Solving Multi-Silo Search for Customer Support Documentation living in Confluence, Zendesk, Google Drive with Lessonly Knowledge

Knowledge silos pose a productivity inhibitor for customer service and support teams because they create confusion as to where the answers to customer questions live. Consider some of these use cases.


Product and engineering teams have popularized Confluence (and other parts of the Atlassian suite of tools) as a wiki, of sorts, for documenting technical information. When customer support agents are confronted with a complex technical inquiry, it helps if they can access the same documentation that the Product or Engineering teams access to ensure that they are delivering the most accurate information to customers. Customer support agents frequently access documentation in this silo to handle the more technical questions coming in from customers.


Zendesk is a market-leading customer support software with a broad feature set that enables both customers and support agents the tools that they need to accelerate issue resolution. Most times customers can self-serve support using a Zendesk tool, but when they can’t and need to escalate the issue, they can tap an agent via creating a ticket. Zendesk also offers a knowledge base product that integrates with their core product to support agents in accessing the appropriate knowledge when dealing with customer issues.

Google Drive

While it is not optimized to retain and serve customer support documentation, Google Drive (by way of documents created by Docs, Sheets, and Slides) remains a popular repository for customer success teams, especially for smaller companies. While Google Drive is as good a repository as any, its broad usability makes it prone to getting disorganized and difficult to search. While it is natively a collaborative tool, it lacks features that are critical for knowledge management like verification and feedback/voting.

Understanding the Silo Problem

Using one of these services as a knowledge repository can satisfy your internal support needs, but the challenges come when you have multiple knowledge silos in-play simultaneously. This can create confusion as to where the most accurate and verified documentation resides.

If a customer support agent received an inquiry, a likely scenario is that the agent might waste time looking in the wrong places, or simply give up before arriving at the correct silo. Consequently, the agent will have to tap another agent for assistance and distract them from their own workflow, thereby reducing productivity. In an extreme case, the support-seeking agent might ask open questions in a communication tool like Slack, which could potentially disrupt and distract multiple employees, as they attempt to offer assistance.

Is it Possible to Abandon or Consolidate Knowledge Silos?

Consolidating knowledge silos sounds so crazy, it just might work, right? Well… probably not. It’s difficult to abandon a solution like Zendesk and move all of the knowledge to, for example, Confluence. Never mind that Confluence isn’t a fully featured customer support platform, but there is a laundry list of hurdles that make this an undesirable strategic choice. Imagine the potential costs related to re-training and data migration, to mention just a couple of considerations. The bottom line is, unless you have a systemic problem with a particular knowledge silo that can be rectified with another software solution, chances are you need to find a solution that adds a layer of productivity to what you already have.

But wait, is your Slack a silo, too?

A common problem that arises from multi-silo search is that when a support agent is having trouble locating knowledge, they will often turn to Slack to DM an expert or broadcast a message in a channel looking for help. While this is a disruptive behavior that reduces overall productivity, its often tolerated in the interest of expediting issue resolution.

But, the asking of questions in Slack in itself isn’t the worst consequence of this problem. What tends to happen is that subject matter experts respond with the knowledge required and leave it in Slack. They don’t transfer it to Confluence, Zendesk, or Google Drive, it just remains uncodified to eventually be asked again at some future date. This is a very unfortunate outcome and perpetuates the multi-silo problem.

There are a number of reasons why this practice is counterproductive:

  • Slack is not built or optimized to store knowledge – it’s a communication platform.
  • Slack’s native search tool is a poor choice for searching for structured data such as knowledge or documentation.
  • Repeated questions and answers will all show up in search results obfuscating the “correct answer”.

Multi-silo Search Solved

Sure, it would be great if everyone could just get along and all jump on the Confluence bandwagon, but in many cases, that’s not practical. In fact, as we suggested, in most cases it’s just plain impossible.

To tackle the problem of multi-silo search, or in other words, to search all knowledge repositories at the same time, you need a third-party tool like Obie that is capable of creating a single source of truth for all of your knowledge search needs.

Obie enables support agents to search Zendesk, Confluence, Google Drive, and more, all with a single query. Agents can query directly from Slack or from our Browser Extension.

The reason Lessonly Knowledge enables search directly from Slack is because that is where so many questions are asked and answers are shared. It’s logical to bring search to the places where the most common workflows occur.

How Does Lessonly Knowledge Work?

Lessonly Knowledge connects with all of your knowledge silos by connecting to their search APIs and makes them instantly accessible without having to make copies of your data and storing them on another server (the most secure and private way!). Learn more here!

Podcasts, Certifications, and Ebooks Galore: Our Favorite Content Picks from 2020
Why Improving Employee Training Should Be on Your Company’s Resolution List

Like what you see? Learn more below.