With an influx of volume and unpredictable spikes, there are good options to better operationalize your agents in Khoros Care. This article highlights immediate automation opportunities and offers some strategy and guidance on how to proceed.
We hosted a webinar where we further discussed this topic. Check it out on-demand!
Welcome Responses (Learn more)
Before more advanced chatbot and automation use-cases, you should consider auto-responding when you are not staffed and setting engagement expectations when you are. Be transparent, and although customers may wish you to respond all the time, they will feel better knowing you’ve heard them and will get back to them. If there is a spike in volume, letting them know it might be five minutes until an agent is available will help keep them from getting as frustrated as fast.
If you have already integrated your own chat-bot into Khoros Care, you may be doing this today with our Automation Framework and Current TAR api. You should have your bot set expectations when it hands over to an agent about when the agent will be able to engage.
If you do not yet have a chatbot, Khoros has been planning to add the ability to configure “Simple Welcome Responses” in our platform. And because of the current situation, we are accelerating this work.
“Simple Welcome Response” will allow you to send an auto-response when customers initiate new private conversations with you. This will be a free feature and we are starting its Early Access program today with the ability to have auto-responses by language, work queue, and business hours . At the moment, Khoros will need to manage the configuration of the Welcome Responses for you and we expect self-service capabilities by mid-May.
To participate in the “Simple Welcome Response” Early Access program and request configuration changes, please contact your Khoros representative.
Automation - Triage and FAQ Deflection Chatbots
Beyond basic welcome messages, relatively simple chatbots can also have a profound impact on your customers’ experience. We recently released Khoros Bot, a chatbot intended for triage and FAQ deflection, but note that it does take at least a few weeks to put a new bot in place. Please contact your Khoros representative if you’d like to learn more.
The key is to always make sure the chatbot is not getting in the way of the customer getting help from a person. The chatbot can gather information to help an agent be more effective quickly, or if it can deflect frequently asked questions away from an agent to give them more capacity.
It is worth highlighting again, that a chatbot needs to get out of the way quickly if it doesn’t understand or a customer wants an agent, especially in these times of emotional stress.
Workflow & Resourcing
With an influx of conversations occurring directly or indirectly about COVID-19, update your tagging, routing, and prioritization rules to account for the additional volume. Additionally, configure shared dashboards to analyze the impact and measure the utilization of your agents.
If you are an admin that needs help enacting the below best practices, our product coaches are here to help.
Click here to book 1:1 time with a product coach to help you manage your increased volume.
Book time with a Product Coach
Supporting Product Doc: Creating Tag Rules
Create a conversation tag with keywords, phrases, and hashtags that specifically mention the coronavirus/COVID-19 pandemic.
(e.g. coronavirus OR “COVID-19” OR #covid19)
Here is a sample lucene rule for you to use.
Create a separate tag(s) for conversations where customers don’t mention COVID-19 specifically, but are posting about issues or problems related to the pandemic. These can be specific to your company, brand, or industry. (These may be tags you already have in place)
When creating the rule, check the option to retroactively apply the tag to all open conversations.
Work queues and smart views
Supporting Doc: Set Up Smart Views
To isolate conversations related to COVID-19 for responding and/or reporting purposes, create a work queue and/or smart view that includes the conversation tags you created above.
Use a work queue if you have a dedicated team of agents assigned to respond to COVID-19 issues and assign the newly created queue to the correct team(s); use a smart view and assign to the correct team(s) if these responding duties don’t fall to a dedicated team.
Supporting Doc: Set Priority defaults
Consider whether coronavirus-related content should have a higher, or dedicated priority. If yes, change your priority rules and map the newly created COVID-19 tags you created above to the new priority rule.
Shared Dashboard & Notifications
Create a COVID-19-focused shared dashboard and configure notifications. Below are several widgets we recommend including on the dashboard. Some of these widgets should be filtered to your specific coronavirus work queue/smart view, while others should display metrics for all work queues, ensuring you have a full picture of your Care operations:
Incoming Conversations - See how volume has increased/spike since the outbreak of COVID-19-related issues.
Online Agents - A live view of how many agents are online. You can also this widget by your dedicated COVID-19 work queue, if applicable.
