Every EAP sales deck in India looks the same: a therapist headcount, a language count, an app screenshot, a logo wall. None of it tells you whether your employees will be helped. The information that predicts that lives in the answers to nine questions, most of which vendors are not used to being asked.
Choosing an EAP provider in India comes down to nine questions: real utilisation and its formula, who the therapists are, what HR can see, manager training, crisis response, reporting, contract flexibility, language and cultural fit, and what the provider refuses to sell. A vendor's answers matter less than whether they have them.
We are an EAP provider, so read this the way you would read a restaurant's guide to choosing restaurants: useful precisely because the author has to survive their own criteria. Ask us these nine too.
What is your real utilisation number, and how is it calculated?
The answer you want is a percentage, its exact formula, and the population it counts, in writing. "Utilisation" is the most gamed number in this industry: some vendors count webinar attendees, app downloads, or newsletter opens to inflate it. The benchmark to anchor on is case level use, where the largest published vendor study, Attridge and colleagues (2013, Journal of Workplace Behavioral Health), found an average of 4.5% of covered employees opening a counselling case per year, median 3.6%. Ask for unique counselling cases as a percentage of covered employees, excluding dependents and excluding non clinical touches. A vendor who cannot produce that number in that form, for a comparable client, is telling you they either do not measure it or would rather you did not see it.
Who are the therapists, and who employs them?
The answer you want names qualifications, an employment model, and a supervision structure. Much of the industry runs on loosely affiliated panels: counsellors on per session payouts, onboarded at speed, with little clinical supervision. Ask what fraction of sessions are delivered by the provider's own supervised clinicians versus panel freelancers, what the minimum qualification is (an MPhil or a Master's in clinical or counselling psychology means something specific in India; a weekend certificate does not), and who reviews difficult cases. The quality of your program is the quality of the person your employee meets in session three. Everything else is packaging.
What will HR be able to see?
The answer you want is: aggregate numbers only, never names, never who used the service, with the boundary stated in the contract. Confidentiality is not a courtesy in India; Section 23 of the Mental Healthcare Act, 2017 gives a person receiving mental health care the right to confidentiality of that care. A provider should be able to tell you exactly what their reports will never contain, and should be visibly comfortable saying no to you, the buyer, on this point. If a vendor hints that HR can get "a sense of who is struggling," end the call. That vendor will also hint it to your employees, and your utilisation will be zero for the most rational reason there is.
Do you train our managers, or only run a helpline?
The answer you want includes a manager training component with rehearsal in it, not a webinar. The WHO's 2022 workplace guidelines recommend manager training for mental health as a core intervention, and for good reason: your employees' daily experience is their manager, not a phone line. A helpline waits for the distressed person to act; a trained manager notices earlier and refers sooner, which is how struggling employees actually reach help. Ask what the training involves, whether managers practise the actual conversation, and what refresh cadence is included in the retainer versus billed extra.
What happens at 2am in a crisis?
The answer you want is a specific protocol with named response times: who answers at 2am, with what clinical qualification, escalating to whom, coordinating with which emergency services, and what the provider commits to in writing. Not "24/7 support available," which can mean an IVR and a callback at 10am. Walk through a scenario with them: an employee calls at 2am saying they cannot keep themselves safe tonight. Ask them to narrate the next sixty minutes, step by step. Any hesitation in that narration is information. This is the question where glossy vendors fall apart and serious ones become obvious.
What does your reporting actually show?
The answer you want is a sample report, anonymised, from a real client, showing outcomes and not just activity. Activity reports count sessions and webinars; outcome reporting shows whether people who used the service got better, on recognised measures, with honest limits on what the numbers can claim. Ask also for the minimum group size below which they will not break down data, because slicing reports by small teams deanonymises users without anyone intending it. A provider who protects your employees from your own dashboard is a provider your employees can trust, and trust is the entire mechanism of the product.
What are the contract terms if utilisation stays low?
The answer you want includes a review clause, not just an annual lock in. If six months in, real utilisation sits near 1%, what happens: do they investigate and fix, restructure the package, or simply invoice? The honest structural answer is that a pure per head retainer pays the vendor the same whether anyone is helped or not, so the contract should build in the correction: a utilisation review at a defined date, an agreed remediation plan, and pricing that can shift toward usage as data accumulates. Also ask about exit terms. A provider confident in their delivery does not need a punitive lock in to keep you.
How do you handle language and cultural fit?
The answer you want goes beyond a language count into matching and context. Therapy in a second language is harder therapy; ask whether an employee in Coimbatore or Indore can be matched to a counsellor in their language, at their hours, and whether counsellors are actually working in the Indian contexts your people live in: family pressure, marriage expectations, hierarchy at work, money sent home. A panel of a thousand counsellors matters less than whether the matching process gets your specific employee to the right three. Ask how matching works, how rematching works when the first fit fails, and how long each takes.
What will you refuse to sell us?
The answer you want is a real refusal, delivered without flinching. Every serious provider has learned, from delivery, that some purchases do not work: a helpline with no internal communication behind it, a program for a team too small to use it anonymously, a wellness app bolted on to look complete. Ask each vendor what they would decline to sell a company like yours and why. A vendor with no answer is a vendor who will sell you anything, and that tells you how the rest of the relationship will go. This question costs you thirty seconds and filters the market better than the other eight combined.
The deck tells you what they sell. The nine answers tell you what they will do at 2am.
One honest complication before you build the scorecard: no provider in India aces all nine, including us. The point of the questions is not to find the perfect vendor. It is to find the one who answers plainly where they are strong, admits plainly where they are not, and puts both in writing. In a market this young, the willingness to be audited is the strongest signal available, and it is also a test of you: a buyer who asks these nine is signalling the kind of client who will actually run the program, which is half of whether it works. The providers worth your shortlist will be relieved to meet you.
What do HR teams ask when comparing EAP providers?
What is a good EAP utilisation rate in India?
There is no published India wide benchmark that survives primary source tracing. The best available anchor is international: an average case level utilisation of 4.5% and a median of 3.6% of covered employees per year, from the largest published study of external EAP vendors (Attridge et al., 2013). Ask any India specific figure a vendor quotes to be shown with its formula and source.
Can HR see who used the EAP?
No, and a provider should put that in the contract. Reporting should be aggregate only, with a minimum group size below which data is never broken down. Section 23 of the Mental Healthcare Act, 2017 gives people receiving mental health care a right to confidentiality, and any hint of individual visibility will hold utilisation at zero.
How long should an EAP contract run?
An annual term is standard, but insist on a utilisation review clause at a defined midpoint, an agreed remediation path if usage is low, and clean exit terms. A pure per head retainer pays the same regardless of outcomes, so the contract structure is where accountability either exists or does not.
What is the single best question to ask an EAP vendor?
Ask what they would refuse to sell a company your size and why. Providers who have delivered honestly have refusals ready; providers who will sell anything have none. It is the fastest filter in the market.
Sources: Attridge, M., Cahill, T., Granberry, S., & Herlihy, P. (2013). The National Behavioral Consortium industry profile of external EAP vendors. Journal of Workplace Behavioral Health, 28(4), 251-324. DOI 10.1080/15555240.2013.845050 (open archive record) · The Mental Healthcare Act, 2017 (India Code gazette text) · WHO guidelines on mental health at work, World Health Organization, 2022. Cover image for this article is AI generated.
