Careers Page Job Alerts: Beat the Job Board Digest

Greenhouse and Lever both publish a company's open roles as a public JSON list that needs no login. Watching that list is faster than any digest, and it is two requests to work out which one a careers page uses.

Careers page job alerts work because most companies publish their open roles as a public list that anyone can read, and the applicant tracking system behind the page hands it over without a login. Greenhouse and Lever, two of the systems you will meet most often, both do this. The alternative is a digest email that arrives the next morning, by which time a popular role has a few hundred applications on it. Everything below was read from its source on 12 September 2026.

  • Greenhouse publishes every board at `https://boards-api.greenhouse.io/v1/boards/{board_token}/jobs`. Its docs say plainly that "Job Board data is publicly available, so authentication is not required for any GET endpoints".
  • Lever does the same at `https://api.lever.co/v0/postings/SITE?mode=json`, with a separate EU instance at `api.eu.lever.co`. "The API does not require authentication for retrieving job postings."
  • Both lists carry a date. Greenhouse gives `first_published` and `updated_at`, Lever gives `createdAt`, so a new role is visible rather than inferred.
  • Greenhouse's own candidate alerts are "a daily or weekly email (8:00 AM in their time zone)", cover "up to 8 job titles" by default, and are listed as available on the Core, Plus and Pro tiers.
  • Visualping puts the problem in one line: "Most job board platforms send email digests hours after a role goes live."
  • Working out which system a careers page runs on takes one look at the URL. `boards.greenhouse.io` or `jobs.lever.co` in the address, or in the frame the page embeds, is the whole answer.

Why is a job board alert so slow?

Because it was designed as a digest, not as an alert. Greenhouse's own documentation for candidate job alerts describes exactly what a subscriber gets: "a daily or weekly email (8:00 AM in their time zone) when jobs matching their preferences are posted". Candidates can set alerts for "up to 8 job titles" by default, and the feature is listed as available on the Core, Plus and Pro subscription tiers, which means the company has to be paying for it before you get anything at all.

Visualping's write-up of the same problem is blunt about it: "Most job board platforms send email digests hours after a role goes live." That is fine for a slow market. It is useless for a role at a company people actively want to work for, where the queue forms on the first day.

How do you tell which system a careers page runs on?

Look at where the jobs actually live. Most careers pages are a company-branded wrapper around somebody else's software, and the wrapper leaks the name. Click a job title and read the address bar, or open the page source and look for the frame the listing sits in.

Spotting the system, and the list behind it. Endpoints read 12 September 2026
What you seeSystemThe public list
`boards.greenhouse.io` or `job-boards.greenhouse.io` in a job linkGreenhouse`boards-api.greenhouse.io/v1/boards/{token}/jobs`, no key
`jobs.lever.co` in a job linkLever`api.lever.co/v0/postings/{site}?mode=json`, no key
Neither, and the jobs are drawn straight into the pageSomething bespoke, or a system with no public feedWatch the rendered list instead

The token is the short name in the link. Monzo's careers links point at `job-boards.greenhouse.io/monzo`, so the board token is `monzo`, and on 12 September 2026 the matching API call returned 69 open roles. Each one carried `title`, `location`, `absolute_url`, `updated_at` and `first_published`, which is more than the careers page itself shows you. Lever's own demo board, `leverdemo`, returned 13 postings on the same day, each with a `createdAt` timestamp and a `hostedUrl` pointing at the application form.

What do those feeds give you that the page does not?

A date on every row. The Greenhouse job board API docs list `first_published` alongside `updated_at`, so a genuinely new role is distinguishable from an old one that had its salary band edited. The Lever postings API gives `createdAt` and also takes filters in the query string: `team`, `location`, `department`, `commitment`, `level`, plus `skip` and `limit` for paging. Filtering server-side means the thing you are watching only changes when a role you would actually apply for appears.

It is also a much smaller thing to fetch than a careers page, which matters if you want to look often. That trade is the subject of a longer piece on how often you can check a website without getting blocked.

How do you set up careers page job alerts?

  1. Find the real list

    Click any job on the careers page and read the address. If it lands on greenhouse.io or lever.co, the company name in that URL is the token you need for the API call in the table above.

  2. Narrow it before you watch it

    On Lever, add the filters to the query string so the feed only holds roles in your team or country. On Greenhouse the list comes whole, so filter with keywords on the monitor instead.

  3. Point a monitor at it

    A Visual Selector monitor lets you click the part of the page that holds the listings and watch only that, which keeps a cookie banner or a rotating staff photo from firing an alert. A Custom monitor does the same job with a CSS selector if you would rather type one.

  4. Filter on the words that matter

    Keywords and negative keywords cut a 69-role board down to the two titles you care about. A board that changes weekly is fine to watch whole. One that changes daily is not.

  5. Send it somewhere you will see it

    Discord and Telegram arrive in seconds, email arrives when your provider feels like it. Setting up notifications covers the three of them.

  6. Check what an alert looks like before you trust it

    Change the monitor's interval to something short, wait for the first alert, and read it. An alert you cannot act on from the notification itself is one you will learn to ignore.

What if the careers page has no public feed?

Then watch the page. That is the ordinary case rather than the sad case, and it is what monitoring a website for changes is for. Pick the element that holds the job list, not the whole page, because a careers page is full of things that move on their own: a tagline that rotates, a cookie prompt, a count of employees that updates on its own schedule. A monitor pointed at the list itself fires when the list changes and stays quiet the rest of the time.

Common questions

Is reading the Greenhouse or Lever API allowed?

Both publish it as a public feature for exactly this. Greenhouse says "Job Board data is publicly available, so authentication is not required for any GET endpoints", and Lever says "The API does not require authentication for retrieving job postings". They are how the company's own careers page gets its jobs.

How often should I check a careers page?

Match it to how fast the company hires. A team posting a few roles a month does not need a 5-second check. Somewhere that posts daily and fills fast is a different question, and the trade-offs are covered in our page on checking a website without getting blocked.

Will I get an alert before the job hits LinkedIn?

Usually, because the company's own board is where the role appears first and the boards are fed from it afterwards. Greenhouse's own alerts go out as a daily or weekly digest at 8am, so anything watching the list directly is ahead of that by design.

Can I watch several companies at once?

Yes. One monitor per careers page is the simplest arrangement, since each one has its own list and its own pace. Keyword filters then decide which of them are allowed to interrupt you.

Does this work for things other than jobs?

It is the same mechanism behind any page that grows a list: a shop adding products, or a resale site adding tickets. Twickets alerts is the ticket version of the same problem, and it is worth reading for how differently a site can behave when demand is high.

Watch a careers page Free monitors check every 60 seconds, which is quicker than any digest.