Your Hiring Pulse report for May 2023

In April’s Hiring Pulse, we talked extensively about AI at work – namely, the latest and potentially greatest destabilizer in the working environment.

And the destabilization could be even greater than COVID-19 in 2020. ChatGPT and all its AI cousins across the board are leading to unprecedented trends in our hiring data.

What’s possible is that this may only be the beginning. Remember March 2020 when those first few COVID-19 numbers started trickling in? That’s the feeling these days.

Let’s have a look at what that means, and stay on for the ride because we have a lot to say at the end.

How we’re looking at data

We’ve adopted two methodologies in how we look at the Hiring Pulse dataset. For Time to Fill and Candidates per Hire, we’re measuring each month using the average of 2019, the last “normal” year, as a baseline index of 100.

For job openings, we’re taking a different route – simply, the average number of job postings per company. This gives us the opportunity to gauge overall recruitment activity and whether that’s going up or down.
Want a more detailed methodology? Jump to the end and check it out.

As always, we look at the worldwide trends for three common SMB hiring metrics:

• Time to Fill (TTF)
• Total Job Openings (JO) 
• Candidates per Hire (CPH)

Let’s start analyzing!

Main highlights

The three main highlights for this month’s Hiring Pulse are:

• Candidates per Hire is at an all-time high for Workable’s hiring data
• Time to Fill is stabilizing – barely
• Job activity is dropping across the board

1. Time to Fill

For this report, Workable defines “Time to Fill” as the number of days from when a new job is opened to when that job opening is filled. It’s important to understand that definition: jobs that are still open as of the end of April are not included in this graph as they don’t yet have an “end date”. Only the jobs that are filled are included here.

Quick clarification, because people are asking: the data in this chart shows the trendline against the 2019 average as an index of 100, not the actual number of days in TTF.

Got that? Good. Let’s have a look at the monthly TTF trend through to the end of April against the average of 2019, based on jobs that have been filled:

After a pretty significant drop in the Time to Fill trend for the first three months of 2023, we’re finally seeing that metric relatively stabilizing to start the second quarter of the year.

In fact, the number has jumped upwards ever so slightly, with the trend jumping 1.1 points from 81.6 in March to 82.7 in April.

It’s still well below the general trend going back to 2020, an indicator of factors such as technology speeding up the evaluation process, more candidates in the talent bloodstream (more on that below), and – perhaps – a desperate rush to fill roles as a stopgap measure in times of high turnover.

On that latter point – there is plenty of labor instability right now. There are reorgs, layoffs, and restructuring all happening on the heels of the Great Resignation (which, while still high, is starting to level off and come down in terms of raw numbers). And a lot is happening in the age of AI as well.

What this means is, in other words, bottlenecks and breakdowns are happening, forcing businesses to move quickly to plug gaps in their workflows.

It’s one explanation, at least. Let’s look at total job openings and see for ourselves if there’s increased job activity across the board.

2. Total Job Openings

Total job openings represent the total number of job openings activated across the entire Workable network.

As stated above, we’re displaying this as an average of job postings per company in the network. And because this is not contingent on job opened/filled dates like TTF and Candidates per Hire, we can simply look at the raw job open numbers up to the end of April.

After a nice levitation in job openings to start the first quarter of this year, we’re finally seeing things tapering off across all three business size buckets. In fact, job activity is down from 7.6 job postings per company in March to 6.6 per company in April.

Overall, what we see is a full point drop from March to April in the average number of jobs posted per company. That’s significant, in no small part due to it being the biggest month-to-month drop in the history of our network data.

We’ve seen nice jumps in the data from one month to the next (especially from the typically slow December to a supercharged January), but we’ve never seen anything quite so dramatic the opposite way.

We would go into depth into each of the three size buckets, but they all see the very same trend for March to April – so we’ll skip that for this month. Instead, we’ll simply point out that the impressively dynamic small business category (1-50 full-time employees) was the usual anomaly in terms of job activity, with five straight month-to-month increases in the trend.

But now, small businesses also took a dip in April. So it can no longer hold itself up as an agile upstart. Last month, we promised to keep an eye on this area – now we’re going to continue watching and see what May brings us.

Now, on to the candidates.

3. Candidates per Hire

Workable defines the number of candidates per hire (CPH) as, succinctly, the number of applicants for a job up to the point of that job being filled. Again, remember, this is a trendline using the 2019 CPH average as a baseline of 100, not the actual number of candidates per hire.

Now that Let’s look at what’s going on here through April:

Um. That chart says it all.

After a momentary stabilization in the CPH trend from February to March, April stands out like a very, very sore thumb, hitting a new milestone of 158.2 for Apil – a huge 21-point jump from March.

To put that in perspective, the biggest jump in the CPH trend in our entire history of data was 18.9 points from February to March back in 2020. The year where the world seemingly changed and the sheer volume of job loss felt unprecedented (for this generation, at least).

