There is a tempting story in a falling vacancy count: fewer employers are hiring, more people are available and recruitment should become easier. It is tidy, intuitive and incomplete.

The ONS’s October release estimated 691,000 UK vacancies in July to September 2026, down 21,000, or 2.9%, from the previous quarter. Yet the CIPD’s summer employer survey found that 31% still had hard-to-fill vacancies. These are different surveys from different periods. Read together, they illustrate the distinction between a labour market with less demand overall and one that can match the available people to the available work. ONS vacancy figures; CIPD summer findings.

For a jobseeker, that distinction changes the question from whether Britain is hiring to where their evidence fits a real requirement. For an employer, it changes the question from how many applications arrived to why suitable people are not reaching—or accepting—the role.

First, understand what a vacancy count can say

The ONS Vacancy Survey estimates the stock of unfilled jobs for which employers are actively seeking recruits from outside their organisation. The headline is a three-month average. It is not the number of people hired during those months, the number of advertisements appearing online or the number of jobs that will be available next month. The ONS methodology guide explains the underlying concepts.

That matters because a post can be filled and replaced by another vacancy while the total stays similar. An organisation can recruit heavily simply to replace departures, without increasing its headcount. Conversely, a business can leave a position unfilled for a long time and contribute to the vacancy stock without making a hire.

The precision deserves attention too. The ONS says the estimated quarterly fall of 21,000 remained within an approximate confidence interval of plus or minus 32,000 vacancies. That is a reason to avoid declaring a sudden turn on one quarterly movement. The broader pattern and other indicators carry more weight. The October bulletin explains the sampling uncertainty and the lag in the headline series.

The largest quarterly decrease by employer size was among businesses with one to nine employees. That does not establish what any particular small firm will do, still less whether a local employer has stopped recruiting. An aggregate can describe pressure while leaving substantial variation underneath it.

Why the advertisements appear to tell another story

In its 26 October Labour Market Tracker, the REC and Lightcast reported about 1.91 million active online job postings in September, 11.8% more than a year earlier. That sounds difficult to reconcile with 691,000 vacancies until the units are made explicit: a count of online postings is not the same measure as the ONS survey’s estimate of vacancies. The tracker is a separate lens on advertised demand.

The figures should therefore not be divided into one another to calculate a supposedly authoritative rate of duplicate or fictitious jobs. Their coverage, methods, reference periods and treatment of advertised opportunities differ. The ONS itself warns that alternative online measures are not directly comparable with its vacancy series.

Nor does an increase in advertisements prove that completed hiring has accelerated by the same amount. To examine hiring activity, it is useful to look at recruiters’ placement evidence as well. September’s KPMG and REC Report on Jobs showed a marginal increase in permanent placements and temporary billings, alongside a continuing, milder fall in demand for staff. The survey covered around 400 recruitment consultancies. Its results describe a cautious improvement in that channel, not a comprehensive census of employment.

The matching problem survives the slowdown

A vacancy is attached to a bundle of conditions: a skill, a place, a working pattern, a wage, a start date and sometimes a regulatory requirement. A person is attached to a bundle of capabilities and constraints. Neither bundle becomes interchangeable because the national unemployment rate changes.

A qualified candidate might be unavailable for the advertised shifts. An applicant from another sector might have most of the necessary skills but be screened out by a requirement for identical experience. An employer might need an accredited specialist whose training takes years. These are illustrative matching problems, not findings about the relative importance of each cause in the national data.

The CIPD’s summer survey, with fieldwork from 24 June to 24 July, gives the issue a measured foothold: hard-to-fill roles persisted even while employers were cautious about hiring. It does not mean every employer faces a shortage, or that every rejected applicant lacks skills. The practical task is to find the specific obstacle.

The payroll figures reinforce the need for restraint. The ONS’s provisional September estimate put UK payrolled employees at 30.3 million, little changed on the month and lower than a year earlier. Payroll counts people on employer payrolls; vacancy counts open positions. Treating one as a direct proxy for the other obscures more than it explains. The October labour market overview brings the measures together.

Search for the bottleneck you can address

For candidates, a useful response is to compare several live descriptions of the work you want. Which requirements appear repeatedly? Which do you already meet but fail to demonstrate? Which are genuinely essential qualifications? Which are preferences worth discussing?

Use that comparison to make a decision. If the same software skill is essential across suitable roles, targeted practice may be worthwhile. If the barrier is a travelling pattern that cannot work, a course will not solve it. If your experience is relevant but sits under another title, make the connection explicit with a concrete example.

For employers, audit the vacancy before adding another filter. Is the salary credible for the responsibilities? Are the hours clear? Could training substitute for a narrowly worded experience requirement? How long does a suitable applicant wait between stages? A larger applicant pool will not repair an offer that the right person cannot accept.

The UK Careers Fair Jobs Board and in-person careers fairs can be used together: the listing supplies detail, while the conversation tests what the detail means. The goal is a better-informed application or a better-defined vacancy.

Britain’s hiring picture is uneven because it contains many different markets. Reading the numbers carefully is the first defence against false hope and unnecessary pessimism. The next step is smaller and more useful: identify the particular work, the particular obstacle and the evidence that could close the gap.

Sources