Turn Slack Conversations into Knowledge Your Teams can Leverage

Posted by Alyssa Verzino on Jun 7, 2019 1:29:31 PM

 

Screen-Shot-2019-01-16-at-8.13.16-PM
We all know that Slack is awesome for knowledge creation. The ability for a team member to send out a question and quickly get a response is incredibly powerful -- we refer to this as transactional communication. For example, it’s great if you’re a support rep and your company has launched a new product. If you’ve been at the company for some time, you need to quickly figure out what’s changed and what’s still valid. If you’re a new rep, you need to quickly get up to speed on all the offerings. Yes, there might be training available, but people can only keep so much information in their head and searching online can be far too time-consuming. With Slack, you can simply post a question in your team's customer service channel, and get almost immediate results. Often, these types of channels have subject matter experts (SME) monitoring them and part of their job responsibilities is to answer these types of questions.

These types of transactional communications are great but aren’t necessarily the most efficient for the parties involved. The person asking for information doesn’t want to have a conversation; they want the best, shortest, and fastest answer to their question so they can then move on. Like a transaction, once it is completed, they move on to the next thing. The information they’re looking for may have already existed in the channel, but finding it would have meant going back through the logs (or, in the case of Slack, the history) and look for the old transaction to find the information. This can be pretty painful and time-consuming, and they’re just not incentivized to do that.

For the person answering the question, this isn’t the most efficient process either. They’re typically faced with a slew of questions and often times, people ask for the same information repeatedly. Notice that I didn’t say the same question repeatedly; this is because people rarely ask for information in the same way. For example, one Customer Service Rep may ask, “What security standards are our product in compliance with?” Another Customer Service rep may ask:

  • “Is our product SOC 2 compliant?”
  • “How do I answer a security question and make sure the customer is comfortable that their information is safe?”
  • “Compliance?”

Each of these questions has an answer based on the same information. With today’s communication, since there’s a human answering every question, the responder needs to tease out the intent behind the question. This is pretty time consuming for an SME. The SME probably intends to capture the request or answer pairs, so they can reuse their answer (think cut and paste). There are good intentions here, but folks get interrupted and this almost never happens. And so they end up repeating this process multiple times a day. I recently had a conversation with an avid Slack user, and they told me their group solved this by having someone periodically go through the channel history, capture the information, and store it somewhere. This is the modern (and time-consuming) version of screen scraping.

As I said earlier, Slack is great at creating knowledge. But what it’s not great at is helping you capture and reuse that knowledge. Chatbots have tried to solve these problems, but have only had limited success thus far. To clarify, a bot in this instance is an application that lives in Slack and offers some service as part of the conversation. For example, a bot can log a ticket in Salesforce, or look up a customer issue in Zendesk. Bots are created so that if you know exactly what to do, you can ask the bot to do it for you. The request needs to be explicit and very task-oriented, to allow the UX to complete the task.

Some of the challenges bots face when trying to capture knowledge include:

  • Not all conversations in a channel are knowledge you actually want to capture.
  • Subject Matter Experts (SMEs) have to manually cut/paste the same answer to the same questions asked multiple times and in multiple ways.
  • Teams are used to asking their questions within the channel or DM’ing an SME for information. It requires a behavior change to get them to talk with a bot.
  • Today's bots don’t help the SME if no one talks to them
  • The history is lost
  • Bots today struggle with getting the right information back to the requestor.

Today’s bots don’t address the issues that teams have with capturing and reusing the knowledge created in Slack. At Talla, we have heard/seen these issues multiple times. With the latest release of Talla we’ve added the capabilities to address these challenges.

Stop Leaving Knowledge Stranded In Your Slack History

Talla can scan your Slack history either once or on a periodic basis to uncover questions and answers to add to your knowledge base. This way, when related questions are asked in the future, the information will be readily available and can be answered automatically. Talla goes a step further and will ensure that information is relevant to your business, weeding out things like “Can we have a company dog?” or “What’s for lunch?” (Yes, we’ve actually seen these types of questions.) If a question was asked that didn’t have an answer, Talla will prompt you to create that information.

Let Talla Proactively Capture and Answer Questions for Faster Information

There’s always a learning curve to getting your team to adopt a new workflow, and getting them to ask Talla first when they need information is no different. You have the ability to set up Talla to proactively listen in on certain Slack channels of your choosing. Even if users don’t ask Talla questions explicitly, it will capture these questions in the knowledge base and try to match up the answer. Talla can even automatically respond to the user within the channel if you’d like.

Capture Knowledge and Reuse It for Better Productivity

A common challenge at many organizations is that SMEs don’t have a reliable way to capture the knowledge being created and reusing it. Some ways that Talla can help:

  1. Talla can be a SME’s digital coworker, allowing the SME to add information to the knowledge base automatically in Slack as part of responding to a user. This means no more cutting and pasting information, and the information can be easily reused in the future.
  2. Since Talla understands the intent of a questions thanks to AI, information can be asked for in many different ways.
  3. SMEs can get the information from Talla and post it to the channel or DM the requestor back personally. If Talla doesn’t have the right answer, then the SME can train Talla so the next time it will know the right answer.
  4. The SME can answer as Talla, so the behavior change to ask Talla instead of the SME starts to be reinforced to the users.

Give Users the Fastest Path to an Answer

As users start to see Talla answering their questions, they are more inclined to start asking Talla. Users can ask Talla their questions directly and Talla will answer as quickly as possible.

So, stop losing knowledge and making your teams less productive. Use Talla to help you turn data created in Slack into knowledge that can be reused by your entire team. Stay tuned for my next post on the Data Science behind how Talla can do this.

New call-to-action

Topics: AI, Slack, Digital Workers, automation

Subscribe to Email Updates

Recent Posts