Here's the quiet tragedy of most EAP purchases. The program was fine. The clinicians were qualified, the platform worked, the contract was signed and filed. And almost no one ever used it.
That's the thing to hold onto while you evaluate vendors: you're not really choosing a service. You're choosing whether, on the worst night of someone's year, they'll reach for help — or sit alone with it. Every feature on the slide deck is secondary to that one question.
So here's how to choose for the thing that actually matters.
Start with the only number that counts: will people use it?
Most providers will lead with their network size and their app's features. Make them answer a harder question first: what's your real utilisation, and how do you drive it?
Because a program nobody uses isn't the cheaper option. It's just a more expensive way to help no one. Ask how long it takes someone to go from "I need help" to actually being helped, and what stands between those two moments. The fewer steps, the better. This is where AI-first models tend to win — the first step is a private conversation, not a phone call to a stranger. (It's the whole design behind Noa, Anshap's AI companion built by psychologists.)
Then ask the question your employees are silently asking
People only use support they trust, and trust dies the instant someone suspects their manager could find out. So press hard here:
Can anyone at the company ever see who used the service, or what they said? Ask to see the actual employer dashboard — not a description, the real thing. It should show anonymised, group-level trends and nothing else. Ask what the minimum group size is before any insight appears, so a team of four can never be reverse-identified. And ask, plainly, how they comply with the DPDP Act, 2023.
If the privacy answers are vague, you have your answer. Walk away. (Anshap's approach — DPDP Act, 2023 native, hosted in India, with a minimum cohort size before any reporting — is on our security page.)
Make sure the clinicians are real, and verified
Anyone can call themselves a wellness coach. Therapy is a different thing entirely. For the human side of an EAP, ask how professionals are vetted, what credentials they hold — including RCI registration where applicable for clinical roles — and what range you get: counsellors, clinical psychologists, psychiatrists. A serious provider verifies its people before listing them, so you're not left vetting credentials yourself. (See what that looks like on our professionals page.)
Walk through a crisis, step by step
This one is non-negotiable, so don't accept a brochure answer. Ask the provider to walk you through, moment by moment, exactly what happens when someone is at risk of harm. A credible EAP has a clear escalation path to urgent human help and visible signposting to national resources like Tele-MANAS (14416) — not a chatbot that politely hands back a phone number.
Ask what the organisation actually learns
The point of the employer-facing side isn't surveillance — it's understanding. Can it show you, anonymously, where stress is rising in time to do something about it? Can it connect to the outcomes you already care about, like attrition and engagement? The best programs help you fix root causes, not just count logins.
The test that cuts through every pitch
When the demos blur together, score each provider on six things: utilisation, privacy, clinical quality, crisis safety, insight, and scale. Then notice which one you were about to pick on price alone. If it's a 5 on cost and a 2 on utilisation and trust, it's the expensive choice — because it helps no one and your people don't believe in it.
Choose the program your people will actually open the door to. Everything else is detail.