Why we are building a Chatbot as a Platform rather than a Mobile App
This is the 9th article of our series.
Nomad’s Land is a 2-month experimental entrepreneurship boot camp in Chiang Mai, Thailand. The goal is to reach Problem/Solution fit on a product built from scratch. If you want to learn more about our methodology and learnings, check the previous articles.
In this article, we will be explaining:
- Why we opted for a chatbot instead of a mobile app
- Some chatbots specificites
- The difference between a Chatbot as a Platform (CaaP) and a Standard chatbot
Chatbot to provide a better experience
Once we determined our product concept, we had to come up with our solution concept. This phase of reflection included thinking of the most suitable type of medium on which to develop our solution. Here are some of the options that we explored: web app, mobile app, chatbot, and browser extension.
Naturally, when it comes to this type of choices, the most important is to think in terms of customer experience. In short, which one could provide the best and frictionless experience for our future users?
Although we see great potential in browser extensions, we are expecting our users to rely mostly on their mobile devices when they will be using our solution. As a consequence, building a web app or extension could potentially create high friction for our user. We didn’t want to take that kind of risk.
Thus, our two last options were between a mobile app and a chatbot. To be honest, it seemed kind of evident from the beginning to us that we would develop a mobile app — as we had a bad opinion on chatbot’s potential. However, before narrowing down our choice too quickly, we decided to explore in-depth both options.
From this moment, we decided to carry out three tasks:
- Analyzing the current trends*
- Exploring the mediums used by alternative solutions.
- Spending more time exploring chatbots as we were neither accustomed nor enthusiastic about them.
* Although we don’t think that following trends should dictate our end-decision as much as focusing on which one could provide the best customer experience.
1. Trends analysis
About trends, a quick research led us to a couple of recent mobile app statistics (here and there) and chatbot statistics (here and there). In short, beyond the ordinary “mobile apps and chatbots will keep growing” trends, there is a specific statement that held our attention:
One of the basic reasons of chatbots expansion is so called “being tired of apps”. Consumers are annoyed by the need to install special applications to their mobile devices. — Chatbot Report 2018: Global Trends and Analysis
This is a pretty interesting insight as it reveals an important mobile apps’ source of friction. It is true that for most of us, downloading an app may turn out to be an inconvenient task.
2. Alternative solutions
First of all, we looked at the alternative solutions that we mentioned in our previous article. Most of them (Bond, Stay In Touch, and Call Reminder) are mobile apps but none seems to have experienced considerable growth.
Another alternative — mentioned by one of the interviewees — was the built-in extension “Plans” inside Messenger. Plans helps to seamlessly schedule an event with a friend.
After trying it ourselves, we agreed that the fact that this simple feature was already integrated within Messenger made an important difference in terms of user experience.
3. Chatbots exploration
The next step for us was to spend more time trying chatbots on Messenger to see what kind of experience we could potentially provide through this medium.
As soon as we started using a couple of bots, we grasped the potential that could represent this medium for our solution.
We expect our solution to be a very efficient app from end to end. As every productivity apps, the primary objective of our product is to get the Job-to-be-Done (JTBD) as fast as possible for our user and shouldn’t be time-consuming.
An excellent aspect of Messenger chatbots — as compared to mobile apps — is the onboarding part. There is no need to download anything, a simple click on the “Get Started” button and that’s it.
Also, our user research interviews showed that Messenger seems to be the first channel of communication used by distant friends to keep in touch. WhatsApp, on the other hand, seems to be the main one for family-relationship.
Thus, for all the reasons mentioned above — in our opinion — building our solution on top of Messenger makes a lot of sense and we are quite confident that we can provide a better customer experience through a bot. Eventually, we changed our initial idea of building a mobile app.
In summary, we agree with P. Daniel Tyreus’ conclusion in his article “I turned my mobile app into a chatbot. Here’s why.”
I’m willing to speculate that it’s easier to acquire a user if the user doesn’t have to download a new app to use a service. I’m also willing to speculate that users are more likely to continue using a service that’s integrated into an app they already use. Finally, I’m willing to gamble that more and more people are comfortable chatting with a bot on a messaging platform.
Nevertheless, there are also some downsides of building a chatbot on top of Messenger platform which is mainly associated with the lack of flexibility and freedom in some aspects. For instance, the User Interface (UI) is restricted to a conversational UI with little possibilities in term of design and user interactions.
Chatbots particularities
Building a chatbot is more accessible than ever before, thanks to companies such as Chatfuel and Manychat which give the ability toto craft one intuitively without coding.
For all these reasons, a chatbot is often seen as a simple product to develop.
The reality is slightly different though. Technically and design-wise speaking, it is true that building a standard chatbot is easier than a mobile app, however, creating a delightful conversational experience is, in reality, nothing but easy.
It is relatively simple to make that point after observing the low ratio of breakout chatbot successes.
Also, most of the bots created today are used as support functions for big companies’ customer service, marketing or sales and more rarely as an end-to-end product with its own business model.
Chatbot as a Platform (CaaP) vs. Standard chatbot
The major difference between our chatbot and most of the others lies in the fact that we are building a CaaP, not a standard one.
To have a better understanding of what it implies, let’s have a look at the two diagrams below.
In the upper one, there is a direct interaction between the bot and a single user. It is the most common type of chatbots, and this model is perfectly suitable for companies that want to use it as a support function for their marketing, sales and customer service.
However, in our case, this model doesn’t work for us. We need our bot to be the intermediary between a user and his fellow friend. Thus, as shown by the second illustration, the architecture of our chatbot is similar to the one of a two-sided platform.
Technically-speaking, this change brings more complexity and doesn’t entirely fit anymore with the standard chatbot format that companies such as Chatfuel and Manychat support. In essence, shaping our bot isn’t that straightforward and requires from us to go beyond the basic features of bot-building platforms.
By the end of this week, we should have our MVP ready and start running usability testing with our first users. 😄
Coming Next… ⚡️
- A feedback article on our two-week prototyping.
- Additional articles along the way to share our learnings, feelings…
Thanks for the read and stay tuned! 😃
Check this out to get a sneak peek of what is a digital nomad’s lifestyle in Chiang Mai! 📸
Any question or simply talk? Feel free to reach out to us either by email or on Twitter; we will be more than happy to follow up with you! 😉
