Is Chat Ready to Grow Up?

My credit card recently expired, which meant updating accounts such as Netflix, cable, etc. where the card was used for automatic payment.

It was a strangely inconsistent process from company to company.

My cable company's website wasn't working, so I had to contact a live agent. I opted for chat, hoping this would be the most convenient. It was not a great experience.

I waited for two minutes to get connected with an agent. Once the agent came online, her responses were so clearly templated that I sincerely questioned whether she was a human or a bot. She replied "I am a robot :) beep beep," which tells me she was (probably) a human.

There was also a long lag between her responses. I later learned she was handling three chat sessions at the same time. The chat session took 10 minutes just to update my credit card expiration date.

This experience is similar to what many chat customers encounter every day. If your company offers live chat, it's time to do better.

chat.jpg

Chat Satisfaction is Down

Customer satisfaction for chat was down in 2017. 

The LiveChat Customer Service Report 2018 shows an overall chat satisfaction decline from 86.35 percent in 2016 to 83.54 percent last year. Comm100's Live Chat Benchmark Report 2018 showed a similar decline from 84.06 percent down to 80.68.

Two major pain points are average wait times and average chat length, both of which I experienced in my chat with the cable company. Here are the averages for 2017 from the LiveChat report:

  • Wait time: 51 seconds 
  • Chat length: 11 minutes, 34 seconds

Companies should drive those numbers way down if they want to modernize their chat channel. Keep in mind that customer satisfaction with wait time is a function of the actual wait combined with how that wait was spent.

 

Best Practices to Speed Up Chat

I reached out to leading chat providers such as Comm100, LiveChat, and Zendesk to learn some best practices.

 

Leverage Chatbots

Comm100 launched a chatbot in 2017. Its clients were able to use the chatbot to handle 20 percent of their chat volume, on average. Let's go back to my credit card example. The agent added no humanity (recall I initially couldn't tell if she was human) to what was really a simple transaction. A chatbot could have handled the same issue with no wait time and no lag. This also would have freed up the agent to assist another customer with a more complicated issue.

 

Integrate Customer Data

Szymon Klimczak, CMO at LiveChat suggests leveraging available information. "Building a positive experience is all about using the available data appropriately. Hence, knowing your customers really well seems to be the key to success." For instance, I was signed into my cable company account when I initiated my chat session, so the chat program could have been programmed to recognize that data and allowed that chat agent skip the four questions she used to verify my identity.

 

Route Intelligently

Tony Sandhu, Comm100's Customer Success Manager, suggests making sure chats get routed to the best available agent, rather than just assigning chats on a random or round robin basis. "Long wait times can be eliminated by using intelligent routing rules that automatically route requests based on departments, customer value, or competency of agents."

 

Share Content 

Caitlin Henehan, Zendesk’s Vice President and General Manager for Chat, describes how sharing helpful content in chat helped one client solve issues faster. “The ability to send video, screenshots, and links to help articles via chat has allowed our customer TeamSnap to reduce their work time per issue by 20% and increase customer satisfaction."

 

Limit Simultaneous Chats

The number of simultaneous chats an agent can handle at one time should be capped to improve customer satisfaction, reduce lag time, and prevent errors. The exact number varies from company to company and should be determined by working closely with your agents to observe what's most efficient. In my cable company example, running three chats at once created a negative experience because the agent took took long to respond and wasn't able to inject any personality into her initial responses.

 

Take Action!

Chat has the potential to be a really great channel if used correctly.

Evaluate your own chat function from the outside in by conducting a mystery shopping exercise to experience what your customers experience. Look for opportunities to apply best practices to reduce customer wait times, increase customer engagement, and solve problems faster.


Three Big Trends from LiveChat's Customer Service Report

Chat software provider LiveChat recently released its 2017 Customer Service Report. The latest report is a treasure trove of data on chat. Here is a snapshot of what was included:

  • 13,500 companies that use chat
  • 22 different industries
  • 24 billion website visits
  • 235.7 million chats
  • 11.1 million tickets

A number of important trends jump out. You can read the entire report here or skim below to see what I think are the top three.

