You know you are onto something when you have 800 people register for a customer service webcast. And no, it wasn’t about social media. It was about an even a hotter topic, believe it or not: Web chat for B2B support. Tuesday’s webcast, sponsored by CitrixOnline, included the “Top 5 Web Chat FAQ’s” from me, and a great case study about deploying Web chat by Judy Bendy, End User Computing Practice Director, Long View Systems. For those of you who missed it, here is a link to register for the OnDemand version of the webcast: http://www.focus.com/webcasts/why-chat-matters/
There were lots of audience questions we didn’t have time to answer live, and I’m attempting to make up for that with this post. Here are some of the questions we received:
Q: Can chat be used effectively as a sales tool as well as a support one? Do you have any strategies for implementing this successfully?
Absolutely. Proactive chat is commonly used for sales. The best example is retail websites, including Amazon. If you are browsing high end products, or have a certain dollar amount in your shopping basket (at one point the Amazon threshold was around $300), the customer is prompted with a chat option, such as, “May I answer any questions for you?” In this way, a sales agent can help push the customer toward purchase by answering any questions they have or reassuring their fears.
But sales can also be successfully incorporated into a support chat interaction, easier than by phone or email. For example, instead of just asking the customer if they are interested in your new service option or an upcoming training class, since you are connected via chat you can push a webpage to the customer advertising the offering, or take them to a product page for something you want to try upselling to them.
The key, and this is a topic for a whole other blog post, is incorporating soft sales training into customer service training, so support employees begin to see selling as part of the service transaction. And as one company told me after successfully implementing sales offers into support calls, “When agents hear ‘YES’ often enough, they stop thinking they are selling. It is all just part of servicing the customer.”
Q: Could you elaborate on the Proactive Chatting? You mentioned something about asking users on a website where if they are repeating some actions, the website will prompt a question for additional service/chat.
Proactive chat uses a rule engine in your web self-service portal to detect when certain things happen. When a rule is met, the customer is proactively prompted with an option to chat with an agent. Rules can be defined to be very inclusive, or very restrictive. Common rules for proactive chat include:
- Customer value. For ‘platinum’ level customers, or other customers designated as high value, proactive chat is a great way to give these high profile customers the best possible experience. In fact, proactive support could be an option included in a premiere support package.
- High value purchase. As I mentioned in the first question, you can also proactively prompt customers for chat when they access web pages about expensive products, or when they have a certain value of merchandise in their shopping cart.
- Unsuccessful self-service. You can also get very creative in writing rules that detect when a customer is unsuccessful with self-service, and ask if they need help. Rules can be based on amount of time on the site, number of searches, number of articles viewed, etc.
Q: Do you think chat works better in call center environments (customer service, taking orders etc) vs. IT Help Desk or technical support?
Clearly Web chat works fantastically for high volume, low complexity environments like communications, retail, and financial services. These quick one-and-done questions work great with chat, and keep the customer in their channel of preference–the Web. But, we have seen case studies from multiple TSIA members including EMC, Symantec and Cisco, that Web chat lends itself for most technical support issues, as long as they aren’t TOO complex. We discussed on the Webcast the importance of having an escalation procedure in case a chat question becomes too complex and needs to be moved to phone.
Q: I support a very technically demanding software package, has Chat been shown to cause questions which are not fully thought out… i.e. email would require users to take long and think about their questions.
This is not a complaint I’ve heard before, but I can understand the concern: instead of attempting self-service, a lazy customer would just reach out to you via chat. I would argue that if the customer wants you to do all the work, they probably would have just picked up the phone and called you anyway. The only thing I can offer, and I’m not saying this is a best practice, is to only reveal the chat option on the website after the customer has at least made some attempt at self-service, such as a search of the knowledgebase, or after viewing a knowledgebase article. Though customers may be confused why the chat option is there sometimes and missing sometimes, at least that way you are pushing them to at least try self-service before chat.
Q: Is it possible to have chat agents at different physical locations responding to chats from the queue?
