Industry Guides

Connecting Your Website to an ATS: Options and Trade-Offs

Embedded widget, feed import or API: how each way of connecting a website to an applicant tracking system affects branding, search visibility, the apply step, data freshness and what happens to expired jobs.

Connecting Your Website to an ATS: Options and Trade-Offs

There are three ways to get jobs from an applicant tracking system onto your website: embed the vendor's widget, import a feed, or connect through the API. Each can be the right ATS integration, and each fails in a predictable way when it is chosen for the wrong reason. The widget is live in an afternoon and invisible to search engines. The feed gives you real pages and data that can be hours old. The API gives you everything, along with a piece of software to maintain.

This comparison is for agencies and in-house recruitment teams deciding how far to go. It applies to Bullhorn, JobAdder, Greenhouse, Workable and any other system, because the vendor's developer documentation will tell you which of the three it supports.

Three kinds of ATS integration side by side

Embedded widgetFeed importAPI integration
How it worksA script or iframe from the ATS draws the job list inside your pageYour site fetches an XML or JSON file on a schedule and creates its own job pagesYour site requests jobs from the ATS and posts applications back to it
Job pages on your domainNoYesYes
Control over designColors and fonts at mostFullFull
How current the jobs areAlways currentAs recent as the last importClose to real time
Apply stepThe vendor's formUsually a link to the vendor's formYour own form, sent to the ATS
Effort and upkeepLowestModerateHighest, and it continues after launch

When the embedded widget is enough

If you are an employer with a handful of openings and candidates reach your careers page from your own site or from job boards, the widget is a reasonable choice. Paste the snippet, write a proper careers page around it, and stop there.

Know what you are giving up. Content inside an iframe is generally credited to the vendor's domain by search engines, and a list drawn by a script has no separate address for each job on your site. That means no job pages of your own to rank, carry structured data or share. Styling is limited to what the vendor exposes, and the apply step happens on their domain, so your analytics see a visitor leave and nothing after that. For a staffing agency, whose website exists to attract candidates, those limits usually rule the widget out.

What a feed import gets right and where it goes stale

With a feed, your site builds a real page for every job, in your design, on your domain. A scheduled task fetches the file and, for each job, looks up the ATS job ID on your site, then creates a listing, updates the existing one or closes one that has left the feed.

Four rules separate an import that runs for years from one that causes trouble every month:

  • Match on the ATS job ID, always. Never delete everything and re-create it on each run. That changes addresses and breaks links in job alerts.
  • Map values, do not copy them. Categories, locations and contract types in the ATS rarely match the filters on your site. Keep a mapping table, and send anything unrecognized to a review list.
  • Clean the description. Job text pasted from a word processor arrives with inline styles. Strip it down to paragraphs, lists and headings before saving.
  • Treat an empty or failed feed as an error. If the file cannot be fetched or contains no jobs, stop and alert someone. The classic import bug closes every job on the site because the vendor's server had a bad minute.

The weakness is time. A job closed at five past nine stays on your site until the next import, collecting applications nobody wants, so run the import as often as the vendor regenerates the feed. On WordPress, do not rely on the built-in scheduler, because WP-Cron only runs when someone visits the site. Turn it off in wp-config.php and let the server's own scheduler trigger due tasks with WP-CLI:

define( 'DISABLE_WP_CRON', true );

Then add a line to the crontab of the site's system user. This one runs whatever is due every fifteen minutes:

*/15 * * * * wp cron event run --due-now --path=/var/www/example.com

If your jobs are a custom post type, as they are with a job board plugin such as WP Job Manager, the import writes to that post type and the theme and search need no changes.

What an API integration adds

An API connection does everything the feed does and removes its two limits. Jobs can be updated within minutes, either by asking the ATS for records changed since the last check or by receiving a webhook when a job changes, if the vendor offers them. And the application can be posted straight to the job record in the ATS, with the resume file, the source and the consent choices, while the candidate stays on your site and your analytics see the whole funnel.

