When is the best time of year to look for a job? The surge is mostly in the applicants

It is September, and a video has told you the jobs are about to arrive. Or it is late November, and a friend has told you to stop until January, because nobody hires in December. Both pieces of advice describe a season. Neither says what the season is made of, or who is in it with you.

The monthly figures that answer that exist, and almost nobody looks at them, because the numbers in the news have had the seasons removed. Put them back and the calendar looks different from the folklore: the swing is real, it is mostly in other people's industries, and the crowd that arrives in January is applicants, not adverts.

What is being counted

Three different things get called "hiring". The US Bureau of Labor Statistics separates them. A job opening is a position that is unfilled "on the last business day of the month", where the work exists, "the job could start within 30 days" and the employer is actively recruiting outside its own walls. A hire is any addition to the payroll during the month, counted when the person starts, not when the offer is made.¹ A posting, the thing you see on a job board, is a third measure that belongs to whoever runs the board.

Every headline version of these numbers is seasonally adjusted, which means the statisticians estimate the usual pattern for each month and remove it, so that a fall in December can be compared with a rise in April. That is the right thing to do for the economy and the wrong thing for you, because the usual pattern for the month is exactly what you want to know. So everything below uses the seasonal pattern the statisticians estimate and remove: for each month, the unadjusted figure divided by the seasonally adjusted one, averaged over the seven ordinary years from 2017 to 2019 and 2022 to 2025, with 2020 and 2021 left out for the obvious reason and 2026 left out because it is not finished, and rescaled so that a typical month is 100. Dividing rather than averaging raw levels keeps each year's overall level, and its rise or fall within the year, out of the pattern.

What each month looks like in the United States

All four columns are indices with the annual mean at 100; the last is the unemployed factor divided by the openings factor.²³⁴

Start with hires, because a hire is the thing you want, with one caveat: the count is every addition to a payroll, and the BLS definition includes "rehired employees", "seasonal employees", "transfers from other locations" and employees "recalled to a job at the sampled establishment following a formal layoff lasting more than 7 days".¹ A hire is not always a job won from outside, and the count cannot say how many were. December is the quietest month by a distance, at 74, and it is not close: in 2025, employers added 3.9 million people to payrolls in December against 6.1 million in June.² November and February are the next two troughs. Then April to August all sit above 106, with June at the top at 115. September, the month with the surge named after it, is 101. October is a modest 106.

Openings move less, because an opening is a stock and a hire is a flow. The last-business-day count peaks in April at 108, has a second peak in July, eases through late summer, recovers to 105 in October and then falls through November to a December low of 92.³ In 2025 there were 6.1 million unfilled positions on the last business day of December; by the end of April 2026 there were 8.2 million.³ In the averaged pattern the range from the December trough to the April peak is about a sixth; the 2025 to 2026 gap is wider because the market itself was recovering. That is the whole seasonal range in the official count of unfilled jobs, with the trough in December and the peak in spring.

So two of the three sayings are half true. December really is the low. January really is a recovery: hires climb from 74 to 97 and openings from 92 to 100, with the rest of the climb coming in the spring. And the September surge, in these numbers, is not visible at all. Hires in September sit at the annual average, and so do openings. LinkedIn's data, as reported by Fortune this month, show postings on LinkedIn rising to 14 per cent above March levels in September and 11 per cent above in October, and LinkedIn's Kory Kantenga says there are more postings on the platform in September "than you do any other time during the year".⁹ That may well be true of LinkedIn. Indeed's Cory Stahle, in the same report, says of Indeed's data that "it's not typically a very large bump", and on Indeed's own US index the September seasonal factor is about 1 per cent above the annual mean.⁹⁶ In the government's count of actual hires, September is an ordinary month.

Where the seasons actually live

The averages hide the more useful fact, which is that the seasons belong to particular industries.⁵

Construction hires run at 129 in April and 124 in May, then 55 in December. Leisure and hospitality peak at 128 in May and 127 in June for the summer season. Retail has its own calendar, with hires at 121 in October and 118 in November for the holidays and 78 in December once the shops are staffed. And the September surge, as far as the payroll data can find one, is the school year: hires in state and local education hit 276 in August and 155 in September, against 52 in April. Indeed's own year-end seasonal postings, which its economists tracked in November 2025, were driving jobs up 153 per cent on the year and loading and stocking up 49 per cent.⁸ When a headline says seasonal hiring is strong, that is what it means.

