Off Center
With 57 percent of customers calling contact centers for support after attempting to find answers online first (according to the Customer Contact Council), it makes sense (and cents) for your organization to look for ways to optimize web self-service.

Now, you’re probably getting tired of my take on everything, so I’ve brought in a knowledgeable guest – someone even smarter than I think I am. Following is my exchange with Ashley Verrill, a call center analyst and self-service expert with Software Advice, who was happy to discuss some of the winning online self-support practices of top customer care organizations.

What is the most important web self-service feature, and how does not offering this feature impact the customer experience?

I would say having a really effective search bar is crucial. Often, customers will land directly on an article because they typed their question into Google and your self-service content was among the results returned. However, if the article they navigate to doesn’t directly answer their question, you’ll want to provide them a simple “out” for quickly finding the right content – otherwise you risk them switching to a more labor-intensive channel, such as phone or email. Also, the longer it takes for customers to find the answer, the more likely they are to become frustrated. There’s nothing worse than landing on a self-service support homepage only to find a long list of FAQs or discussion threads. It doesn’t leave the impression that finding the right solution will be easy or fast.

I’ve heard (and read) you mention that it’s essential to offer an “escape” – an easy way for the customer to chat with a live agent if he or she can’t find an answer. But what about proactively “chatting up” users once they arrive to your site? Which of these is a more critical feature to offer?

It really depends. Proactively chatting with every website visitor can be really labor intensive – particularly for websites that experience thousands of visits on any given day. I would recommend a proactive chat feature only if it could be used to directly drive more revenue, like if you’re able to offer more consultative advice to new opportunities that could lead to a sale, or if existing customers have the potential to become return customers. Some very large organizations have the ability to dynamically offer proactive chat based on characteristics about the site visitor. For example, I’ve seen proactive chat solutions that can be programmed only to appear if the site visitor is recognized as being in their marketing “sweet spot” – based on data from their IP address, social, potentially mobile and other sources.

Another alternative might be having your contact center agents proactively serve up chat only to visitors who navigate to your support pages. This wouldn’t help you generate more revenue in a direct way, but at least it’s a way to more exclusively target those people looking for support…and it might improve the customer experience by not having them sit on hold or wait for an emailed reply.

What are the best ways to showcase an online community moderator, and how should he or she go about identifying customer service opportunities? Does having a community moderator impact the customer’s perception, or does it simply ensure that questions get answered?

I think having a community moderator is imperative. One of the biggest obstacles companies face in driving engagement in an online community is the perception that customers won’t actually get an answer, or at least they won’t get one quickly. So, if they dive into a discussion thread that matches their question only to find no one has responded, they probably won’t ever try the channel again.

For this reason, moderators should be present to proactively provide an answer if it doesn’t come from the community. As far as how long a moderator should wait before intervening, I’ve heard average response time ranges between 1-3 hours. Many tools provide features that can automatically notify a moderating agent if a community question goes unanswered. I’ve also seen a lot of communities that will add some kind of visual indicator to call out moderators so it’s really obvious. This usually comes in the form of a branded icon or color-coded indicator.

With 67 percent of customers preferring to find answers online (according to Nuance), what are some quick tips for improving web self-service with minimal effort?

I’d say first you need to make sure that your community is stocked with answers to your most common customer questions. So, take a survey within your contact center and identify the top 20 most popular questions. Write solid content that answers those questions, then add them to your community. Then, ask agents to record instances where customers said they tried to find an answer online. This will identify gaps in your content or improvements that could be made to the presentation of your content.

For additional info on web self-service, you can check out my post from a couple of years ago:
“Web Self-Service that Won’t Self-Destruct”. Keep in mind I drink more than Ashley does.

When it comes to social customer care (providing service and support via social media channels), there are two key practices that contact centers must embrace: 1) monitoring; and 2) monitoring.

No, I haven’t been drinking, and no, there isn’t an echo embedded in my blog. The truth is, I didn’t actually repeat myself in the statement above.

Now, before you recommend that I seek inpatient mental health/substance abuse treatment, allow me to explain.

Monitoring in social customer care takes two distinctly different though equally important forms. The first entails the contact center monitoring the social landscape to see what’s being said to and about the brand (and then deciding who to engage with). The second entails the contact center’s Quality Assurance team/specialist monitoring agents' 'social' interactions to make sure the agents are engaging with the right people and providing the right responses.

The first type of monitoring is essentially a radar screen; the second type of monitoring is essentially a safety net. The first type picks up on which customers (or anti-customers) require attention and assistance; the second type makes sure the attention and assistance provided doesn’t suck.

