How to get people hooked to your chat bot?

In April 2016, at F8, Facebook launched the API for chat bots. Just three months later, there are more than eighteen thousand bots. From ordering pizza to checking news to buying new shoes, to even getting advice on your lipstick shade- all possible services are available on Facebook Messenger now. The bigger question is how many of these bots will really survive? Do they have the capability to change user behavior and create new habits?

The enemy for forming new habits is the past behavior.

Can we suddenly expect our customers to accept the new way of ordering pizza? With bots, they have to answer questions like- which topping would you like? As opposed to their regular way of using check boxes. A chat bot might reduce the number of actual steps to get a pizza at your door step, but it is still challenging our old way of doing things.

How can we get consumer to change their habits?

Nir Eyal, the author of the famous book, Hooked, says that for a product to become habit forming, it should either be high on frequency or on perceived utility. Reading news is a frequent activity. If a user keeps reading on a chat bot, it has a high probability of becoming a habit. On the other hand, something like booking air tickets is a relativity low frequency activity. So, for a user to use bots to book tickets, there should be a high-perceived value in his mind.

How to make habit-forming chatbots?

Nir Eyal proposed a four-step hook model — 1) Trigger, 2) Action, 3) Variable Reward and 4) Investment.

According to him, these steps should be embedded into your bot subtly. One should then run a consecutive cycle of these steps to urge the user to keep coming back to the chat bot. This will get him addicted.

Let’s understand what these steps are and how they can be embedded in chat bots.


Trigger is like a spark plug, that creates an urge to use your product. It could be an external — something like an ad, media coverage, an email campaign or simply word of mouth.

It could be an internal- when the product becomes tightly coupled with a thought, an emotion or a pre-existing routine. Each time this thought or emotion occurs, it urges the user to come back to the product.

Let’s apply this to chatbots. Chatbots live within messaging apps such as Facebook Messenger, WhatsApp, Telegram etc. These messaging apps are already a vital part of our lives. So much so that every time there is a notification beep, we feel a strong internal desire to lift our phone and check it. Thus, chatbots already have some foundation for instigating the user. Push Notifications are a perfect way for chatbot to create a trigger to bring back people to their product.

Keep instigating the user to engage with your bot, that’s the first step to get someone addicted to your bot.


How a user engages with your product after a trigger, is what constitutes the action. In case of apps, it requires a user to download it. But, action in case of chatbots would simply be opening the messaging app and talking to your bot. People know this environment. Infact, they spent most of their phone time on one of them. It is therefore easy to prompt them to action.

The biggest issue the apps developers are facing these days is to get downloads of their apps. But for chat bots the users don’t have to download a new app. Thus, taking action is relatively simple. One just has to open their favorite messaging app, search for their favorite bot and start talking. There is definitely an ease to take action!

Though it is easy to get user on chat bot, but to keep him engaged needs real value. As a founder of Instalocate, we are helping user track their flights in real time. We send them alerts about their loved ones who are flying. They don’t to keep checking the whereabouts of their loved ones. This is a great value to them as compared to google or any other information source which is more static. And that can be their reason to return to use Instalocate.

Variable Reward

The third step in the Hook Model is the variable reward phase, in which you reward your users, reinforcing their motivation to take action. According to Nir Eyal, it is not the sensation of getting the reward that draws actions. It’s the variability of reward. So, products should use variable rewards to entice users to get addicted.

Variable rewards can be of three types: the tribe, the hunt and the self.

Reward of tribe: We are a part of a social system. Our brain seeks out rewards that make us feel accepted, important, attractive and included. Every time we tweet or post; don’t we wait for the social validation in forms of likes, views or upvotes? That’s the reward of tribe.

We at Instalocate are building a personal travel assistant on Facebook Messenger. We let our users share their itinerary with their friends and family and even on their Facebook wall. Our motivation behind introducing this feature is to reward users with social validation.

Reward of hunt: Man as an animal has always been on a look out. Where we once hunted for food, today in this age of technology our hunt is more for deals and information. Companies like Groupon, LivingSocial utilise this form of reward- reward for hunt.

Olabot, with the mission of creating a personal bot, can really leverage on this type of reward. Their celebrity bots can be an interesting way of rewarding fans with a sneak peek into the lives of celebrities. Every time a user returns, he gets some new information about their favorite celebrity. Thus satisfying a general hunt for information.

Reward of self: Finally, there is variable reward we seek for personal gratification. All the gaming apps, or business productivity apps like To Do list, provide a variable reward system build to satisfy our personal desire to seek mastery in the world around us.

Chatbots like PokemonGo bot, or Meekan which helps you in maintaining your calender utilises this kind of reward. They satisfy a user’s personal desire to excel. PokemonGo bot which lets you find Pokemons on Facebook Messenger rewards by helping in outdoing our friends on PokemonGo game.


Investment is the last step in the Hook Model that is critical for habit-forming products. Before users create the mental associations that activate their automatic behavior, they must first invest in the bot. Higher the time and effort, a user spends on a chatbot, more value he adds to it.

If a user has invested time in saving his address and choice of pizza in a chatbot, there is a very high chance of him coming back to order at the same place. Or a news bot, where a user invested in selecting his areas of interest is likely to see him return.

Chatbots is a relatively new phenomenon but with each passing day we see new features being released to enhance them. These features can give bot makers a lot of scope to create new avenues of user engagement. With the way things are progressing, Facebook will soon allow bots to be added to the groups. There will be scope to create profiles, or networks within bots. These will really increase the scope of investment in the product from the user.

Chatbots are here to stay. They will be new apps of the future. There is a lot of energy and potential in the ecosystem. It will be very exciting to see which products will get our users HOOKED.