(You may also wish to read past blog posts on chat trends here and here.)

Trend #1: Tech Support Loves Chat

Tech companies had the highest customer satisfaction for chat among the industries covered in the report:

  • Web Hosting: 92.91% satisfaction
  • IT Businesses: 92.66% satisfaction
  • Software: 91.17% satisfaction

One of the natural advantages of using chat for technical support is agents can have a nearly real-time conversation with customers while also sharing helpful links and screenshots.

The slight delay between responses actually creates another advantage. LiveChat was kind enough to quote me in the report:

The natural latency of a chat conversation gives customers built-in time to implement the steps required to fix their issue.

 

Trend #2: Chat Demand is Increasing

The average LiveChat customer saw a 4.11 percent increase in chat demand in 2016 compared to the previous year.

Forrester's Kate Leggett offered one explanation why customers may be increasingly choosing chat—they want to avoid the phone.

Organizations can quickly connect customers to an agent with the right skills to answer the question without them having to navigate an arduous interactive voice response.

That high demand leads us to the third big trend.

 

Trend #3: Companies Are Struggling to Keep Up

The report noted businesses experienced a 31.15 percent increase in the average number of monthly tickets.

This number stands out because a ticket is created in LiveChat when a chat agent isn't available (due to high demand or after hours) or a customer's issue can't be resolved on the initial contact.

Meanwhile, average first response time is 56 seconds, which can feel like an eternity to a customer who is waiting for assistance.

Customer service expert Shep Hyken noted that making chat easy for customers is much more important than the length of a chat.

When I get on live chat, whether it is a live agent or AI [artificial intelligence], it doesn’t really matter whether it takes 6 or 12 minutes. If I’m having my question answered and there is little friction between the time I start and the time I get the answer, I will be completely satisfied.

 

What You Can Do

These trends all suggest a few key actions for contact centers that offer chat:

  • Empower your agents to resolve issues in the first conversation
  • Staff chat adequately to prevent long wait times and excess tickets
  • Take advantage of chat's natural latency to provide value-added service

Inside LiveChat's New Ecommerce Holiday Shopping Report

Is your customer service team ready for the holiday season? 

A new report from chat software provider LiveChat provides some insight that can help you prepare. The report studied 8.8 million customer service chats from 1,400 ecommerce companies in 2015. The results were further segmented into off-peak (January - October) and peak (November - December).

The data reveals when customer chat volumes will be high and provides some tips for staffing to meet that demand. You can also use this data to improve service quality.

You can read the full report here or browse through a few highlights below.

Managing Volume Is Tricky

LiveChat's report reveals that the average company experienced a 63 percent chat volume increase in November and December.

That's not too surprising. The stat that really jumped out was that chats lasted an average of 1 minute longer in November and December (peak season). That jumped up to 1:23 longer on Cyber Monday.

Many customer service leaders will plan for the volume increase. The big question is whether contact centers will be prepared for longer chats.

The extra volume and longer chats means customers may have to wait longer for a live agent. LiveChat's data shows that it took 7 seconds longer for agents to respond to chats during peak season.

A 2015 report from Zendesk shows that customer satisfaction cliff dives after customers wait longer than 50 seconds for a chat response, so that extra 7 seconds can be a problem.

 

Predicting Peaks and Valleys

Not surprisingly, Black Friday and Cyber Monday were the two busiest days of the year for chat volume.

What happened next was a little surprising.

Chat volume dropped 53 percent the Saturday after Black Friday. For some contact centers, this may have meant being overstaffed if they planned for volumes similar to Black Friday.

Cyber Monday had 10 percent higher chat volume than Black Friday. But, look at LiveChat's graph and you'll see that volume dips but remains relatively high throughout the week. It's only the weekends when chat volume declines.