That control has running costs. Access tokens expire and must be refreshed, rate limits cap how many requests you may make, and vendors retire old API versions. Find out early whether API access is included in your subscription and whether there is a sandbox to test against. Keep the API credentials on the server. A key placed in browser JavaScript is public: anyone can copy it and call the API in your name.

A common middle route is a feed for jobs and the API for applications only: indexed job pages and your own apply form, with a smaller integration to maintain.

Design the apply step so no application is lost

The worst outcome of any ATS integration is a candidate who applied and whom nobody ever saw. It happens when the website sends the application to the ATS while the candidate waits, and the ATS is slow or down at that moment. Build the step in this order instead:

  1. Save the application and the resume on your own server first, in private storage.
  2. Confirm to the candidate immediately, on screen and by email.
  3. Send the application to the ATS from a background queue.
  4. On success, record the ID the ATS returns and schedule the local copy of the resume for deletion.
  5. On failure, retry with growing intervals, and alert a named person after a set number of attempts.

Two details matter inside that queue. Attach a unique reference to each application and check for it before creating a record, so a retry after a timeout does not produce a duplicate candidate. And pass the source along (job board, search, email alert), because recruiters decide where to spend advertising money from that field. The wider pattern is in designing integrations that survive third-party API failures.

Expired jobs and what search engines expect

Once your site has its own job pages, closing them correctly is your responsibility. When a job is filled, either remove the page so that it returns a 404 or 410 status, or keep it, state clearly that the role is closed, link to similar open jobs and remove the JobPosting structured data. Do not redirect every closed job to the home page, which confuses candidates and search engines alike.

Google's job posting guidelines say that expired jobs must not stay marked up as open, and that structured data belongs on the individual job page and not on lists. Update the sitemap in the same run. The markup itself is covered in getting listings into Google for Jobs.

How to choose, and what to ask your vendor

An employer with a few roles and no plan to attract candidates through search should take the widget. An agency or a large employer that wants job pages in search results needs at least the feed. If application volume is high, or you want your own form and reliable source tracking, add the API for applications or go fully API-based.

Before you commit, put five questions to the ATS vendor:

  • Is there a jobs feed, in what format, and how often is it regenerated?
  • Is there an API, is it included in our plan, and what are the rate limits?
  • Can an application be created through the API with a file attachment and a source?
  • Are there webhooks for job changes?
  • Is there a sandbox or test account?

A feed import into a WordPress job board is work a competent WordPress developer can take on. Posting applications through an API with queues, retries and file handling is a proper API integration project, and it is where candidates are lost if it is done casually. Our page for recruitment and staffing companies sets out how we plan this kind of build.

Frequently asked questions

Will an embedded careers widget hurt our SEO?

It will not damage the rest of your site. It means the jobs are not pages on your domain, so they cannot rank or carry your structured data. If attracting candidates through search matters to you, that is reason enough to move to a feed or the API.

How often should a job feed be imported?

As often as the vendor rebuilds the file, since importing more often than that gains nothing. Ask them for the interval. For an agency with jobs opening and closing all day, an import several times an hour keeps filled jobs from lingering.

What happens to the website if we change ATS later?

With a widget, the careers page is rebuilt around the new vendor's snippet. With a feed or API integration built properly, jobs live in your site's own structure with an external ID, so only the connector is rewritten and job addresses, design and search stay as they are. Ask for the connector to be kept separate from the rest of the site.

Add WorldWin Coder to your preferred sources

Add worldwincoder.com as a preferred source on Google, or open this article in your AI assistant to use it as a source.

Swati Priyadrshni
· Founder & CEO at WorldWin Coder

I'm Swati Priyadarshani, Founder & CEO of LMS Advisor and Co-Founder of WorldWin Coder Pvt. Ltd. (est. 2019). Over the last 10+ years, I've helped enterprises build learning platforms that actually get used - not just deployed.

Keep reading

Have something to build, fix or move?

Tell us where you are. A developer reads every inquiry, and you get next steps and a free quote within one business day.