7 Steps to Building a Strong Knowledge Base
A knowledge base is the foundation of your automated customer support. Build it well the first time, and you’ll rarely have to touch it again.
In customer service, a complicated or unclear knowledge base will frustrate customers, not help them. What makes sense to you and your team might not be obvious to your customer, otherwise known as the curse of knowledge.
This guide will help you write knowledge base entries. Our process will reduce the frustration from having to redo your Q&As and tips so they make sense for your chatbot and support pages.
- What is a knowledge base?
- What type of data is provided in a knowledge base?
- Knowledge base examples
- 7 steps for building a company knowledge base
- Pro tips: How to write knowledge base entries
- Userlike: Your knowledge base system
What is a knowledge base?
A knowledge base for customer support is a centralized information hub designed to help customers find answers to their questions and resolve problems on their own.
It reduces support requests by answering common questions, pointing customers to relevant pages on your website and sharing how-to tips and troubleshooting articles.
A knowledge base and FAQ pages are nothing new. In fact, 81% of U.S. and UK customers expect businesses to have self-service options. However, the technology definitely evolved.
Modern knowledge bases are intelligent centralized platforms that power automation features on your website. Take Userlike’s AI Automation Hub for example: the information you upload to the central knowledge base is shared via your chatbot, FAQ pages and contact forms.
And vice versa: Your AI automation tools learn from customer input and feedback, improving your knowledge base data over time.
A knowledge base is also now available on a variety of messaging channels, such as WhatsApp and Facebook Messenger. Customers no longer need to visit your website to get the help they need — that’s what your AI chatbot is for.
It communicates via conversational AI, which makes your customers’ support experience a lot more enjoyable than clicking through dozens of help articles.
Premium guide to automation in support
Learn the top ways your business will benefit from AI-driven customer support.Download for free
What type of data is provided in a knowledge base?
A knowledge base should provide customers with the information they need to successfully use your product or service.
Here’s an overview of what topics matter most to customers. You can also use this list as categories for organizing your knowledge base.
- FAQs: These are the simple, common questions customers ask, plus your approved answers. Depending on your business, these questions will highly vary in topic and quantity.
- Troubleshooting: Step-by-step guides that help customers solve specific problems they may encounter with your product or service. Here, you can link to tutorials or videos that instruct the customer step-by-step on how to fix their specific issue.
- How-to guides: This section can help guide a customer with setup, installation, best practices and more. You can use images, GIFs and other rich media to illustrate your guide. It may also include quick tips and links to articles or videos on your website.
- Product specs: Customers need to make sure your product is compatible with their needs. This section includes all the necessary features, functions, dimensions, materials and performance details of your products.
- Policies and procedures: Here is where you outline your company’s billing and refund policies, data privacy and security, legal compliance and more.
- Business details: This is a helpful section for sharing your store locations, service hours, an “about us” summary and other details relevant to the public about your business.
Knowledge base examples
Alternate knowledge base
German tech store Alternate uses Userlike’s AI Automation Hub to provide a Smart FAQ page and AI chatbot on their website.
Alternate uses categories to organize its knowledge base entries. The answers are written using bullet points, making the content quick and easy to read. For certain categories, such as “Returns,” the user is automatically navigated to the appropriate website page.
Adobe knowledge base
Creative content platform Adobe provides an extensive, dynamic FAQ page in their Help Center. For account holders, they can use the Adobe Virtual Assistant for more personalized help.
Just like Alternate’s Smart FAQ, Adobe’s help center predicts the customer’s question as they’re typing. You can then simply click on the result that matches your request best, or click search for a list of result options.
This is an ideal solution for customers who may not always know the correct terminology or keywords to use.
Allianz knowledge base
Insurance company Allianz provides customers with several self-service and contact options on their website. You can access their phone number, agent router and live chat plus self-service links via a contact button in the bottom right corner of every page.
In their FAQ, their most popular search inquiries are organized into categories. Clicking these will either take you to designated landing pages that have more information, or a separate FAQ page if it’s a complex topic.
7 steps for building a company knowledge base
- Identify the questions customers ask often
- Categorize this information
- Create and collect content
- Choose a knowledge base system
- Test and publish content
- Promote your knowledge base
- Maintain your knowledge base
Identify the questions customers ask often
Unless your business is just starting out, chances are you have a good collection of questions customers ask often in chat, email and phone.
Collect these questions in a spreadsheet, which you will later organize. Be sure to check forums and social media as well to see what customers are saying about your business. Support tickets will also reveal what aspects of your product or service needs more explanation or a how-to guide.
Categorize this information
Organizing your entries makes it simpler for customers to find what they need and for your team to make changes in the future.
