Zendesk on Zendesk: How we provide multichannel support to our customers
This Zendesk on Zendesk is based on a webinar about how we use multiple channels to provide enhanced support to our customers. You can register for the recorded webinar here, Zendesk on Zendesk: How we provide multichannel support at scale. The webinar was hosted by Brian Reuter, a Marketing Manager.
Panelists include:
- Ben Collet, Director of Global Advocacy
- Ana Wiechers, Senior Manager of Customer Advocacy
- Melissa Burch, Knowledge Manager
- Jim Nestell, Senior Manager of Global Advocacy Operation
Our panelists will monitor this post for your questions. And be sure to share your own experiences as well.
See all of the Zendesk on Zendesk series discussions.
--
At Zendesk, an important aspect of supporting our customers is serving them wherever they need us. This means utilizing their preferred channels and quickly transferring between them. As the number of channels continues to grow, switching can increase handle times and impact customer satisfaction. Additionally, interactions in multiple locations can be difficult to report on and measure the overall effectiveness.
We aim to navigate these multiple channels seamlessly to match our customers' unique communication needs and keep their journey short and simple. To help with this goal, we've built a family of products focused on unifying and improving customer relationships across all channels. This article will explain how we use our products together to provide multichannel support.
Multichannel overview
This Zendesk on Zendesk will discuss how we use four of our Zendesk products together to eliminate channel barriers, while still engaging with customers.
- Support is our primary method of tracking, prioritizing, and solving customers' support requests. Support integrates with our other channels to make life easy for our agents.
- Guide adds a cost-effective self-service channel, where customers can hep themselves with product articles, user communities, and previous tickets they've sent to our support team.
- Chat provides us with a quick way of answer customers' questions and solving multiple requests at once.
- Talk creates a personal and productive bridge between our team and a customer. Talk helps us with complex tickets or escalating conversations.
Evolving your support with live channels
Two of the primary ways we help support our customers' unique communication needs is through our "live" channels. Our Chat and Talk products both help solve a customer's request actively and each provide their own benefits.
The most important step in using Chat in our multichannel workflow was figuring out how it would integrate with our other channels. You can read more on how we released Chat in Zendesk on Zendesk: How we rolled out chat, but ultimately there were two main metrics we needed to measure to see if Chat would be a good fit:
- What kind of customers would chat with us? For example, would they be repetitive visitors, more active during a specific time a day, or have similar questions?
- How could customer improve our agents' workflow? For example, how many questions could an agent take while maintaining a high satisfaction level?
To answer these questions, we used chat with a trial sample of customers. Working together with advocates and managers, we put tagging into place to track where chats were coming from, time of day, how many chats we were receiving, and if they were from the same customers. While collecting this data was important for predicting how our scaling efforts would look, to measure the impact on agents we needed to engage in conversations about their new workflow, including number of chats they could take at once and the quality of their conversations.
Our Chat channel is great for quickly answering customer questions to provide fast support, but sometimes there are situations where we need to talk it out. To fully integrate our live channels, we needed to decide when to move from a conversation in Chat to a call in Talk. We've found that not only is Talk the most popular, but it is also the best for critical situations where both agent and customer need to know they've been heard. Talk adds the value of tone to a conversation, which can make all the difference in a positive customer experience.
Using self-service as a strategy
At Zendesk, we consider self-service as a true channel, not only a tool we've implemented. An online search is typically the first step in a customer's quest for an answer, so it's important that our Help Center makes a good first impression. To continuously provide a useful Help Center, we focus on two main self-service efforts:
- Maintaining our growing knowledge base
- Creating great, up-to-date content
To help fuel these efforts in a cost-efficient way, we developed automation initiatives. Our first process, Auto-Archive, helps us identify information in our knowledge base that is not used by customers. Our second initiative, Auto-Flagging finds articles that need maintenance based on set criteria. This process automatically identifies issues that we can then raise within our advocacy or product documentation. In addition to these efforts, a part of creating great self-service is ensuring our content is available to every customer. To do so, we use a hybrid translation method to make our content available in five languages, while staying within budgetary constraints. Our most frequently viewed articles are translated via human translator and our lesser viewed articles use machine-translation.
Our Help Center is developed across multiple teams, so we can build a robust knowledge base with different types of content. To facilitate communication, we implemented a Knowledge Capture Service (KCS) strategy. KCS helps move the resolution process forward quickly and smoothly by encouraging advocates to link articles in tickets, so customers can see content related to their challenge. KCS also encourages advocates to recommend new articles or updates to existing ones.
To help bring our knowledge base to customers, we are constantly looking for new apps and tools to enhance engagement. By communicating across teams and identifying areas to improve, we develop ideas like Knowledge Capture and Answer Bot.
