The first thing a job aggregator does is brag about its inventory. "Search ten million jobs." "The largest catalog on the web." If you have ever used one of those sites, you know the rest of the story. The first page of results is a wall of duplicates. Three of them are the same role re-posted by three different agencies. Two link to a careers page that has already removed the listing. One is a screenshot of a 2022 description that nobody took down. By the time you find a role worth applying to, you have spent twenty minutes filtering by hand for things the site should have filtered for you.
The jobs feed at /jobs is built on the opposite premise. It is small. It is hand-picked. It does not try to be every job on the internet. It tries to be the few hundred you would want to look at this week.
This post is what that looks like in practice, and why we chose this shape instead of the obvious one.
The aggregator's quiet tradeoff
There is a reason every big job site looks the same. Volume is cheap to scale, and it is the only metric that fundraises well. A board that surfaces a hundred curated roles a week sounds tiny next to one that "has ten million listings". The aggregator is selling you the search box. The fact that the search box returns mostly garbage is a problem you are supposed to solve, with filters, with saved searches, with notifications that go off too often.
What gets lost in the volume game is the thing you actually wanted: a short list of jobs that are real, current, and worth your morning.
The cost of the aggregator model is paid by you, in time and in false hope. Every dead listing you click is a small tax. Every ghost role that interviews you for six rounds before going quiet is a large one. Multiply by however many roles you are looking at, and the math gets bleak fast.
What "curated" actually means here
Curation is one of those words that has been used to sell so many things that it has stopped meaning anything. Here is what it means on Applying Pal, in plain language.
A role is on the feed because we decided it should be on the feed. Not because a search index returned it. Not because an algorithm guessed you might be interested. Someone on the team looked at the listing, decided it was current, decided it was worth surfacing, and let it through.
A role comes off the feed when it stops being a real role. Roles that get filled, withdrawn, or quietly abandoned do not stay up. The "no ghost listings" promise is enforced before you see the card, not after you click it.
Tags are honest. Remote means remote. Hybrid means a real in-office expectation. On-site means on-site. Compensation is shown when it is published and absent when it is not. We do not invent a midpoint to make a card look complete.
The feed is small on purpose. There is a cap. We could surface twice the volume by lowering our bar. We chose not to. The number you see when you visit /jobs is the number we think is worth your attention this week.
The filters that earn their keep
A filter is only useful if the underlying data is consistent. Four of them carry most of the weight on the feed.
Location type. Remote, hybrid, on-site. We do not let "remote-friendly" creep into the remote bucket, because that phrase has been doing too much work for too long.
Experience floor. A minimum-years number when the listing provides one, blank when it does not. You filter on a number, not on a job ladder rung that means different things at different companies.
Compensation. When a band is published, we surface it. When it is not, we say so. Applying to a role without comp data should be an informed choice, not an accident.
Time window. A role posted yesterday is a different role from a role posted six weeks ago. The feed leads with recent ones, and the time filter is the one most people end up living in.
We deliberately did not ship a filter for "company culture" or "growth stage". The data underneath those labels is too soft to filter on without lying to you.
What is on the roadmap for the feed
Three things we are working on, in rough order:
- A weekly digest you can opt into, with the new roles that match the filters you care about. Email, sent once a week, no daily noise.
- Saved searches, so the filters you live in persist across visits without you redoing them.
- A view that groups roles by company, for when you are casting a wider net at one specific place rather than across companies.
What we are not building, despite frequent requests: a one-click "apply to fifty roles" button. That is the bulk-spam path, and it does not respect either side of the hiring loop. We would rather make the application you would have sent anyway take ninety seconds instead of nine minutes. The Chrome extension is the answer to that, and it is next on the public roadmap.
The honest tradeoff
You will see fewer roles on Applying Pal than on the big aggregators. That is the tradeoff, and we own it. The ones you see have a much higher chance of being live, being real, and being worth a thoughtful application.
If you would rather scroll for an hour through ten thousand listings, an aggregator is a click away. If you would rather scroll for fifteen minutes through a hundred and find the three worth applying to, open the feed and let us know how it goes.