Queue Backlog - How is your team keeping up with the incoming volume? Is your backlog staying steady, or consistently increasing?
Average TAR , Median TAR , Response Times , and Responses Meeting TAR - These widgets provide a breakdown of your response times. Are you consistently meeting/missing your SLA targets.
Agent States - Have your agents been logged in and available to work during their scheduled shift hours?
Top Tags - How many customers are posting to you about the Coronavirus, and how do those tags stack rank against your other care issues?
Live Workforce Utilization - (coming very soon) This new Analytics widget gives you insights into utilization percentage of teams and individual agents. If your team is fully utilized and still can’t keep up with incoming volume, you should consider adding additional agents.
CSAT/NPS - Filter this widget by your COVID-19 work queue/smart view to measure the quality of agent responses.
Volume Alerts - configure incoming posts widgets to act as an early warning indicator when volume crosses your desired threshold with intervals of 5, 10, 15, 30 min or 1 hour.
TAR SLA Alerts - Track when you are missing your SLA targets in a specific time period.
SHARE - Once your dashboard is created, share it with the appropriate members across your team and organization, including executives and leaders from other departments. Remember, you can share these Shared Dashboards freely, as Dashboard viewer users will not count toward your seat licenses.
New Response Templates
Create response template(s) for repeatable responses specific to COVID-19 and share them with your agents. If there are multiple responses, use the header “COVID-19 -” to organize them
If you have purchased Casual (Flex) User hours, consider using those named users now to help with capacity. Casual (Flex) users are designed to offer flexibility during times of high volume or crisis. They are purchased in bundles of 5,000 hours and calculated cumulatively, when users are active in Khoros Care. Remember, you are not limited in terms of the number of Flex Users -- only the number of hours they are active in Khoros Care.
If you wish to add additional Flex User hours, please contact your Khoros representative.
Supporting Doc: Getting around Manage View
Views can be created and shared within Manage View that target all conversations about COVID-19 for visibility and bulk actions
Configure Columns - Columns within Manage View can configured to view conversations/posts across specific channels that include keywords, hashtags, and pre-created COVID-19 tags
Bulk Actions - Within each Manage View column, users can bulk reassign to a different work queue, change priority, snooze, close, and reopen conversations
Share Manage View - Users with the Manage View Creator or Manage View Admin role can share the newly created view with other Manage View users by clicking the blue share icon in the top right corner
... View more
@jennyreineck You are right to be suspicious if you did not trigger the email. I think you are already proceeding wisely, but please do not engage with any unexpected email that asks for personal information or to set up a password.
... View more
The push-next queue configuration option is a powerful tool to help your team depend on system configuration to set a priority of inbound items and the structure and pace for your agents.
Why would I consider a different queue configuration?
"I didn't know what an efficient agent was until we launched this feature. What a difference!"
- Happy admin of a push-next team.
The below pilot plan is intended for administrators only. Note: the piloting process below can be utilized for any queue configuration. Claim next, push next, or available.
Configure and enable agent states (if not already configured)
1 “yellow” state required for push-next
Yellow states allow agent actions.
Select a group of agents to be your pilot team and select a target date
Ideally, your pilot users will be a sample of an existing team.
Inform these team members and supervisors that you will be piloting a new workflow starting on a specific date.
Train them on switching to “yellow” state at the tail end of their shift You can pass along this post that shows what to expect for common push-next configurations.
Create a “pilot” team placeholder
The team should be configured exactly like your other
Ensure the team has access to your yellow agent state
This document explains how to create teams
Create a dashboard or agent performance widget to compare teams
Create 2 versions of this widget:
1 widget focused on the pilot team, 1 focused on your original team
Take advantage of the agent performance widget and also the new agent activities metrics available . Don't know how to utilize these metrics? Sign up for a free 1:1 coaching session to learn more.
Specifically, be sure to include “In-focus inactive” time and compare between the teams.
Move agents to your pilot team
Run pilot for 48 hours - 14 days
Compare the performance of the 2 teams using the dashboard you created
Connect with agents around their confusion
Look at your agent performance widgets from Step 4 , comparing your original team and team piloting push-next
... View more