How Long Do Sub Jobs Actually Last? We Tracked 82 Districts

Last week we wrote about when sub jobs get posted and promised a closer look at how long they survive once they appear. Here it is — and it did not say what we expected.

We tracked about 5,600 Frontline postings across 82 districts, timestamping each one when it appeared on the job board and again when it disappeared. Every district in this analysis has at least 15 tracked postings.

The range is enormous

In the fastest districts, the median posting is gone in under a minute. In the slowest, the median posting is still sitting there three days later.

That is the same software, the same job category, the same week. The gap between the fastest and slowest district we measured is roughly five thousand to one.

Sorted into three groups by how long the typical posting lasts:

• Seconds districts (median under 2 minutes) — 41% of those we measured. The board turns over while you're looking at it.
• Minutes districts (2 to 15 minutes) — 26%. Postings survive a coffee break, not a class period.
• Hours districts (over 15 minutes, often far longer) — 33%. The board looks calm at any hour.

Across all postings regardless of district, about a third are gone within a minute and roughly two thirds within two minutes.

Resist the obvious conclusion here — we reached for it too, and the data took it away from us. Remember which group your district is in; the rest of this post is about why it matters far less than it appears.

District size tells you nothing

The obvious guess is that big districts are competitive and small ones are relaxed. That guess is wrong. When we group districts by posting volume, the mix of Seconds, Minutes and Hours districts is about the same in every group.

The two highest-volume districts in our data — one a large Florida county system, one a Central Valley district in California — are both Seconds districts. But a large Bay Area district posting nearly as many jobs is an Hours district, where the typical posting sits for well over a day. Volume is not competition.

The part that surprised us

Here is where we expected to write "so find out which kind of district you're in." The data would not let us.

We also track how often our own auto-accept wins a job it goes after. In one large Bay Area district, the typical posting sits unclaimed for more than 40 hours — and we still lose most of the jobs we chase there.

That looks impossible until you ask which jobs we were chasing. In that same district, the typical posting lasts 41 hours, but the postings our users' filters actually matched had a median lifetime of 35 seconds. The average job on that board lasts roughly four thousand times longer than the ones anybody actually wanted.

Pooled across every district we measured, what predicts whether a job gets caught isn't the district at all. It's how long that particular posting lasted:

• Postings gone in under a minute — caught 22% of the time.
• Postings lasting more than an hour — caught 78% of the time.

Same software, same users, same filters. The only thing that changed was how contested that individual posting was.

For the contested ones, how fast you react is close to the whole story. Among people acting on an alert by hand, those who tap within 30 seconds win nearly four times as often as those who take a minute — and the drop-off after that is steeper still.

The average job is not the job you want

A district's median posting is, almost by definition, an unremarkable one: a date three weeks out, a school across the county, a grade or subject most subs on the roster don't want. Those postings linger, and there are a lot of them, so they set the median.

The jobs people actually want — tomorrow, nearby, a grade you like — are a small slice of that pile, and they move far faster than the median suggests. A relaxed-looking job board is often a large stack of postings nobody is racing for, sitting on top of a handful that disappear in seconds.

So the honest finding is not "some districts are hard." It's this: the average job in your district might sit for two days, and the job you actually want is still gone in under a minute. They are not the same job.

Which also explains a frustration we hear a lot: "my district isn't even that busy, so why do I never get anything?" Now you know. The board looking calm tells you nothing about the three postings you'd have said yes to.

What this actually means for you

The practical takeaway is smaller and more useful than a district ranking:

• Don't judge your odds by how full the job board looks. A board with plenty of open jobs is not evidence that the good ones are catchable.
• Speed only matters for jobs you'd accept. Being fast on everything is exhausting and pointless. Being fast on your actual criteria is the whole game.
• Narrow filters and fast alerts beat constant checking. The jobs worth racing for are a small share of postings, and they're the ones that don't wait.

What we can't tell you

Two honest limits, because they change how much weight these numbers deserve.

We can't see below about a minute. We check each district's board many times an hour, but a posting that appears and disappears between two checks looks the same to us whether it lasted five seconds or fifty. "Under a minute" is our measurement floor, not a finding. The fastest districts are at least that fast and may be considerably faster — we're currently running a test to sharpen this, and we'll publish what it shows.

Our window is four days. Postings still unfilled when we ran the numbers are counted at the age they'd reached, so slow districts are "at least this slow." The three-day figure is a floor too.

We've also deliberately left district names out. Publishing a list of which districts are easy would mostly serve people racing you for those jobs.

Where Sub Hero fits

This is the problem the app exists for. Sub Hero watches Frontline continuously and sends you a push notification the moment a job matching your filters appears — which is what puts you inside that 30-second window instead of finding out at lunch. On the Pro plan it can go further and accept the job for you, typically about a second after the posting shows up.

It won't help with jobs you wouldn't have taken anyway, and it can't create postings in a district that isn't posting. What it does is make sure the handful you actually wanted didn't go by while you were teaching, driving, or asleep.

Learn more: app.getsubhero.com

Plans start at $6.99/month. Works with Frontline Absence Management. September is when calendars fill — setting up now means you're watching for the busiest stretch of the year.