Go back

3 Ways To Gather Topics For Your Knowledge Base

Supporting customers is a road paved with constant improvement. The products we support change and so are all the tools we use daily to deliver the best customer experience. But what about knowledge base content? What are the best ways to gather and select topics for new articles? How do support interactions play a role in it?

We’re exploring these questions and more with two teams that need to collaborate – knowledge management and customer support leadership.

Support Driven and Cleverly turned to seasoned customer support professionals to answer these questions and more. We asked active members of our Slack community to participate in a round table about everything knowledge management. Stay tuned to learn about approaches and ideas shared by our talented pros.

Representing the knowledge management perspective

Representing the customer support leadership perspective

In a previous article, we shared that collaborating with other teams is the key to be on top of your knowledge management game. The same goes for learning what features should be covered by your knowledge base.

Building relationships with the product and engineering teams would help the support team to stay on top of things, and not miss any new releases in terms of educating both customers and themselves.

When we release new features, we always look at knowledge base content needs to be updated or created to provide customers with information on the new functionality. Tadas Labudis, Founder and CEO at Prodsight
I am a member of our Product Release Communication team.This is made of reps from Product, Engineering, Marketing, and me. New releases are discussed, we have a chance to ask in-depth questions and take notes. Sarah Betts, Senior Customer Support Agent at Alyce
It might be something that comes up from our Product team when I'm learning about a new feature from them. They may mention something that isn't blocking the feature from being released, but that could be perceived as a pain point by consumers until they iterate on the feature later. I'd like to address that pain point to ease the user experience in that interim. Sarah Collins, Community Manager at NerdWallet

Surely, making writing support articles a part of your release circle is a sure-fire way to educate customers and build support self-service. 

We also should not forget about the essential element of any support function — support interactions. When customers raise similar questions every day, that's a signal for any support pro to dig deeper and work on some knowledge base updates. Or create a new Jira ticket :)

My queue time HIGHLY influences my work in the KB. Oftentimes, I will go ahead and create an article as I'm working on a ticket when I know the question is being asked a lot and an article makes more sense than a saved reply/macro. Sarah Betts, Senior Customer Support Agent at Alyce
Our product is highly technical and complex; our team handles a wide array of queries. What adds to the complexity is the fact that we operate in many different countries; hence we have to ensure that we abide by regulatory requirements specific to each of these countries. The queries we deal with can range from something as simple as how to use our dashboard or something quite compleX like an API integration with one of our partners. Our knowledge base is dictated by the most common queries that our team encounters as well as support really niche complex queries. Dimple Pattani, Frontline Customer Support Team Lead at GoCardless
That said, I think it's important to have articles that actually address real customer questions. At my last company, we built a text field into our support ticket form for "documentation gaps"-- any scenario where an article COULD have answered a customer question, but there was no existing doc yet. Having the support team explain the situation and what kind of article would have helped give me the context I needed to update our docs. Claire Littell, Customer Success Program Manager at Pendo
Our help Center provides comprehensive knowledge around every Airtable feature; as part of that, we provide answers to the most frequently asked questions that our customers have around each feature. Support interactions play a major role in this; we use various metrics and insights to determine the most common questions our Support team receives, and we build and support content to meet the needs of our customers. Jason Schulke, Support Content Operations Manager at Airtable

So far we have learned how to derive new ideas for support articles from questions and release logs. 

What about things nobody asks about? Somebody may encounter a blocker and never report it, simply losing interest in your product. Proactively thinking about these areas of improvement or encouraging desired customer behavior could also be a way to enhance your knowledge base.

Other articles are written to proactively address blockers that people might not bother to ask us about or to encourage behavior that supports company goals, like new registrations. For example, people don't often ask whether our product is free or how secure it is to link a bank account, but I want to make sure that information is easily accessible, especially for new visitors. Sara Collins, Community Manager at NerdWallet

It's also a great idea to use support metrics or other available data to navigate patterns and blind spots. We can analyze often used tags, macros, or saved replies to learn more about frequently asked questions. 

I also use 1:1 support interactions to help measure the knowledge base's effectiveness. I've added a specific tag to every macro whose subject matter is covered by a knowledge base article. Theoretically, we shouldn't receive any tickets from people whose question is answered by our knowledge base, so increased use of this tag indicates that the help center could be more visible or its content could be clearer. This support-related metric is an especially helpful barometer for the success of experiments designed to make the knowledge base more engaging or visible across the site. Sara Collins, Community Manager at NerdWallet
We believe the best source of information in this regard is data so we always follow a data-driven approach. First, starting with a strong labelling system that is actionable. Most teams use taxonomies/label systems that are not focused on the customer intent but are product-focused, depending on the company structure, and so on. First of all, having a labelling system that focuses on the customer’s intent is key. For example: "order refund request", "can't order a new card", "update payment method". Then using analytics based on volume per label, understanding if content exists, is relevant, being used, or needs updates.Cristina Fonseca, Co-Founder and CEO at Cleverly.ai

When working on the knowledge base content, the best way is to try a little of everything before learning what mix would work the best for your team and your product. Or maybe you have a completely unique approach that works wonders for you. Tell us in the comments!