Diversity recruitment software claims tested, what candidate sourcing software and candidate sourcing tools actually do, how candidate screening should be structured, and where candidate tracking software fits between the two

Structuring selection before automating it

1. Agree the criteria before seeing anybody. Written, job related, and the same for every candidate. This costs an hour and does more for consistency than any product, and it is the precondition for automation being safe rather than risky.
2. Ask what any automated screen is scoring. If a vendor cannot explain the basis on which candidates are ranked, you cannot defend the decision. Opacity is a reason to decline rather than a proprietary advantage.
3. Monitor outcomes by group. Run the numbers on who progresses at each stage. A tool that shifts those numbers is doing something, and you want to discover that in your own reporting rather than in a complaint.

What sourcing tools actually deliver

Reach into people who are not applying, mostly through professional networks and aggregated profiles, plus the campaign tooling to contact them at scale. The value is real where your roles are scarce and near zero where you already receive more applications than you can read.

Before buying, count your applications per vacancy. Employers drowning in applicants who buy sourcing tools have bought a solution to somebody else's problem, and the screening end is where their money should have gone.

Screening, where the risk concentrates

Knockout questions, keyword filters and scored assessments all reduce a pool by rules, and rules that are not job related can exclude groups of people systematically. That exposure exists whether the rule was written by a person or learned by a model, and it belongs to the employer rather than to the vendor.

The practical protections are unglamorous: job related criteria agreed in advance, a human review of anybody rejected automatically at the first pass, and outcome monitoring by group. Those three make automated screening defensible.

Diversity claims, and what to ask for

Products in this space claim blind screening, inclusive language checking and balanced shortlists. The first two are mechanical and verifiable. The third deserves care, because engineering a shortlist by group is a very different act from removing bias in how it was assembled, and the law on it is not the same everywhere.

Ask any vendor for the evidence behind the claim and for the option to switch each mechanism off. A tool you can audit and disable is one you can adopt; a tool whose behaviour you cannot describe is one you will have to defend without knowing what it did.

Common questions

Is automated screening lawful?

Where the criteria are job related, applied consistently, and outcomes are monitored, generally yes. Where the basis cannot be explained, an employer is carrying a risk it cannot assess.

Does blind screening actually help?

Removing names and other identifying details at the first pass is a simple change with reasonable evidence behind it. It is also the easiest of these tools to adopt and to reverse.

Is candidate tracking software the same as an ATS?

In practice yes, with the emphasis on the candidate record rather than the vacancy pipeline. Agencies care about that distinction far more than employers do.

Should rejected candidates always get a human look?

At the first automated pass, yes, at least as a sample. It catches mechanical failures and it is the control that makes the rest of the system defensible.

Coverage by country

• Employer of record vendors covering Singapore
• Employer of record vendors covering Mexico
• Employer of record vendors covering Spain
• Employer of record vendors covering Colombia
• Employer of record vendors covering United Kingdom
• Employer of record vendors covering France
• Employer of record vendors covering Hungary
• Employer of record vendors covering New Zealand