Is having a useful product enough to make people use it?

What building a WhatsApp chatbot for ANMs in rural India taught me about adoption, trust, and human behaviour.

I work in design research at the intersection of social development and technology. Currently, I am working on a WhatsApp-based chatbot for ANMs (Auxiliary Nurse Midwives) in India, where they can ask any question related to pregnancy and get an accurate answer. On paper, it is a good product — accurate, protocol-based, approved by the right authorities, with little to no delay. It lives on WhatsApp, a platform ANMs already use daily, and it works through both text and audio, which matters because rural India is a heavy voice-based user of information. And yet, none of this led to immediate adoption by the majority of ANMs who were introduced to it. That got me questioning, is a product being well-intentioned and useful actually enough? Not just for ANMs, but for people like us too?

The burden of too many tools, none of them yours

ANMs are the backbone of rural India’s healthcare system. They are also expected to use a large number of digital tools, mostly for data collection, record keeping, and inventory management. Almost all of these arrive the same way: top-down. The government introduces a product, expects usage, and monitors compliance. As one ANM in UP told me plainly:

“Humare paas itne saare apps hain phone mein, par sab monitoring waale hain. Koi learning ka hai hi nahi.”

(We have so many apps on our phones, but all of them are for monitorin. Not a single one is for learning)

So when the WhatsApp chatbot arrived (something actually designed to help ANMs get answers to their own questions) you’d think it would be welcomed. And it was, but only initially. Post training, most ANMs stopped using it after 0 to 3 questions. The questions they did ask were generic, lacking the nuance of what they actually face on the ground.

When we dug into why, we found that ANMs already have well-established ways of getting information — peers, senior ANMs, Google, YouTube. These systems have been built over years. Simply introducing a new tool, however well-designed, does not displace them. What’s missing is explicit communication of value. And this is not unique to ANMs. Think about the last time you knew about a website or app that seemed useful but never actually tried it out. A good product without a clear value narrative is just another icon or tab on the home screen.

Peers are the real distribution network

Think about the last time something genuinely made your life easier. For me, it was ChatGPT, and Claude. But I didn’t discover them through a product website. I found them through Instagram reels, colleagues mentioning them in passing, a friend screen-sharing during a call. That is how most of us actually adopt new tools — through people we trust, in contexts that feel relevant.

ANMs are no different. They have peers. They have WhatsApp groups. They have senior colleagues whose opinion they trust. But in our case, the chatbot wasn’t generating any chatter among them. It wasn’t finding its way into group conversations. And that pointed to something important — the conversation about any product needs to travel beyond the product itself. Peer testimonials, word of mouth, even informal demonstrations by a respected colleague carry more weight than any training session. Adoption driven by a peer recommendation feels like a choice. Adoption driven by a mandate feels like compliance.

The product itself has to show people what it can do

Even when someone is aware of a tool and has heard good things about it, there’s still a gap between knowing it exists and knowing how to make it work for you. Most of us don’t have time for self-exploration. And so it becomes the chatbot’s responsibility to bridge that gap.

Think about the last time you searched something on Google and ended up going down a rabbit hole you didn’t plan on. That wasn’t an accident, those suggested searches at the bottom of the page kept pulling you in. The same way Amazon keeps you browsing with ‘customers also bought,’ a well placed nudge can turn a one-time interaction into a habit.

The same logic applies here, with the WhatsApp chatbot. Think about what happens after an ANM asks a question and gets a good answer. That could easily be the end of the interaction — question answered, chat closed. But what if, right after the answer, the chatbot surfaced two or three follow-up questions she might not have thought to ask? Something like: ‘You might also want to know about managing this in the third trimester’ or ‘Here’s a related question other ANMs have asked.’ The tool should teaches you what it’s capable of, at exactly the moment the user is most receptive.

A good product is the starting point, not the finish line

I hope this makes it a little easier to understand why going beyond building a good product matters, especially in contexts like this. And what’s interesting is that the ways to do it — communicating value clearly, using peer networks, designing for continued engagement aren’t new ideas invented for ANMs. They’re behaviours we recognise from our own lives, things we experience every day but forget to notice.

Because even when the user is an ANM in rural UP and the designer is sitting in a city, the basics of human behaviour turn out to be more similar than we expect. The context changes. The product changes. But the question of what makes someone genuinely choose to use something — that stays the same.