Office work swings too, and for two of its three big categories by about as much as the total, but on a different calendar. Professional and business services, the broadest white-collar category, moves between 78 in December and 111 in April, with February at 90 and September at 97. Financial activities runs from 73 in December to 115 in July, with April and May at 114. Information, which includes software publishing, runs from 67 in December to 113 in June and again in October. All three have something the seasonal industries do not: January is a real month for them, at 104, 109 and 110 respectively. Hires are counted when people arrive on the payroll, and for office work January arrivals are at or above the annual average. Unfilled positions agree: on the last business day of January, openings in professional and business services stand at 104 and in financial activities at 110, both above their annual average and well above December.⁵

So the swing is there for office jobs, narrower for professional services and about as wide as the total for finance and information, and it is a spring and summer peak and a December trough rather than a September one. A job seeker in finance who waits from June until September is waiting for a month that is worse, not better, on the average of the years here.

The other side of the counter

None of this is what makes January feel like hiring season. What makes January feel like hiring season is that everyone else has also decided to look.

Indeed's economists measured it directly this year. Searches on Indeed were "up to 31% higher in January 2026, compared to the early-December 2025 average", while "job postings in January remained roughly the same as in late 2025".⁷ The sectors that had hired for the holidays, retail, driving and logistics, were the ones whose postings fell most.⁷ LinkedIn finds the same shape on its side: applications on the platform peak between January and May and then decline for most of the rest of the year, whatever postings do in the autumn.⁹

The official figures agree. The count of unemployed people in the US is at its high in January, at 110 on the same scale, stays elevated in February at 107, dips to 91 in April, climbs back to 108 in July, and falls through the autumn to 94.⁴ Divide that count by the number of openings in the same month and you get a crude gauge of competition: how many unemployed people there are for each unfilled job. Rescaled the same way, it runs at 109 in January and 112 in February, falls to 85 in April, sits at 95 in September and reaches its second low, 89, in October.⁴ On that measure there are roughly a quarter to a third more unemployed people per opening in the first two months of the year than in April or October.

Two limits. Unemployed people are not the only applicants; a large share of applications come from people with jobs, whom this ratio cannot see, and the numbers on how many applications a single advert draws are measured differently. And an opening in the official count is not necessarily a job the employer intends to fill this month. But the two measures that can see applicants directly both rise at the turn of the year: searches on Indeed in January against early December, and applications on LinkedIn, which peak between January and May. The unemployed count rises then too. That is as far as this evidence goes: on every measure available, the applicant side fills up in the first months of the year, which is exactly when the folklore tells you to join it.

That is also the honest case for the autumn. Kantenga's version, from the Fortune report, is worth keeping: "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."⁹ The unadjusted quits data fit that timing: resignations peak in August, at 125 on the annual scale, and LinkedIn's analysis puts job transitions at their peak between July and September.⁵⁹ What the quits series cannot say is where those people went.

Outside the United States

Indeed publishes its job postings index for several countries in two forms, seasonally adjusted and not. Both are index levels, with 1 February 2020 set to 100, so dividing the unadjusted level by the adjusted one for the same day gives Indeed's own estimate of that day's seasonal effect. Averaged over 2022 to 2025, with the annual mean at 100, the pattern for the stock of all postings is mild everywhere.⁶ In the United States it stays within about 3 per cent of the mean all year, lowest in January at 96.7 and highest in October at 101.6. In the United Kingdom the range is wider: 94 in both December and January, rising through the summer to 104 in September and 105 in October. Canada peaks in September at 105 and bottoms in January at 92. Australia is highest in October and November, at 105, and lowest in January. Ireland peaks in September and October at 104. France peaks in June and July at 105 or 106 and is lowest in January. Germany barely moves, staying between 97 and 102 with no clear season.⁶

New postings, those on Indeed for a week or less, swing far more, because they are the flow rather than the stock. In the UK new postings run at 81 in December and 91 in January, then jump to 104 in February and stay above 100 through to September, which at 105 is the year's high. Canada and Australia show the same December collapse, to 80 and 85, and the same September peak, at 108 and 109. France has August at 89, for the holidays, and September at 106. Germany's January is 89 and its February 107.⁶

What that says, for anyone outside the US, is that the year-end pause in new adverts is deepest in the UK, Canada and Australia, at 15 to 20 per cent below the mean in December, while Germany's trough is January and France's is August; that the January rebound in new adverts, where there is one, arrives in February; and that the September rise in postings that is hard to find in US payroll data is clearer on Indeed's boards in the UK, Canada, Australia and Ireland. It is a rise of about 5 per cent in the stock and about 5 to 10 per cent in the flow of new adverts, not a floodgate.

