5 Practical Tips to Make Self-service Customer Support AI Ready

We are fascinated by the way businesses adopt newer and emerging customer service methodologies to WOW customers. Businesses are taking a 360-degree approach to deliver great customer experience by considering not one but multiple support channels.

Emails and phones are definitely not leaving the customer service world anytime soon and we see businesses incorporating various other channels like live chat and social media to provide faster and efficient service. With the holistic approach comes the difficulty to manage and scale these channels in order to make them available 24*7 (round the clock).

The ideal solution to fill the gap is to incorporate self-service methods into your customer support strategy by creating a knowledge repository and capturing the best practices and know-how. Self-service is all about empowering customers to find answers themselves without reaching out to support teams.

However, with the advent of chatbots, businesses can become sophisticated and more modern by introducing chatbots into their self-service strategy as well as differentiate themselves from competitors.

Now, the important thing to remember is that, to make a bot work, you need to power it up with information. More the information, better the performance. Hence, for your bot to function at its best, your knowledge base needs to be up to date and better optimised for search. You need a content strategy for your knowledge base more than a bot-based support strategy.

Here are our top 5 tips to build a great knowledge base that will make your chatbot smart and efficient to handle your customer’s questions:

#1 Understand Your Customer Personas

The key aspect of creating good knowledge base content is to make it tailored and relevant for your audience. By doing so, you cultivate empathy for your customers and it leads to a stronger relationship with them. Ideally, you should know what issues your customers might run into and create relevant content makes your bot capable of answering their questions.

To achieve it, the first thing is to sketch out your customer personas — to put yourself in their shoes and to start creating content that will add value to them. This will help you capture certain patterns of their behaviour, their challenges and goals. Here’s a template that you can use.

Customer support AI Customer persona
Add a personal touch to your customer support

While you are creating personas, there are three things you need to remember:

— Avoid any assumptions about your personas. Conduct proper research to create a detailed persona and cross-reference with any existing data
— Involve your entire team to create the personas (what I’d call the wisdom of the crowd)
— Create not more than 3 personas; 5 is an overkill.

#2 Proactively Identify Customer Problems

Once you’ve nailed the personas, map out your customer journey and identify all the potential touch-points where they engage with your product/service. Here’s what the customer journey for a SaaS product looks like.

Customer Journey for SaaS products

Now, start curating the list of problems faced by your personas at each of the touch-points in their journey. Once you’ve done that, begin generating content for your knowledge base.

Next, scan extensively through your existing support channels/outlets like customer emails, social media, and chat messages for frequently asked questions (FAQs) from your customers. Create a repository of all these questions with the ideal solutions and map them under the respective touch-points.

This is a crucial step when you are implementing a bot-based support strategy because the information that’s collected will help in solving the cold start problem as the bot will already be trained on the data. Additionally, it makes the bot aware of your customers and the types of questions they usually ask, making the bot learn and become smarter over time.

#3 Structure Your Content

With the whole heap of knowledge base content that pops out from the previous steps, it’s time to curate them. You need to create a well-structured hierarchy for easy discoverability — not just for your users but also for the bot.

An easy way to do this would be to group the content based on the touch-points in your customer’s journey. Follow a content structure and map articles with relevant metadata allowing the bot to scan the content easily. It helps the bot to create a correlation between understanding the business context and mapping against user’s intentions.

Freshdesk knowledge base

A content guideline will come in handy to help your team maintain consistency. Here are four easy rules to stick to while creating articles:

— Keep the content simple, relevant, and consistent
— Write for the average user
— Use minimal technical jargons and buzzwords
— Include relevant visuals such as screenshots, GIFs, or videos

#4 Keep Your Content Up-to-date

You need to ensure you constantly update the content to make it accurate, current and relevant. With the help of Artificial Intelligence (AI) and Machine Learning (ML), a constant stream of learning will be consumed by the bot and it is vital to keep the content up-to-date to make the most out of the learning. And any piece of outdated content suggested by the bot will leave the customer confused and frustrated, hampering the relationship built over the days.

Here are fours steps to help your team scale content for your knowledge base (and your bot):

— Create a cadence to review and update the content in regular intervals, ideally every 3 months
— Define triggers like product releases or service updates to revisit the content and make it part of the process
— Make your teams use Knowledge Base extensively to learn about your products/services and its related updates
— Constantly monitor the feedback received on articles to make necessary changes to the content.

#5 Measure & improve

If you can’t measure something, you can’t improve it. The most important part of your chatbot strategy is to find out whether it is working for your customers and your business. You need to constantly monitor the performance of your bot and take necessary steps to improve it.

Ask yourself the following questions to understand the impact after implementing chatbots for your business:

— Do your personas engage with your chatbot?
— How many questions are being deflected by the chatbot?
— Do you groom and train your bot regularly?

Pro tip

Start training your bot right from the start/launch. Interact with it as much as possible so that it learns how to respond to questions about your product correctly.


As a business, you should create multiple layers of support and with the potential of bots, the first level of defence becomes stronger to deflect as many queries as possible. Your support agents can focus on responding to critical issues that require a lot more attention and empathy. Bots are definitely ready to play a key role in the future of customer support. And the future is now!

Try Freshdesk

Leave a Reply

Your email address will not be published. Required fields are marked *