Absolutely, and this is become a common scenario with more home-based workers. The same push/pull options for chat queueing work whether all the agents are together in a contact center, or working remotely. With the ‘push’ method, the next chat in queue is popped to the next available agent. This is a basic feature of Web chat platforms. With the ‘pull’ method, agents monitor a queue of pending chat interactions and pull the next customer in queue into a chat conversation.
Q: One of the questions i have time and again been faced in IT Support is that :: Does chat aids reduction of Phone AHT or not? since resolution by way of typing may add more time for resolution. So will it take more agents to resolve same set of queries while doing it on chat.
We discussed this a bit on the webcast, and I admit there are some differing opinions on this topic. In my experience, the AHT (average handling time) for a chat interaction is slightly longer than for the identical question handled via phone. Why? There are often lags in the conversation while the customer is browsing the web or on a phone call and ignores the chat dialog for a few minutes, and as you note, typed conversations take longer than spoken conversations. However, for consumer centers with less complex issues, experienced agents can handle multiple concurrent chat sessions, which boosts productivity and more than makes up for individual issues taking a few minutes more.
Q: using chat for a customer sat survey vs. eamil or web… is the acceptance rate higher… and does it depend on the age range?
I don’t have a lot of data on this, but I do know that customer survey response rates are higher the sooner the customer is surveyed after the interaction. The advantage of Web chat is you can prompt the customer with the survey at the end of the chat session, while the interaction is fresh in their minds. Not only is the response rate higher because you are surveying immediately–as opposed to sending an email survey 24 hours later–but the answers are more accurate.
Q: What are some common mistakes that companies make when they first start up a chat program, and how can those mistakes be avoided?
There are quite a few common problems, and we talked about some of them on the webcast. Here are a few:
- Dedicated chat agents. A 2 year trail at 2 large communication firms showed that dedicated chat agents had higher productivity than blended agents (those handling chats, phone calls and emails interchangeably).
- Chat skills. Chat requires a different set of skills than phone, so don’t assume your high performing phone agents will automatically be a whiz at chat. You need to incorporate chat into employee screening and training to find agents with the right skills for chat.
- Volume management. This is the single biggest problem I see–companies put a chat option on every page for every customer and are overwhelmed with the volume. Instead, only put the chat option on a few pages, and consider limiting chat availability to certain customers or certain geographies to keep chat volumes low while you get out the kinks and agents have figured out the chat tools.
- Bad technology. You can get free chat tools from a variety of sources, but I don’t recommend them. Chat platforms for support have strong UI design, ties to knowledgebases and CRM, bulletproof security, auditing capabilities, supervisor monitoring capability, etc. Get the right tool for the job to start with.
Q: What are the key KPIs you enabled for measuring chat traffic in contact center?
- Average handling time/chat time is the most important metric, and be sure you have a supervisor dashboard showing all chat sessions in progress with elapsed time on each. The supervisor should be able to ‘drop in’ if a chat session is going too long, so they can see if the agent needs assistance. Also, agents should be trained to move the chat to a phone call if the issue is becoming too complex for typing.
- Satisfaction. Since you can prompt customers with CSAT surveys at the end of the session, be sure you are keeping an eye on the ratings and take immediate action if a low score comes in.
- Wrap up time. Chat conversations can automatically be written into CRM incidents, so there should be little wrap up time after a chat compared to a phone call.
Of course, the key is to capture enough metrics to identify averages and high/low performers, regardless of the metric.
Q: What are the keys to more efficient chat? Integration with KM tools is one thing that was mentioned, what is meant by that specifically? what are the other keys?
First of all, pick a chat tool with a UI designed for high productivity. Options like a response library for common problems and frequently used text can really speed conversations. Integration with KM is important so that you are presented with a list of knowledgebase articles automatically based on the chat conversation, and can cut/paste KM text into the chat dialog if necessary. And, writing the chat conversation into CRM is a critical piece to make sure the “360 view of the customer” is preserved.
Thanks to everyone who attended the webcast and asked questions! And as always, thanks for reading!