When creating your categories, keep them simple and logical. For example, “product details,” “pricing” and “troubleshooting.” Once you’ve outlined your most important topics, organize your list of Q&As you collected before into these categories.
Create and collect content
Sometimes showing is better than telling. In your how-to articles, images and videos help illustrate your points and show the customer how something is done.
Take screenshots, collect links to helpful videos and guides and create diagrams and more that may help customers digest your information quicker.
However, remember that media is just a supplement to your text. Using more than two or three images may work in a FAQ article, but will overwhelm the user when chatting with your bot.
Choose a knowledge base system
When choosing a knowledge base system, look for one that uses AI and is ideally connected to customer messaging software.
For example, Userlike’s AI Automation Hub is a self-learning knowledge base system connected to our live chat software – that way you allow customers to easily transition from your FAQ page or bot to a chat with your support team.
To create entries, you can either write them manually or import your prepared spreadsheet.
The AI uses Natural Language Processing (NLP) and Natural Language Understanding (NLU) so customers can type whatever they want and the AI will interpret their question and deliver an accurate response every time.
Test and publish content
Before going completely public with your knowledge base system, we recommend testing it with a small group of customers. This will ensure the content is helpful and easy to understand.
With the AI Automation Hub, you can even test the knowledge base before you buy. To get a good impression, you can create a few categories and activate or deactivate them as needed.
This is also a helpful feature if you want to add new categories or only want to display information in a certain timeframe, such as weekends or after business hours.
Once you’re satisfied with the look, presentation and content of your knowledge base, go live on your website and the channels your customers use most, like WhatsApp.
Promote your knowledge base
Make your self-service portal stand out! On your website, you can use your AI chatbot in proactive mode so it offers help to new site visitors. Make sure to also include a link to your Help or FAQ page on your homepage.
We also recommend promoting your new self-service features on your social channels. Invite customers to chat with your new bot or check out your knowledge base articles on common topics.
Maintain your knowledge base
Keep your knowledge base up-to-date by regularly reviewing and updating the content based on customer feedback and changes in products or services.
The AI also assists with maintenance by improving the accuracy of its answers over time based on customer conversations and feedback.
Pro tips: How to write knowledge base entries
This is the most important step in the knowledge base creation process. Your entries are displayed across all modules, so they need to meet certain criteria.
Entries need to not only be helpful and relevant to your customers, but also fit every module format.
Use full sentences, not keywords
Never use single keywords as a question. For example, “shipping” could mean several things, such as “shipping costs,” “shipping times,” “shipping status.”
“Shipping” works as a category, however. You would then create different entries for each shipping topic, like the examples above. But again, use complete sentences, “What are your shipping costs?”
This makes it possible for the AI to extract the intent and answer the customer.
Create separate entries for questions even if the answer is similar or the same
Mixing questions, even if they’re similar and have the same answer, is confusing for the AI. You’ll inevitably run into problems.
For example, these two questions have a different intent, but essentially the same answer:
Intent 1: Why are you raising my electricity bill?
Intent 2: I don’t accept the billing increase.
To make it easier for the AI to learn the intent, create separate entries for these questions. Plus, your answers may change in the future. Combining intents may lead to you needing to rebuild the knowledge base and will reverse all the progress your AI made.
Keep answers consistent for all automation modules
Since the knowledge in your database is shared across modules, the information should be concise with just enough details.
Longer answers with multiple images may make sense in a FAQ page, but not in a chatbot message.
Condensing your answers to apply to all modules will save your team from spending extra time rewriting the same answers to apply to different formats. Of course you can create different answers per module, but it will make your setup far more complex.
It’s better to have many specific and targeted questions, like in the Tushy example above, than a short collection of long help articles. It’s easier to digest and reads well in a chatbot conversation.
When needed, link to blog posts or landing pages that go into more detail about a topic.
Userlike: Your AI knowledge base system
Are you convinced by the power of an AI knowledge base system? Or do you need to see it to believe it? No matter which side you’re on, it doesn’t hurt to test the technology for yourself.
At Userlike, we provide you with a central knowledge base system that supports three modules: an AI chatbot, Smart FAQ and Contact Form Suggestions. It’s connected to our live chat and customer messaging software, so everything you need for customer service is in one place.
With our tips above, you can create intents that the AI can use right away. The AI Automation Hub is the self-service portal that meets customer’s needs and the most modern support tool available.
Userlike makes getting started with a knowledge base easier than it looks, so don’t hesitate to sign up for a free trial to get to know our product. If you like what you see, reach out to us in the chat on our website. We look forward to hearing from you!