Integração Adzuna

Adzuna positions its job listings product around reach and match quality rather than raw volume. According to Adzuna's own employer product page:

1. The platform reaches more than 10 million monthly users, with 65% described as a unique audience versus other major job sites.
2. Listings can be distributed across search, email, mobile apps, and a network of more than 100 partner sites.
3. Adzuna applies AI-based matching that it states improves application rate per user by roughly 20%, aiming to route more relevant candidates to a listing rather than just more clicks.
4. The platform offers integrations with 18+ applicant tracking systems, and access to benchmarking data drawn from its historical archive of job postings.

For a feed-management workflow, the practical benefit is straightforward: once your job data is mapped correctly to Adzuna's schema, every new or updated role in your source system (ATS, careers page, or HR platform) can flow through to this audience automatically instead of being re-entered by hand.

Unlike a typical product feed, Adzuna's own XML specification includes a dedicated bidding mechanic built into the feed itself: the cpc tag. Advertisers who support programmatic bidding can include a per-job cost-per-click value directly in the feed, formatted as a decimal like 0.20. Adzuna's documentation notes that your account manager agrees a minimum/maximum CPC range with you upfront, and the value supplied in the feed lets you bid within that range on a per-job basis rather than setting one flat rate for an entire account.

This means feed structure and monetization are linked for programmatic Adzuna accounts — a feed pipeline that can populate cpc per job (e.g. bidding more for harder-to-fill roles) has a direct lever that a static, one-time export doesn't.

Adzuna's job feed specification is an XML format with a simple, strict structure. Per Adzuna's published XML specification:

1. The root element is <jobs>, and every individual posting is wrapped in its own <job> tag.
2. An XML declaration is strongly preferred, e.g. <?xml version="1.0" encoding="UTF-8" ?>, with UTF-8 (Unicode) encoding.
3. The file must be well-formed, valid XML — Adzuna specifically references validating it with a tool like xmllint --noout test.xml antes da submissão.
4. CDATA sections are optional. If you don't use them, angle brackets and ampersands inside field content must be escaped as < e & — literal unescaped characters will break the XML.
5. The feed can be delivered two ways: hosted at a URL Adzuna's crawler downloads on a schedule (the URL can be password-protected), or uploaded to Adzuna's FTP server from an IP address Adzuna has explicitly whitelisted — uploads attempted from a non-whitelisted IP fail outright.
6. Compressing the file with a common algorithm like gzip or zip is explicitly called out as highly preferable.

A minimal example matches this shape:

<?xml version="1.0" encoding="UTF-8" ?>
<jobs>
  <job>
    ...
  </job>
</jobs>

Adzuna's XML specification documents a defined set of job-level tags, split between mandatory fields, "highly preferable" fields, and optional enrichment fields:

Additional optional fields in the spec include geo_lat / geo_lng, postcode, salary_min / salary_max / salary_frequency (accepted values: hour, day, month, year) / salary_currency, date (initial publish date), and cpc for programmatic bidding.

Feedance sits between your source job data and Adzuna's feed requirements, so the general workflow looks like this:

1. Connect your source — your ATS, HR platform, careers page CMS, or a spreadsheet/database export — as a data source in Feedance.
2. Map your job fields to Adzuna's schema: job title to title, your internal reference to id, the full job description (with HTML formatting, 100+ characters) to description, the live apply page to url, and so on through location, country, salary, company, category, contract type and time, and remote status.
3. Feedance generates a compliant XML file with the <jobs>/<job> structure, proper escaping or CDATA handling, and hosts it at a stable URL.
4. Provide that feed URL to Adzuna (or, if you use FTP-based delivery, share the IP Feedance will upload from so Adzuna can whitelist it).
5. Feedance keeps the feed refreshed on a recurring schedule so new postings, filled roles, and edited details stay in sync without manual re-submission.
6. Use Feedance's feed diagnostics to catch structural issues — missing mandatory fields, descriptions under the minimum length, malformed XML — before they cause jobs to be skipped by Adzuna's crawler.

Adzuna's own specification states it downloads the feed and updates its search index at least once per day, and that the file you produce must be updated with at least that frequency.

