In this week’s Data Byte, we explore customer service. We discovered that email is still king when it comes to customer service, but picking up the phone is still preferred to chat or social media combined.
In this article, I first cover the most interesting results of our research. Further down, I talk about the two common pitfalls of research that we fell into, and how you can learn from our mistakes and avoid them in the future.
First, our findings.
When Customers are Looking for Answers, They Start on Your Site (and Then Probably Ask Google)
This will come as no surprise. The first thing most of your customers are going to do when looking for the answers to their question is reference your own materials. Whether it’s through marketing pages or in depth tutorials and help documentation, the best way to minimize customer service calls is to provide the answers where they can find them.
Interestingly, older customers, those aged 45 and over, are more likely to turn to a company’s help community and/or forums to find the answers to their questions, while anyone aged 44 or younger will turn to search engines like Google when they can’t find what they’re looking for from your main site.
For Contacting Customer Service, Email is King
I’ll admit, we asked this question for selfish reasons. SurveyGizmo is conducting research internally to measure the success of our own customer service team, and one of the big questions we have is: how do our customers want us to talk to them?
To find out, we asked respondents to rank their preferred methods of contacting customer support.
Across age groups, email is king, followed closely by online communities and help forums, with the trusty telephone coming in third.
What surprised us most is that this pattern holds mostly true across age groups. Yes, I know you hear that Millennials and Gen Yers hate picking up the phone, but it turns out, when it comes to customer service, the phone is still considered one of the best ways to get in touch with support staff. Chat support just doesn’t compare.
For Award Winning Phone Support, Whatever You Do, Don’t Transfer Me!
We also asked respondents to vent their customer service frustrations in an open text section.
The most common pet peeve was being transferred too many times, with multiple mentions of hating having to re-verify account information every time a transfer occurs.
This frustration is entirely understandable. No one likes repeating themselves again and again, particularly when they have to explain their problem and recite their account information.
Almost as annoying as being transferred, our respondents hate being placed on hold.
For companies hoping to improve their phone-based customer service, keep these peeves in mind. Your goals should be to:
- Keep transfers to a minimum by empowering the person who answers the phone with the training and authority to help.
- If the customer does need to be transferred to a different department or escalated along the customer support channel, pass as much information as you can to the next support person. Don’t make the customer repeat themselves.
- Keep hold times to a minimum. SurveyGizmo has been having success with setting a call-back time, essentially scheduling a one-on-one, in depth service call for a time that is convenient for the customer. This way, no one has to sit on hold.
What Not to Do When Conducting Research With Panels
This week’s research also reminded us of how easy it is to fall for some of the most common pitfalls of conducting research with panel respondents.
On the surface, conducting your own custom research sounds like it should be easy. First, you decide what questions you want to ask. Then, you ask them! Your panel company will find the respondents for you, so it’s easy to think that all you have to do is sit back, relax, and watch the insights roll in.
Unfortunately, this isn’t the case.
In addition to keeping the principles of Great Survey Design in mind, you have to plan carefully and avoid the most common pitfalls of surveying a broad audience – particularly when your questions revolve around a highly specific topic.
We were reminded of this the hard way. Once we collected and looked at our data the first time, we realized that we had walked head first into two big mistakes.
- Never forget to include demographic questions.
- Don’t take any information for granted.
I will run through each of these points quickly, and show you how we fixed it.
When Conducting Research with Panel Companies, Always Ask Demographic Questions
Panel companies do not provide demographic information, and while you can request certain categories or traits – say, medical doctors in their forties – the panel company isn’t foolproof. Even when you have requested a specific target demographic, you should always include disqualifying questions that will exclude anyone who does not fit the bill.
Demographic questions are important even when you’re surveying a general audience.
This is where we made the rookie mistake.
Even if you think the questions you’re asking don’t need demographic information, ask them anyway. You may be surprised by how much more detail you can see when you start breaking down questions by age and gender.
The first time around, we omitted these important demographic questions, and found that our results were lacking. The second time through, we were able to glean so much more from the data.
Don’t Take Any Information for Granted
This is especially important to keep in mind when you’re conducting research about a topic that you know really well.
Yes, we made this mistake, too.
In our initial idea for this survey, we wanted to check the pulse of the customer service industry with a particular focus on the SaaS industry.
Now, we work for a SaaS company. We spend a lot of time thinking about the unique challenges that SaaS companies face.
Do you know what a SaaS company is?
While we live and breathe software here at SurveyGizmo, it turns out, not everyone else does.
The biggest problem we faced with the first version of our survey is the fact that people don’t know what SaaS is. And, even among those who claim to know, when it came time to fill in our unprompted recall short answer questions, it became clear that people did not know what we were talking about.
Which leads me to the lesson: don’t assume your respondents know what you’re talking about.
When you design your survey, ask yourself. Would the average person know what this is? Do I need to explain to people what we’re looking for?
In our case, we weren’t able to ask our panel company for “users of SaaS products” as a target demographic. In our first try of this survey, our disqualifying logic wasn't thorough enough to weed out the majority of people who really didn't know what a SaaS company was.
The second time around, we improved our disqualifying logic and added additional language that defined the exact kinds of services that we were asking about.
That said, a survey llike this still required a lot of manual work to remove incorrect answers from the data. I spent a lot of time cleaning responses one by one by analyzing open text questions.
While the process of data cleaning in this way is time consuming, putting in the extra effort does wonders for data accuracy.
Start Improving Your Customer Service Today
What's the best way to find out exactly what your customers are looking for? Asking them! If you're not already surveying your customers to find out how you can improve, you're missing out on insights that can take your service fromm good to great.
And, learn from our mistakes don't forget the best practices of survey design and research.