the pillar did not say is who that labor prosperity pays for. My perspective is that the additional labor portion is going to informed people, during the seats that would have gone to 22- to 25-year-olds in exposed jobs are quietly remaining empty. Gartner doesn’t divided promotion labor by seniority, and Stanford doesn’t interrupt out marketing, so I can’t demonstrate it. But a prosperity that raises labor portion during freezing entry-level hiring has not told the CFO what it is funding, and I think that omission volition be the archetypal item a keen backing chief finds.
See also: How To Restructure Your Marketing Team & Budget For The AI-Search Era
My September Column Needs A Sharper Test
In my another part concerning how an SEO or paid director can rotate into a director, I stated the managers most apt to move up are the ones who can demonstrate AI pays for itself during keeping their teams intact.
The Stanford document finds the adaptation happens through hiring and not through layoffs. A squad can be completely intact on the day a director presents the scorecard and motionless be shrinking, since the younger place that would have been filled this twelvemonth never was. “Intact” tells you who is motionless there. It says nothing concerning who was NOT hired.
If the entry-level pipeline narrows, the pond for the next generation of managers narrows alongside it. That is my inference, not item the document tests, and it is the part of this narrative I would observe most closely.
One additional reconciliation. The Harvard Business School study I cited in September looked at job postings through March 2025, so it is a petition signal. Stanford measures realized occupation through June 2026. They item the identical direction, but I would not add the 13% and the 19% together.
Kevin Indig Gives The Gap Somewhere To Land
Kevin Indig’s Growth Memo on preparedness 2027, summarized in a LinkedIn post, starts from a premise I share. Salaries, tools, and infrastructure are fixed costs, so the genuine prosperity inquiry is which capabilities to fund. Proven activity specified as specialized SEO, satisfied refreshes, and digital PR that earns citations gets complete funding, during unmeasurable bets get their own row according to judgment.
The memo goes additional than the post. It adds a capped bucket for learning work, alongside one owner and a decision date, and argues that community activity needs a named owner and fixed weekly hours or it gets cut first. It suggests approximately 5% of the promotion prosperity for community, sometimes 10%.
His three measure layers additionally inhabit a gap in my September scorecard, which tracked hours saved, hours returned, outcomes, and cost. Kevin says to set targets for “exposure (AI citations, mentions, portion of voice), behavior (branded search, straightforward visits), and endeavor outcomes (pipeline),” and to peruse them together. If visibility rises during tagged hunt stays flat, he says, you have a allocation problem.
Put Stanford and Kevin side-by-side and a younger function looks different. The tasks I listed in September as most exposed, recurring reports, bulk heading tags, and bid adjustments, are the ones juniors used to study on. Those are my inferences concerning SEO, and Stanford did not test them. Community work, digital PR, and tracking anywhere AI engines cite you are the activity Kevin says needs an owner and hours, and they are much harder to hand to a model.
My stance is that a younger place built about that activity belongs in Kevin’s learning bucket, alongside a named owner, set hours, and a decision date. That is a stronger case to a CFO than keeping a place since it has continually existed.
See also: Why Digital PR Fundamentals Beat AI Tactics: What Marketing Leaders Should Focus On
How To Use This In Your 2027 Plan
Sort all younger project by what AI does to it. List what your 22- to 25-year-olds really do all week. Mark all project as one a example can substitute for or one that depends on judgement and relationships. Stanford did not study SEO, so run your own check, equal a duration of uncomplicated period tracking. It repeats the re-tagging stage from my July column, lone at the project level.
Write the entry-level function into the learning bucket. Give it a named owner, fixed weekly hours for community, off-site mentions and citation tracking, a quarterly review, and a decision date. If the function cannot name its weekly output in those terms, it is a place you are keeping out of habit.
Put a hiring row on the scorecard. Next to hours saved, hours returned, outcomes, and cost, display the seats you chose not to inhabit and why. Report Kevin’s three layers monthly beside it. When the CFO asks why labor portion is up, you can display what it buys and what it does not.
Where I Land
I went into this exercise assured that my July and September columns were broadly right, and they are. What I got incorrect was how small they stated concerning the group who never appear on a headcount report.
My stance for 2027 is that an intact squad is the incorrect approach. A squad you are motionless training is the improved measure. If AI pays for itself in your shop, several of the funds should display up as a younger place alongside a genuine job attached. Stanford calls its own data early, and so do I. But a prosperity approved this autumn volition assistance decide who is in the area whenever the next board are chosen, and I would fairly contend concerning that now than detect it later.
More Resources:
- The SEO Skills Gap: Why Technical Expertise Alone Won’t Cut It Anymore
- Harvard Found The Public Has Little Objection To AI Taking Search Marketers’ Jobs
- Using AI To Assist With SEO Work, Not Replace The Worker
Featured Image: Anton Vierietin/Shutterstock
