Helpful Tips and Fundamentals About Creating a Better Chatbot Strategy

Source: Pixabay

Remember the days when businesses had to hire someone to answer customer’s simple, silly questions? Me neither.

Oh, now that I think of it, I do. I remember reading or listening to something like ‘Imagine a day when businesses will use a computer to auto-answer their user’s queries’. It must have been pre-2015⁠— a time when chatbots were in a nascent stage. Siri was a thing but not smart enough. And Google assistant didn’t even exist.

Fast forward half-a-decade, and we can see stats displaying the rise of chatbots.

Today, people expect sites to have a chatbot that gives super-fast, 24/7 support. On the other hand, companies use it to provide a better customer experience. Moreover, many companies use it as an alternative to a customer support member, i.e., they don’t have to pay an extra employee.

For example, If an individual chose Magento vs Shopify for, say, better SEO for their ecommerce store, they lose on customer support, which Shopify provides for a bit extra. In such case, chatbots can come in handy at least for first few months/years till the shop is profitable enough to hire a support person.

Either way, there are obvious benefits of using a chatbot.

So, if you are a business not leveraging smart-chatbots, you should. And the first thing you need to do once you embed a chatbot is: create achatbot content strategy.

Chatbot content strategy:

It is the sum of the Q & As which builds the conversation tree. Here, the core focus is on business goals and user experience. 

To build a chatbot content strategy, you must start with an objective.

Start with goals

Primarily, the purpose of a chatbot is one of two:

1. Customer service: This chatbot helps the customer. Based on customer queries, it can answer/redirect to the appropriate answer or transfer the chat to the relevant team. 
2. Sales: Chatbot can be your 24/7 lead generating machine. For e.g., Chatbots can replace forms and take user details(i.e., leads). Here’s a great example(I tried on the Landbot site):


Fig 1: Landbot chat example.

In either case, the customer doesn’t feel like they are filling a form or sending an email. That’s what customer experience is about⁠— a smooth, flawless journey.

Moreover, chatbots can do the job of landing pages too. Just as you weave a buyer persona in landing pages, same can be done with chatbots as well. Make it relevant, funny, and sound.

In a nutshell, lock your objective. Once the goal is defined, practice the following content strategy to achieve results:

1. Start with greetings:

Humans begin their conversation with greetings, and that should reflect on your chatbot. Begin with common greets such as “Hi there”, “Hello”, “Good morning”, etc. You can greet based on context also. For e.g., The screenshot (above) is of a chatbot embedded at the end of an article, so it greets me with “Hello there, reader”. 

Alternatively (and this is what most do), Move on to introduce the bot and ask the name of the user and purpose. E.g., 

Bot: “This is XYZ bot.”

Bot: “May I know your name?”

User: “ABC”

Bot: “Hi there, ABC. How can I help you?”

In short, insert the human elements.

2. Develop options:

The Landbot example we saw was for sales. Therefore, they had two options which lead to their end goal:

• “Try Landbot” for direct sell.
• "Subscribe to newsletter” for lead generation.

Similarly, give auto-generated options that serve the purpose. For e.g., If your business is an e-commerce store, options such as “My orders”, “Track my order”, “Buy my wishlist”, etc. can fasten the customer’s process.

3. Conversations:

Only a few tags can be presented as “options”. In most cases, users want more. Thus, dialogue or custom message mode is much needed. The thing to remember is, it should feel like a conversation⁠— fun & helpful.




How to improve chatbot conversations flow

1. Make conversations more funny and improve user experience.

a) Use GIFs & Emojis:

Again, to the Landbot example. They have used emojis extensively and a GIF to display their happiness because the reader liked their content. 

I was testing it as an example, but if someone actually conversed with that friendly-bot, they would have undoubtedly given their email for the newsletter. 

That’s the power of visual components.

b) Use short messages:

Don’t throw a chunk of information for the user to read or act upon. For e.g.,

Instead of:

Bot: “What’s your name and Email address?”

User: “ABC and [email protected]”

Try:

Bot: “What’s your name?”

User: “ABC”

Bot: “And email?”

User:...

This makes it a two-way communication, which gets the job done smoothly.

c) Insert non-lexical sounds:

To make the user feel like they are talking to a human, add non-lexical words to the conversations.



Source: Medium

Non-lexical fillers are extra words that come meaninglessly in case of verbal communication. Typical non-lexical fillers in English are: er, erm, um, mm, hm, h-nmm, hh-aaaah, hn-hn, unkay, nyeah, ummum, uh and um-hmm-uh-hmm.

That’s how we express & speak, and so should your chatbot.

d) Add voice:

What better than text to converse with your customers? Voice! Voice assistance adds a personality to your chatbot. This keeps the user hooked because of better UX.

2. Chatbots should be able to help users

a) Display categorized FAQs:

In most cases, users are going to throw in questions. Your bot needs to be ready to tackle those. One of the first implementations in your chatbot should be helping the bot anticipate the question and map to nearby FAQ. 

The image below is an excellent example of FAQ bot:



Source: Faqbot.nz

Such FAQs will grab an excerpt from an already existing content, either by answering or redirecting to the appropriate resource.

b) Forward: 

If nothing’s found in the FAQ categories and your conversation tree is not AI-powered or deep enough, asking the user to chat with an executive is the last option. Forward their problems and let human take care.

That’s the last resort.

4. Learn:

Learning from past failures can help upgrade the conversation tree. The more the chatbot fails to answer user queries, the more information a business has to implement for the next user.

The ideal computerized chatbot never forwards a message to human support. So, evolve the chatbot and take it further to the ideal scenario. 

Final words:

Although today’s chatbots are not perfect for replacing human support, it sure has made the job of the customer service department easier.

Cost-effectiveness, 24/7 service, Quick responses, and better customer experience are some of the things a chatbot brings to the table. Its effectiveness in customer support and lead generation is excellent. However, the effectiveness entirely depends on the content strategy and its impact on the user.  

What we discussed today are the fundamentals— the bases one must cover how to nail the chatbot content strategy. 

I hope you found this article helpful. If you have any additional queries, the comment section is open. We can discuss it right there. Until then, I bid you adieu.

Goodbye. Happy Chatbotting.