What to do with a calendar

Do not stop in November and December. Openings are at their lowest then, and by the official gauge the queue per opening is a little shorter than it will be in January and February.⁴ What the data show next is that January is a normal hiring month for office work, with hires at or above the annual average and unfilled positions in professional and business services and in finance above their annual average on the last business day of the month.⁵ A search that runs through December is in front of the employer when that happens.

Do not wait for a month. The whole seasonal range in unfilled jobs is about a sixth, trough to peak, and in the stock of postings on Indeed it runs from about 5 points in the United States to about 14 in Canada, trough to peak.³⁶ The difference between your industry's calendar and the average is bigger than the difference between any two months of the average. If you are in construction or hospitality, spring is your season. If you are in education, it is August. If you are in an office, it is spring and summer: April is the high month for professional and business services, July for finance and June for information, and January is fine in all three.

Treat January as the crowded month, not the open one. The applicants arrive in January. Where Indeed's files record the flow of new adverts, the January low gives way in February in the UK, Germany, France and Australia and in March in Canada; in the US, where the files stop short of that flow, the count of unfilled positions is back above its annual average only by the last business day of January.³⁶ If your search has to start in the new year, the data give a modest case for starting it in the first week of December instead.

Keep looking through the summer and the autumn. Applications fall from May onwards while hires stay high through August, and the two months with the fewest unemployed people per opening are April and October.⁴⁹ The months the folklore skips are not the weak ones.

The seasons are real, and they are small, and they are mostly not yours. The thing that changes month to month is less the number of jobs than the number of people looking at them, which is a reason to be looking when they are not. A search that runs every day does not need a theory of the calendar at all, and if you would rather not check the boards through the quiet weeks yourself, JobCraftly's daily agent will search on your behalf and email you the roles worth a look, whatever the month says.

¹: US Bureau of Labor Statistics, Job Openings and Labor Turnover Survey: definitions. Accessed 10 September 2026. ²: US Bureau of Labor Statistics, Hires, total nonfarm, not seasonally adjusted, JOLTS series JTU000000000000000HIL. Month indices computed by JobCraftly as the unadjusted series divided by the seasonally adjusted series (JTS000000000000000HIL), averaged by month over 2017–2019 and 2022–2025; method in the source notes. Accessed 10 September 2026. ³: US Bureau of Labor Statistics, Job openings, total nonfarm, not seasonally adjusted, JOLTS series JTU000000000000000JOL. Indexed by the same method, against the seasonally adjusted series JTS000000000000000JOL. Accessed 10 September 2026. ⁴: US Bureau of Labor Statistics, Unemployment level, not seasonally adjusted, CPS series LNU03000000. Indexed by the same method, against the seasonally adjusted series LNS13000000; October 2025 is missing because of the lapse in appropriations, so that month rests on six years. The last column of the table divides the unemployed factor by the openings factor and rescales to a mean of 100. Accessed 10 September 2026. ⁵: US Bureau of Labor Statistics, Table 7. Hires levels and rates by industry and region, not seasonally adjusted, JOLTS news release; the industry indices use the corresponding JOLTS series, listed in the source notes, over the same seven years and indexed by the same method. Accessed 10 September 2026. ⁶: Indeed Hiring Lab, Indeed Job Postings Index, data repository, files to 4 September 2026. Seasonal factors computed by JobCraftly as the daily ratio of the unadjusted to the adjusted index, averaged by month over 2022–2025; method in the source notes. Accessed 10 September 2026. ⁷: Laura Ullrich and Sneha Puri, February 2026 US Labor Market Update: New Year, Same Resolutions, Indeed Hiring Lab, 19 February 2026. Accessed 10 September 2026. ⁸: Cory Stahle, Seasonal Hiring Picks Up, but Signs of Hesitance Remain, Indeed Hiring Lab, 25 November 2025. Accessed 10 September 2026. ⁹: Tatiana Sataua, Job seekers are waiting for the 'September Surge.' Economists say there's something behind the hype and explain why it's a good time to apply, Fortune, 5 September 2026. Accessed 10 September 2026.