Howth Technology Factory
All tools

Home Depot Jobs API: the careers site as a queryable feed

A queryable feed of what The Home Depot is currently advertising, from the Atlanta support centre to state-level virtual roles.

Unofficial. Not affiliated with, endorsed by, or sponsored by The Home Depot, Inc., or by Workday, Inc. Structured facts only by default; no personal data is collected.

apify.com/apifmcpfactory/homedepot-jobs-api
Home Depot Jobs API — sample run
Illustrative preview
Success
Deterministic
Fast
Typed output
What it does

In plain language.

Reading a large retailer’s careers site by hand tells you almost nothing, because the interesting pattern is in the distribution of roles rather than any single posting. This returns the whole advertised set as rows so the distribution is visible at a glance.

What came back on 2 August 2026 was not what most people picture when they think of this employer. Nearly a quarter of the sampled requisitions sat at one address: the Store Support Center in Atlanta. Corporate hiring, not shop-floor hiring, dominated the run.

The second pattern was geographic in an unusual way. Several postings carried locations such as Georgia Virtual and Massachusetts Virtual — roles tied to a state for employment purposes but not to a building. Anyone measuring remote hiring from city names alone would misread these entirely.

Distribution appeared too, in the form of a Rapid Deployment Centre posting in West Columbia. Together with Pennant Park in Atlanta, the sample covered three distinct kinds of workplace under one employer brand.

Titles ranged from Senior Product Manager to Night Replenishment Manager, with Installation Coordinator, Operations Experience Manager and Pro Experience Manager in between. Very nearly all were full-time; a small number were part-time, and the field records which is which.

Every row in that run resolved to the United States. Posting dates clustered tightly across the final days of July and the first days of August, which is the pattern you would expect from a retailer refreshing its board continuously rather than in seasonal batches.

A hundred rows is where a default run stops. That is a deliberate ceiling on cost, not a statement about how many vacancies exist, and it moves when you tell it to.

Because each requisition keeps the identifier the source gave it, a second run can report precisely which postings are new and which have disappeared, instead of handing you another undifferentiated list to compare by eye.

Apply links resolve to the employer’s own posting pages, so nothing you build on top of this becomes a stale mirror of someone else’s content.

Results come back as JSON or CSV, and the same run can emit a white-label HTML hiring report when the audience is a person rather than a pipeline.

Who it's for

  • For retail and supply-chain recruiters. See at once whether a competitor is staffing stores, distribution centres or head office this month, because all three appear in the same table with the location that distinguishes them.
  • For engineers wiring up a pipeline. Typed fields, stable ids and a documented shape. Schedule it, land it in a warehouse, and skip the brittle scraping layer that a careers site would otherwise force on you.
  • For AI agents. Callable as an MCP tool with no arguments. An agent asked to report on retail hiring can fetch the rows, compare them with the ids it saw last week, and describe only the change.
  • For workforce and location planners. State-level virtual postings and named facilities sit side by side, which makes it possible to separate genuinely distributed hiring from hiring at a specific site.

Pricing

Pay per event

$1.50 per 1,000 jobs returned.

Billed once against the final result set. Nothing is charged for a run that comes back empty beyond the start event.

Honest limits

What it does not do.

Published vacancies only. Anything filled internally, advertised through an agency, or not yet posted is outside what a public careers site exposes, and therefore outside this.

Facts by default, prose on request. The advert body belongs to the employer. Structured fields are returned as standard and the description is left off until you switch it on deliberately.

Nothing about individuals. No recruiter names, no hiring-manager details, no applicants. Only the vacancy itself.

Virtual is not the same as remote. A state-level virtual location is an employment arrangement, not a guarantee of remote work. The raw location string is preserved so you can judge it rather than trusting a derived flag.

Departments are frequently blank. This employer publishes little department metadata, so that field arrived empty across the sample. Where the source is silent, the field stays null rather than being guessed.

Unofficial. Not affiliated with, endorsed by, or sponsored by The Home Depot, Inc., or by Workday, Inc. The Home Depot is a trademark of its owner, used here only to identify whose public careers listings this tool reads. All data originates from the employer’s own public careers site.

Why it's built this way

Fewer knobs. Better defaults.

Press Start, no setup

The careers site is already wired in. There is nothing to configure on the first run.

Raw locations preserved

Facility codes and virtual designations survive intact instead of being flattened into a city.

Diff, not just dump

Hand back the ids you hold and get new and removed roles rather than a fresh full list.

For AI agents

Same tool, callable as an MCP server.

Home Depot Jobs API runs on Apify, which means it's exposed as a remote MCP tool with no extra setup. Point any MCP-compatible agent (Claude, Cursor, and others) at the endpoint below.

View full API reference on Apify
# MCP endpoint
https://mcp.apify.com?tools=apifmcpfactory/homedepot-jobs-api
# Example call
{ "maxJobs": 250 }