Setting agents up for success
The most important component of using multichannels for us is making sure our agents are provided with a detailed on-boarding process and resources for supporting customers across multiple locations. In short, a happy agent equals a happy customer.
In our on-boarding process, one of our focuses is ensuring that customers feel like they're having real conversations. While we have a robust technical training program, it's also important our agents understand when to pause those technical conversations and empathize with the situation instead. We use role-playing exercises to practice good, bad, typical, and escalating phone calls and tackle any "tone-deaf" problems. There are a few other tricks we teach as well, such as if you're smiling the customer can hear it, limit the on-hold time, and set expectations to guide your customer's experience. Using this training, we've found that the customers who call us continue to call us.
Giving 24 hour support isn't easy, but it's what we need to do for our customers. When handing-off tickets and chats between timezones, it's important that there aren't gaps in agent availability. We work together with different timezones to make sure the transition goes smoothly and customers are consistently supported. By identifying hand-off pain points, we can improve our process. The ultimate goal is to make the transfer as seamless as possible with the help of views, triggers, and macros to collect all of the necessary information, so a customer doesn't need to repeat themselves.
Another challenge of multichannel support is avoiding channel fatigue. We found that it makes more sense to break up what channels our agents are on throughout the day, rather than focus on one. At the same time, we don't want to switch our agents between channels too frequently or else it could interrupt their conversations and workflows. We found that implementing a batch schedule is useful for planning out what channels our agents will be on and diversifying our channels.
Determining your channel health
After we implement a new channel, we need to evaluate how the channel is performing. If you have already created a custom tailored model for your organization, stay with that. If you haven't, we find this model particularly helpful:
First, we take the inventory of our tickets and divide by our support budget. We call this number our Cost per Contact. We can then break out our costs into individual channels. This will determine how each channel contributes to our overall support budget. While this is an important metric to measure, we cannot only take this factor into account as some channels will naturally be more expensive than others. To fully evaluate all of our channels, here are some red flags we look for:
- If chat costs are higher than email costs, there could be some problems in our deploy of Chat.
- If phone cost are too high, there might be an issue in our self-service and chat experiences.
- Which channels have the highest handle times and lowest CSAT scores?
Let us know how you use multiple channels to support your customers! How do you navigate between your different channels? What channel do you find most useful for your organization? How do the different channels impact your agents' workflows?
-
Nice tips! Just a question, though: how does the auto-flagging work in Guide? Is it available to all customers? Where can I find more info on that?
-
so which support channel is the most cost effective for you?
-
Hi Nadia,
Auto-flagging is not available natively in Guide today. We implement this technique by running scripts against the articles we create inside the Advocacy (Support) organization here at Zendesk. The scripts identify the articles that match our specified criteria. Our publishers then review the list of articles and update each one as required. We are still in early implementation of this method and are still learning as we go.
We combine the use of auto-flagging scripts and manual flagging by our support agents to keep our articles up-to-date.
-
Hi Геннадий -
As a general rule chat is the most cost effective for our team for a number of reasons but mainly due to throughput. Our advocates can solve more tickets while on chat than any other channel in a given hour, with consistently high CSAT.
-
Thanks Melissa. It would be nice to somehow be kept up to date semi-automatically of outdated articles. It would be an extra reason for me to promote Guide over our current solution to my superiors ;-)
Any idea if this feature is going to be available to Guide customers in the (near) future?
-
Hi Nadia,
Melissa is checking to see about sharing her team's script with you. Automated scripting or manual intervention is the best way to identify out-of-date content now. We don't have a specific native feature planned for that.
We do have some features that might help with content curation, including our Knowledge Capture app (in early access now) and Answer Bot. Both of those have reporting dashboards that might help too. You need Guide Pro for both of these, and Answer Bot is an add-on to that.
We are working on another feature that might be able to recommend new content you can consider creating.
I hope you can convince your team to move to Guide at some point. Let us know if you have more questions about it.
-
Hi Nadia,
As Jennifer mentioned, I checked to see if we are able to offer a peak at the script we use. However, we aren't able to share that because as you might expect, it contains too much internal details to be useful for you. If it would be helpful, I'm sure that the Zendesk Services team would be happy to help with this.
-
Thanks Jennifer and Melissa. I understand that the script cannot be shared.
Someone from the support team has contacted me with some more questions in the meantime, we'll see how that goes :-)
-
That's good to hear, Nadia!
Thanks for taking time to provide more feedback via the ticket. I hope it is mutually beneficial. :)
Please sign in to leave a comment.
9 Comments