Image Source:  LiveChat

Image Source: LiveChat

You may have similar peaks and valleys in your volume. It's a good exercise to look at your historical data to see when you need extra staff and when you don't. This is old hat to large contact centers, but smaller contact centers tend to keep a more static schedule from day to day.

 

Customer Satisfaction Dips Slightly

The increased volume takes a minor toll on customer satisfaction (CSAT) as survey scores dipped an average of just .7 percent during the November - December peak season.

Here are the averages from LiveChat's report:

  • 87.1% - average CSAT during off-peak (January - October)
  • 86.4% - average CSAT during peak (November - December)

Scores declined a bit more on the two busiest days:

  • 85.8% - average CSAT on Cyber Monday
  • 84.5% - average CSAT on Black Friday

To me, these stats can be misleading. It may seem that a small drop is OK, but this is a missed opportunity. 

Holiday shopping is a stressful time for many customers. You can distinguish your company if you can find a way to alleviate some of that stress through fast, helpful service.

After all, you want your customers to remember you during the off-peak months of January through October.

 

What Should You Do?

It takes the average contact center three months to hire and fully train a new agent. That means it's probably too late to add staff now if you haven't already started.

But, there are still things you can do. I shared three tips for handling increased volume without adding extra staff in LiveChat's report.


Zendesk 2015 Q1 Report Reveals Live Chat Insights

A new report from Zendesk reveals new insights into live chat.

Zendesk is a leading provider of customer service software. Their 2015 Q1 Benchmark report analyzed live chat data from 2,261 companies.

This post summarizes a few of Zendesk’s more interesting findings. For example, live chat delivers the highest customer satisfaction rating among contact center channels with a 92 percent average. 

You can download the full report here.

Faster Problem Resolution

Implementing chat in a contact center can lead to faster problem resolution.

Companies that introduced Zendesk’s Zopim live chat software saw a significant drop in API and Web Form contacts. These are channels that customers typically find when searching for support on a company’s website. 

Moving to chat allows customers to resolve their problems faster. Zendesk recorded an average chat duration of 10 minutes and 35 seconds. Most companies take an hour or more just to respond to contacts submitted through an API or Web Form.

The grainy graph below shows the percentage of contacts by channel prior to implementing live chat in green. The percentage of contacts after implementing live chat is show in yellow.

Source: Zendesk

Source: Zendesk

Slow Reply Times

There’s still room for improvement when it comes to reply times.

When a customer initiates a chat, they typically have to wait for an agent to respond. In some ways, this is like being on hold when you call a customer service line.

Predictably, customers prefer a faster response.

The average time to first reply is 1 minute and 36 seconds. However, Zendesk noted that customer satisfaction begins to drop once first reply time crosses the 50 second mark. 

Another large drop in customer satisfaction happens just after the two minute mark.

Source: Zendesk

Source: Zendesk

An Emerging Channel

Zendesk’s data also shows that live chat is still an emerging channel.

Most companies are seeing very low chat volume:

62 = Average monthly chats per company

Individual agents aren’t handling a high average volume either:

22 = Average monthly chats per agent

Companies with a low volume of chat contacts face a number of challenges:

  • Live chat agents typically need to support other channels too
  • Training can be difficult when volume is so low
  • Scheduling is tricky at low volumes

 

Whats Next?

Live chat offers two advantages over phone that companies could capitalize on.

First, it has the potential to deliver higher customer satisfaction. It’s a perfect channel for today’s multi-tasking, keyboard addicted consumer. 

Second, chat’s can be more efficient to serve. A typical agent handles two to three chats simultaneously. This means chat agents have the opportunity to serve more customers per hour than phone.

The key to getting this right is response time.

Live chat has to offer a response time of less than 50 seconds to be competitive with phone. Otherwise, the channel will simply continue to divert contacts away from slower channels.