How to Design a WhatsApp Chatbot
How to Design a WhatsApp Chatbot
With over 2 billion monthly active users, it’s no surprise that WhatsApp is a fantastic channel to build a chatbot for. Beyond its fantastic global reach, it also integrates well with popular chatbot functionalities, making it a natural fit.
More so, it’s a platform users have shown they love using chatbots on. Major markets such as India have seen significant uptake of these functionalities, meaning there are tons of potential users out there who already know the ins and outs of using a chatbot.
In this post, we’ll walk you through everything you need to know to get up and running. But first, be aware that WhatsApp is a bit more challenging to build for, and requires patience. In an effort to avoid having the platform be swarmed by bots, WhatsApp has added a few upfront steps to ensure you are using its platform in good faith. Rest assured, once you have completed these steps, the results are well worth it.
1. Upgrade to the Pro Plan
There are a couple of items you’ll want to be aware of before you start building your first WhatsApp chatbot for your website.
First, you’ll need a Pro Plan from SnatchBot. Thankfully, it’s affordable, and offers outstanding value with a great deal of additional features. As there are multiple Pro plan options, be sure to choose the version that includes access to the WhatsApp channel.
You can find even more details about the Pro plan and all it includes here.
2. Provide additional documentation
WhatsApp requires some additional documentation and information in order to use the channel to its fullest potential. Make sure you have the essential details to sign up your bot with WhatsApp:
* Valid phone number
* Display name, compliant with WhatsApp display name policy
If your Facebook Business is not Verified, WhatsApp usage will be limited to sending up to 10 messages to 2 different phone numbers per day.
Please note that any previously existing WhatsApp accounts connected with the provided phone number will be closed, this cannot be undone.
You can find more detailed steps to complete this critical process here.
3. Add your Avatar image and profile information
WhatsApp does not allow the same high degree of customization that you’ll find on some other channels. This is likely to maintain the cohesive look of the platform. It does, however, allow you to add an Avatar image and profile information.
To do so, select Channels from the left-hand menu, followed by WhatsApp, and finally, Avatar and profile.
When choosing an Avatar, it’s recommended that you use your organization’s logo. Essentially, you want potential users to both immediately recognize you, and trust that this is in fact an account connected to you. However, you can also opt for a more human-driven image if you plan on giving a name/personality to the chatbot.
WhatsApp allows you to add the following fields to your profile:
- Business description
- Location and address
- Type of business
- Email address
- Website URL
- Product images
4. Add Company Information
To add even more helpful details to your chatbot, be sure to include your company information. You should do so by connecting information about your company with the purpose of your chatbot.
There are two main reasons to do so:
- Make it clear to users who they are interacting with.
- Provide reassurance to users that the chatbot is in fact affiliated with your organization.
Here’s how you can do it:
- Select My Bots from the left-hand menu
- Select the chatbot you’d like to customize
- Select Configure
- Update the Bot name and Bot description fields
The Bot name field does give you the opportunity to establish who exactly the user is communicating with. There are a few different naming conventions that are quite popular:
- Company name (most common)
- Company name support (common for support-focused chatbots)
- Invented names (often named as a person or representation of the organization)
While inventing a name for your chatbot may seem fun, and is fairly frequent, keep in mind that it impacts the entirety of how your chatbot is written. You’re essentially committed to writing every interaction from the point of view of that character. While engaging, it’s an extra level effort, especially for first-time chatbot builders.
When adding your company information, be sure to focus on whatever is most critical. The description is limited to 150 characters (including spaces), and your users probably wouldn’t read much more than that anyways.
Top 3 WhatsApp chatbots
With so many WhatsApp chatbots out there, finding the best one can be challenging. So, we’re breaking down three chatbots in detail.
Company: JioMart, the e-commerce marketplace of Jio Platforms (one of the largest Indian businesses).
Challenge: As the COVID-19 pandemic began, people turned to online shopping in significant numbers. While this rise in sales volume was certainly a good thing for JioMart, the rise in accompanying support requests were not. They needed a way to bolster customer support outside of traditional web channels.
What they built: JioMart created a WhatsApp chatbot that effectively handles common support requests, such as order status, delays, replacements, and refunds.
- 90% of queries managed successfully
- 1,500 new daily sales
- Average response time cut by nearly 60%
Company: Government of India.
Challenge: The Government of India needed a way to fight misinformation about COVID-19 and give citizens important details during the pandemic.
What they built: In under 5 days, they were able to create a WhatsApp chatbot that could respond to some of the most common questions. This includes items like details about symptoms, travel regulations, and common myths.
- 109M queries answered
- 84M users
Company: Tata Mutual Fund, an Indian investment solutions company with more than 1.9M customers.
Challenge: As more young people became customers, Tata Mutual Fund needed a support channel that would fit with that generation’s preference for mobile communications.
What they built: Tata Mutual Fund rolled out an AI-powered WhatsApp chatbot capable of handling routine requests.
- 40% support call deflection
- 90% end-to-end query resolution
- 67% drop in referrals to call centers