Customer-Care Principles Give Employers A Better Test For AI Hiring Tools

A candidate screened out of a hiring process usually receives a form rejection, with no stated reason and no way to correct a misread. Criteria's 2026-2027 Hiring Benchmark Report puts AI in the screening phase at 46% of hiring professionals, the most common use anywhere in the funnel. A hiring team that asks the right questions while it's still choosing a tool can make the screening more candidate-centric and keep the speed AI brought to it.

Leehla Huerta is the creator of the The Five Human Truths™ framework, a speaker, and an author. She's spent more than three decades leading customer care and experience organizations across hospitality, telecommunications, legal services, and e-commerce. Her framework rests on five things she says people want: to belong, to be understood, to feel valued, to keep a sense of control, and to reach their potential. Huerta first mapped The Five Human Truths onto customer standards inside a contact-center operation.

"AI can identify a potential mismatch and provide a score, but it should never be the sole authority making a final judgment about a person’s capability or potential," says Huerta. She spent eighteen months applying for work before her current role, and worked through enough applicant tracking systems in that period to form her view from the candidate's side. She has since turned her framework on hiring tools, as a set of questions a team can run before one goes live.

Five questions, one tool

The Five Human Truths began as a customer-care framework, and Huerta rewrote each one as a question to put to a hiring system. She starts with "belong," which asks whether everyone an employer meant to reach can finish an application, including a candidate whose qualifications arrive as equivalent experience in place of a credential. "If there's education and there's equivalent work experience, how does AI work through that with a resume?" says Huerta.

Next comes "understood." "The questions that I would ask under understood would be, 'Can this tool see context, and does that context change the evaluation of that particular candidate?'" says Huerta.

"Valued" counts how often a system asks for information it already holds. "Control" covers what a candidate can do about a result: correct an error, challenge a score, or reach a person. Huerta's reference point is an interactive voice response menu, where pressing zero once reached an operator. She applies the same test to an application portal. "Much like an IVR, there's no zero escape anymore. Even when you scream 'agent,' it doesn't work anymore," says Huerta. "In the candidate role, there's no zero escape to talk to someone."

Rounding out The Five Human Truths is "potential," which separates trajectory from track record.

The second layer of review

Huerta's proposal keeps AI on the first read and adds a step after it. She draws on an example of setting a threshold at a candidate who meets 75% of the stated criteria, an application a tool may be set to otherwise reject on its own. “Allow AI to organize the volume and identify candidates who appear not to meet the requirements, but require a second layer of review before closing the door.”

The measurement Huerta wants comes from contact centers and e-commerce. Application abandonment can be tracked the way a retailer tracks an abandoned cart, and a repeated question can be counted the way a service team counts a customer retelling a story. "Companies would benefit from looking at candidates as a customer," says Huerta. "They should be measuring cart abandons just like an e-commerce platform."

Criteria's Hiring Benchmark Report names candidate perception and experience among the top concerns holding hiring teams back from AI, cited by 28% of respondents and by 43% of those in technology. Huerta would put a number on it directly, adding a possible survey question to candidates after a rejection as well as at submission. "What you hear in your feedback would be what was missing in the requirements conversation during implementation," she says.

Testing the tool yourself

Cost, legal, operations, and HR all have someone in the room when a company buys a hiring tool, and the CEO signs off. Huerta calls the candidate’s seat an empty chair, and would like to see more executives test drive the application process themselves. "AI is a trust decision. I’ve always said that the top executives that make a decision on a tool like this should go through the process and have others go through it,” she advises.

A demo can make a product look like it does everything a buyer wants. Huerta has been through a long list of them, and she asks a vendor to prove it while she watches. "I'm going to upload my resume, we're going to run it through your tool and let's see how it comes out," she says. Huerta turns the same question on an employer, asking about culture when she's the one being interviewed. "Am I going to feel like I belong at that location, at that company, in that job, with that boss? Ultimately, that's important to a lot of us," she concludes.

Huerta invites readers interested in learning more about The Five Human Truths framework to connect with her on LinkedIn.

The views and opinions expressed are those of Leehla Huerta and do not represent the official policy or position of any organization.