And the highest CPH trend with the exception of February’s 138.8 and March’s 137.2 was in the high 120s and very low 130s from mid-2020 to early 2021.

And now? 158.2. To put it in visual perspective, look at the chart from January 2020 onwards:

Let’s go back to a quick quote from last month’s Hiring Pulse:

“We discussed the Great Resurgence in [February]’s Hiring Pulse – that’s still happening, of course, but the candidate pool is not a bottomless one. Are we finally reaching the crux of this data point? Or is this just a hiccup and more are on the way? We shall see.”

Well, we are seeing now that the bottom of the candidate pool isn’t yet discovered. It may be at a murky depth not unlike Lake Baikal in Siberia, known as the world’s deepest lake with a bottom that’s ​​5,315 feet (1,620 meters) deep.

We don’t really want to talk about Mariana Trench at 36,201 feet (11,034 meters) because that’s uncomfortable to think about and we can’t predict whether or not we’ve hit a certain limit in terms of depth and breadth of the candidate pool.

Instead, should we just try and understand what’s going on here?

What’s going on here?

We talked a lot about AI last month. It’s still very relevant now and will continue to be so going forward. And it is absolutely impacting the working world in myriad ways. Our day-to-day is affected, and our hiring processes are changing, and above all – jobs are ultimately changing.

Consider this – according to Goldman Sachs, 300 million jobs worldwide could be affected by this new tsunami of generative AI that started with ChatGPT in December.

Many other companies are actively encouraging the use of generative AI technologies in their working environment – including one CEO who has purchased ChatGPT licenses for his entire staff base to the tune of $2,400 a month. For the record, that CEO says productivity has gone through the roof.

Others, like IBM, are phasing out some jobs altogether – to the potential tune of 30% of non-customer-facing roles – as a result of increasing AI capabilities.

Meanwhile, the Biden administration called together the CEOs of Alphabet, Microsoft, OpenAI and other AI-driven companies to discuss the potential risks and opportunities of the new technology.

There is a lot more going on, of course, but at the core of all of this is jobs. When we started 2023, people weren’t really talking about artificial intelligence beyond how cool ChatGPT seemed to be. Some early adopters were taking on ChatGPT to help in their work, but overall, generative AI wasn’t really in the everyday lexicon. Instead, talk of a recession was.

Now, we don’t see a lot of talk about an impending recession. Is there even one happening? Who knows? What we do know now is the tremendous rise of AI and jobs in everything we’re talking about:

Our hiring data is starting to show it. Time to Fill is getting shorter – potentially because of the use of generative AI tech in the hiring process. Job openings are dropping – not because of a recession, but because some jobs are becoming redundant and companies are figuring out how to get more done with fewer people.

And finally, candidates per hire is surging – perhaps due to the double whammy of layoffs ahead of a (possible / speculated / who knows) recession and the rise of AI technologies in workflows.

Does this mean AI is coming for your job as well? Not necessarily. One saying that’s making its rounds is this one: “AI will not take your job. People who use AI will.” But humans are still at the center of it all – AI is a great enhancer to your work, not a great replacer. And the human touch is still paramount in hiring.

If humans weren’t important, then why do we still see lineups in banks for that more personable service and in supermarkets with checkout cashiers? Why do we grumble about pressing ‘1’ to do this and ‘2’ to do that when trying to get service on the phone?

And so on.

Things may change if (or when) AI gets to a point where it becomes general intelligence, but right now – we are still the drivers.

For how long, though? Let’s keep watching this space – the data is still telling us a lot.

Thoughts, comments, disagreements? Send them to [email protected], with “Hiring Pulse” in the subject heading. We’ll share the best feedback in an upcoming report. Watch for our next Hiring Pulse in May!

The Hiring Pulse: Methodology

Because one of the three metrics (Job Openings) is different from the other two metrics (Time to Fill and Candidates per Hire), we’re adopting two very distinct methodologies.

To bring the best insights to small and medium (and enterprise-level) businesses worldwide, here’s what we’re doing with the Job Openings metric: we’re taking the number of job openings in a given month and dividing that by the number of active companies in our dataset, and posting that as an average. For example, if July 2022 shows the average Job Openings per company as 7.7, that simply means each company posted an average of 7.7 jobs that month.

For the Time to Fill and Candidates per Hire metrics, we’re comparing a specific month’s trend against the full average of 2019, and we show the result using that 2019 average as a baseline index of 100. For example, if July 2022 shows an average Time to Fill of 30 days for all jobs, and the monthly average for all of 2019 is 28, we present the result for July 2022 as 107.1 – in other words, 7.1% higher than the average of 2019.

And we chose 2019 as the baseline because, frankly, that’s the last normal year before the pandemic started to present challenges to data analysis among other things.

The majority of the data is sourced from businesses across the Workable network, making it a powerful resource for SMBs when planning their own hiring strategy.