Having a powerful social media monitoring tool that enables agents to quickly spot and respond to customers via Twitter and Facebook is great, but it doesn’t mean much if those agents, when responding…
  • misspell every other word
  • misuse or ignore most punctuation
  • provide incomplete – or completely incorrect – information
  • show about as much tact and empathy as a Kardashian.
  • fail to invite the customer to continue his/her verbal evisceration of the company and the agent offline and out of public view.
All of those scary bullet items above can be avoided – or at least minimized – when there’s a formal QA process in place for social media customer contacts. Now, if you’re thinking your QA and supervisory staff are too busy to carefully monitor and evaluate agents’ Twitter/Facebook interactions with customers (and provide follow-up coaching), then what the Zuckerberg are you thinking even offering such channels as contact options? I’ve said it before and I’ll say it again (and again, and again): If your contact center isn’t ready to monitor a particular contact channel, then it isn’t ready to HANDLE that channel.

Customers don’t applaud organizations for merely being progressive. If Toyota came out with a new automobile that ran on garbage but that had a 20% chance of exploding when you put the key in the ignition, customers’ response wouldn’t be, “Deadly, yes, but I might make it across the country on just banana peels!”

Social customer care is still new enough where organizations offering it are considered progressive. If your contact center is one such organization, are your customers applauding the strong and consistent social service and support your agents are providing, or is your center overlooking the quality component and losing too many customers to explosions?  

For more insights (and some irreverence) on Social Customer Care, be sure to check out my blog post, “Beginner’s Guide to Social Customer Care”. Also, my book, Full Contact, contains a chapter in which best (or at least pretty good) practices in Social Customer Care are covered.

Many call centers do a good or at least passable job of forecasting and scheduling for phone contacts, but when it comes to the other channels they handle, such as email and chat, they show about as much concern as I do for Renaissance festivals or people who say “irregardless”.

According to a 2010 ICMI study on call center workforce management practices, nearly three in four centers (72.9%) handle customer email, but just over one in three (37.5%) of these centers actually forecast and schedule for email contacts. Likewise, only 49.5% of centers that offer chat as a contact option take the time to formally forecast and schedule for chat.

Failure to adequately staff the call center for text-based interactions is asking for trouble. While e-contact volume may not be nearly as heavy as call volume is in most centers, customers who receive inefficient and ineffective service online often cause even more problems for the company than unhappy callers do. After all, an online customer who is slighted or ignored today in the Age of Social Media is just a few mouse clicks away from creating a virtual public relations and brand nightmare for the organization. Angry rants on sites like Twitter and Facebook can turn an ignored email or a poorly handled chat session into an attack on your company that goes viral and infects many other existing and would-be customers. Ashton Kutcher might even take notice and re-tweet someone’s complaint to each of his 17 billion followers, some of whom are old enough to read.

Department of Redundancy Department

Possible brand damage and customer defection aside, failure to effectively forecast and appropriately schedule for email and chat can cause expensive increases in call center operational costs. Customers who become impatient with long delays in email responses are quite likely to contact the center via phone to check the status of their query and/or rip an agent a new orifice out of frustration.

So now the center is handling repeat contacts via multiple channels AND coping with increased toll-free costs and decreased agent availability due to long customer diatribes. Not a fun situation for anybody involved, unless your center has made it a habit of hiring masochists, which actually isn’t a bad idea if you are looking to reduce turnover.

I Never Said It Was Easy 

Accurately calculating expected email and chat volume, and scheduling the right number of agents not crippled by Carpal Tunnel to handle that volume, is a little easier these days due to advances in workforce management software; most WFM solutions today feature multichannel forecasting functionality. However, many centers don’t have the budget for such advanced tools, and those that do don’t always know how to use the technology for email and chat. Thus, in many multichannel call centers, WFM can quickly turn into WTF.

So, if you are fortunate enough to be equipped with a modern-day multimedia WFM system, don’t insult the system’s intelligence by using it only to forecast call volume. That’s like buying an iPhone and not using the data plan or watching an Angelina Jolie movie in only 2-D.

If, on the other hand, your call center can’t afford a window let alone an advanced WFM system, you still can’t just blow off forecasting for email and chat. (Unless, of course, your center handles neither; numerous studies have revealed that managers who forecast for non-existent contact channels are idiots or very drunk.)  Plenty of feisty “blue collar” call centers do a solid job of manual workforce management for text-based transactions – tracking how many (and what type of) email and chat contacts the call center receives every day, when these inquiries typically occur, and how long the average interaction lasts, as well as factoring in events (e.g., marketing campaigns, etc.) that may impact email and chat volume. Over time, such analysis has enabled these centers to uncover key historic trends on which they can base staffing decisions.

Of course, you could just ignore everything I’ve written like my mother always does and simply continue to chase after chats and emails as they arrive. Who needs best practices when you can just hope for the best? Sure, your customers will despise you, your brand will begin to disintegrate, and your agents will start bringing weapons to work, but at least you won’t have to spend time doing any nerdy staffing math or worrying about chat-astrophic deluges before they occur