In practice, job listings are more time-sensitive than most product catalogs — a role that's been filled but stays live on Adzuna wastes clicks (and, on CPC accounts, budget) sending candidates to a dead application page. A feed pipeline that re-syncs whenever a job opens, changes, or closes — while still meeting Adzuna's once-daily minimum crawl cadence — keeps what candidates see aligned with what's actually open.

Adzuna doesn't publish a public list of feed error codes, but its XML specification is explicit about the constraints that cause jobs to be rejected or misread. The most common issues trace directly back to these documented rules:

1. Unescaped special characters. If you're not using CDATA sections, any literal < or & inside a field breaks the XML document. Escape them as < e &, or wrap the field in CDATA.
2. Description too short. Adzuna requires a minimum of 100 characters in description. Short, boilerplate descriptions fall under this threshold and get rejected.
3. Inconsistent country identifiers. If you use custom country strings instead of ISO 3166-1 codes, they must be used consistently throughout the file — mixing identifiers for the same country (e.g. "UK" in some rows, "United Kingdom" in others) can cause jobs to be rejected.
4. Location/country mismatch. Adzuna only maps a job to the country stated in its country tag — a job in Munich tagged with country code UK, for example, is rejected outright rather than reassigned.
5. Multiple locations in one tag. Se um location field contains more than one place, Adzuna only reads the first — the rest is silently ignored, which can produce jobs geo-matched to the wrong area.
6. Duplicate or non-unique URLs. Cada url must be unique and lead to a page where the candidate can actually apply. Reused or placeholder URLs undermine this requirement.
7. FTP uploads from a non-whitelisted IP. If you deliver via Adzuna's FTP server, uploads from any IP other than the one Adzuna has whitelisted will fail.
8. Missing filter fields. remote, contract_type e contract_time aren't mandatory, but Adzuna's spec notes that without them, your jobs simply won't surface when candidates filter search results by those criteria — not a rejection, but a visibility loss.

A couple of these are lifted straight from Adzuna's own guidance; the rest are general feed-hygiene practices worth labeling as such.

1. Keep titles clean (Adzuna-documented). Adzuna's spec is explicit that title should be an actual job title — not padded with location, salary, an internal ID, or urgency keywords — because that padding actively hurts search relevancy on their platform.
2. Use Adzuna's own category list where you can (Adzuna-documented). Mapping to Adzuna's 29 standard categories, rather than passing through unmapped internal category names, is called out as improving how accurately a job gets indexed and surfaced.
3. Supply real salary data (Adzuna-documented). Apesar de salary is only "highly preferable," not mandatory, providing it avoids relying on Adzuna's automated market estimate and gives candidates a more accurate picture up front.
4. General best practice: write substantive, scannable descriptions. Beyond the 100-character minimum, structured HTML descriptions (clear responsibilities, requirements, benefits) tend to perform better with both candidates and Adzuna's own relevancy signals.
5. General best practice: prune stale listings promptly. Since Adzuna re-crawls at least daily, closing out filled roles quickly in your source system limits how long dead listings linger and continue drawing clicks.

What format does Adzuna require for a job feed?
XML, with a <jobs> root element and each posting inside its own <job> tag, per Adzuna's published XML specification.

How often does Adzuna pull in feed updates?
Adzuna's own documentation states it downloads and re-indexes feeds at least once per day, so your feed needs to be refreshed at least that often.

Can I submit my feed from my existing ATS or careers site?
Yes — Adzuna accepts a feed hosted at any URL (including password-protected ones) that it can crawl on schedule, or delivery via FTP upload from a whitelisted IP.

Does Adzuna support remote job postings?
Yes. Per the spec, set location to the relevant country and use the remote field (accepted values 1/Yes/Remote or 0/No/Non-Remote) to flag it.

Does Adzuna charge to list jobs?
Adzuna's own XML spec confirms a cost-per-click (cpc) field exists for advertisers on programmatic bidding arrangements, with ranges agreed per account manager. Beyond that, exact plans and rates vary by market and account type — confirm current pricing directly with Adzuna rather than relying on a fixed number.

Why would a valid-looking job not show up on Adzuna?
Most non-rejection visibility gaps trace back to optional-but-influential fields: missing category, contract_type, contract_time, ou remote won't cause a rejection, but the job won't surface when candidates filter by those criteria.