Invest1 distinct publisher3 min readPublished
Postings in the US run 14% above March in September, and applications do not follow. Employers are bidding into the thinnest applicant flow of the year in a month when payrolls grew 162,000 and revisions added another 55,000.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
Postings are not hires, and the gap between the two explains the September pattern. LinkedIn's own analysis puts hiring and job transitions at their peak between July and September, with a sharp December fall and a January rebound partly made up of people who accepted offers in the closing months of the prior year and delayed their start dates [13]. The mechanism Kantenga points to is finance, accounting and other professional-services firms that recruit in September and October for workers who may not actually start until the following summer [12]. So a share of the September posting count is a forward booking of next year's headcount rather than a bid for anyone available in October.
The arithmetic on the window itself is straightforward. US postings sit 3% below March in August and 14% above March in September [5], which is a rise of about 17.5% inside a month once you divide 1.14 by 0.97 [16], and October gives back roughly 2.6% of it [17]. Applications, by LinkedIn's account, peak between January and May and decline through much of the rest of the year, with no autumn rise even in the countries where postings turn back up [6]. Hold application volume flat and that same posting arithmetic leaves roughly 15% fewer applications per opening in September than in August [18]. Kantenga frames the seasonality from the candidate's side, arguing that five jobs and one applicant is still not a bad position to be in [7]. From the employer's side, September is the month with the most competition for the same candidate.
Two things could undo that read. Cory Stahle of Indeed Hiring Lab says it is not typically a very large bump, and some years show only a small September uptick [9], which makes the 17.5% a seasonal average an individual year can miss entirely. And Indeed attributes the Labor Day ramp to employers preparing for the fourth quarter and the holidays, with the added demand concentrated in retail and in transportation and warehousing [8], so the aggregate index may be measuring a labour market with little bearing on whoever is trying to fill a controller seat. Accounting shows how far the sector calendar can diverge from the national one: Indeed's data have accounting postings jumping roughly 21% from July to August last year, ahead of year-end reporting and tax season [11].
Stahle points to a duller explanation that may matter most for anyone actually scheduling interviews: hiring managers and HR staff take holidays over the summer, which slows the interview process, and September and October land after that slowdown but before holiday scheduling gets difficult again [10]. That makes September a throughput story rather than a supply story, and throughput is a constraint a hiring team controls.
The underlying labor market stayed modest through all of this. LinkedIn's hiring rate rose 2% from July to August [14], unemployment held at 4.1% [3], and Friday's report put 217,000 jobs on the books once the 55,000 of June and July revisions are counted alongside August's 162,000 [15]. Employers posting this month are still competing for a thin pool of applicants in that same modest market, not a booming one.
Ranked by verification strength, evidence, and original report placement.
US employers added 162,000 jobs in August, according to the Bureau of Labor Statistics.
Estimates for June and July were revised up by a combined 55,000 jobs.
A 2025 LinkedIn Economic Graph analysis provided to Fortune shows US job postings in August dip 3% below March levels, then rise to 14% above March levels in September and 11% above in October.
LinkedIn found applications typically peak between January and May before declining through much of the rest of the year, and that even in countries where postings rise again in the fall, applications generally do not show the same increase.
Kantenga said: "If there are only five jobs available, but you're the only person looking, that's still not a bad position to be in, assuming that you qualify for one of those roles."
Distinct publishers with included, body-backed reporting in this cluster.
Follow any of these and your For You feed starts watching them — no settings page required.
invest
The 4.1% jobless rate is now a composition story, not a stability story1 distinct publisher
product
Hiring teams buried in AI applications are asking for friction, not better filters1 distinct publisher
invest
Goldin's birthrate paper names a variable that no budget line can buy1 distinct publisher
product
Rillet's $100M reads as proof mid-market ERP is rip-and-replace, mostly at the cheap end1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Checkable payrolls, unpublished vendor curves
The payroll spine is verifiable by anyone: 162,000 for August, 55,000 in revisions, unemployment at 4.1%, all from a public BLS release. The seasonality spine is not. LinkedIn's 2025 Economic Graph analysis was provided to Fortune rather than published, and its numbers are ratios to March instead of counts, so a reader cannot recover how many jobs the September peak represents. Indeed's index is public, but the 21% accounting jump reaches us through its economist's summary of it, and both datasets describe platform activity rather than the labour market.
No behaviour measured
Nothing here counts anyone acting on the September calendar. Postings and applications are volumes, and the closest thing to behaviour, LinkedIn's 2% rise in its hiring rate from July to August, describes the market rather than uptake of the seasonal theory. Employers stating that they time requisitions to the autumn window, or job seekers timing applications to it, would be the missing measurement.
Label outruns the numbers, and Fortune says so
The "September Surge" name promises more than the data delivers, and this reporting undercuts it in its own paragraphs: Stahle's "not typically a very large bump" and a 2% move in LinkedIn's hiring rate sit beside the surge framing, with hiring still more than 20% below pre-pandemic levels. The finding that survives is narrower than the label, namely that postings rise while applications do not, thinning competition per opening. The gap stays modest because the counterweights are in the story rather than absent from it.
Both datasets sold by recruiters
Every seasonality number in this story comes from a company that makes money when people post and apply. LinkedIn's analysis was provided to Fortune, and Kantenga's supporting note went out the same Friday as the payroll report; Indeed's figure arrives through its Hiring Lab economist. The conclusion that September is a good moment to apply routes traffic to both platforms, which is worth holding even though Stahle is the one talking the effect down. Fortune has been covering the "September Surge" since 2023 and has its own franchise in keeping the frame alive.
One outlet, one unresolved contradiction
A single publisher, two vendor datasets, and a conflict left standing: Kantenga says September carries more postings than any other month while the same analysis puts the peak in spring around May. The arithmetic follows cleanly from the published percentages, and the BLS figures are solid. An independent read of the seasonal curve, from anyone without a recruiting business, would move this assessment more than any other input.