Support & Success Metrics — Measure What You’ll Use; Use What You Measure
This post is based off a lightening talk I gave for a CSxCS Boston meet-up in May 2017. Thanks CSxCS!
There are a series of questions I think everyone should ask when considering metrics for Success & Support teams. (Full confession, I think everyone should ask these when considering metrics for just about any team, I just say “Success & Support” because Success & Support is what I know.)
What do I want?
Given what I want, what do I need to know?
What am I going to do with what I learn?
What is the best (least difficult, least annoying) way to learn it?
After answering these questions there are three simple steps: Gather the data, do what you said you were going to do with it, go back to the first question — Did you get what you want? Could any part of the process be better?
Here’s how this can work in the real world:
You want to know if your Support & Success team’s customer interactions are going well. But is that really the root of what you want? You want your customers to successfully use your product and be happy about it, therefore you have a Support & Success team. Because you have a Support & Success team, you want to know how well that team is doing.
What you need to know — Are my customers happy with my Support & Success team’s interactions with them?
What are you going to do with the data when you have it? Ideally, you’re going to brag because your team is awesome and gets awesome CSAT scores (and by bragging to the right people, you secure continued internal support for your team) and you’re going to work with your customers to turn around poor experiences and you’re going to use the feedback to help your awesome team train and improve. That’s three actions from one data set — pretty useful data.
Best way to get your data? For this, you need to directly ask your customers. You can measure if they’re using the product successfully by monitoring how they use it, but you can’t tell if they’re happy by looking at page clicks. Since you’re asking them for something, ask in the least annoying way possible while getting yourself sufficient data to act on. Value their time and their information.
Currently, I favor an Up/Down survey with optional comments when a customer support case is closed. I don’t ask after every interaction, just when my team thinks the issue is resolved. I also don’t ask for more than a single click, Up/Down reaction. This is quick, actionable, and uncomplicated, both for me and for my customers.
Act — Respond to the negatives (and some of the awesome positives if you have time). Report out to your organization. Train. Improve. Repeat.
You want to get your customers the help they need, when they need it (because you want happy, successful customers).
You want to know how much help your customers need and when they need it, aka your ticket volume and effort both overall and per average customer.
What are you going to do with your data? Staff appropriately. What else are you going to do with it? Track the reduction of customer need for support as your product and documentation incrementally improve (because of course they improve, right?). You can also report on how bad any particular product emergency is from a customer standpoint, by measuring the spike in customer contact during said emergency.
Best way to get the data? Have case management software. You don’t need to bother customers for this, just to track what you do. Track it as seamlessly as possible and you’ll have more time for other fun things.
Act. Plan your staffing as far ahead as possible. Remember that people are humans who take sick days and go to meetings and get distracted. Use case volume to argue for product improvements and then use reduced case volume to show that those improvements worked. Keep going.
What do you want? Your organization to succeed by having awesome, loyal customers.
What do you need to know? Do you have awesome, loyal customers? Is there anything you can do to get more of them? or to have them be even more awesome and loyal?
What are you going to do? OK, seriously, if you’re not going to do anything based on NPS data, don’t gather it. (Same rule for any data you directly ask customers for.) If you do gather it, use it to communicate with your customers about their overall experience. Use it to convert “passives” to “promoters”. Use it to drive improvements in products and user experience. Use it to track how your organization is doing over time and when you may need to change course.
Best way to track NPS? You have to ask, so ask in the least annoying way possible. (Consider every question in a survey that includes NPS — is it needed? Did you change the question to accurately reflect your product and your customers? If your company makes an HR platform, don’t ask people if they’d recommend it to family or friends.) Consider your population — does it make sense to ask everyone? a rotating, random sample? How often can you actually go through the data and respond? Who are you comparing yourself to? Should you consider seasonality?
Act. If your survey isn’t anonymous and you have something constructive to say, respond to your detractors directly. Even better, respond to your passives to discover what would turn them into promoters. Consider responding to your promotors to thank them. Collate the comments and find ways to improve your products and user experience. Get that information to the right people. If your organization makes a great change based on someone’s comment, let that someone know.
What do you want? Your organization to succeed by having awesome, loyal customers who keep re-upping and maybe even upgrading. (Why yes, this may be somewhere you can get some of the same benefits as an NPS survey, without doing an NPS survey.)
What do you need to know? Are your customers sticking around and buying more? Are they leaving? If they are leaving — when, where, and (possibly) why?
What are you going to do with what you know? Make improvements directly targeted at the points where your customers leave. Reinforce the points at which they buy more. Step 3: Profit.
Best way to track? Probably? As automagically as possible with targeted requests for customer input. Look at what you know already, based on web analytics, customer inquiries, sales & cancelations. Find your most likely inflection points and determine if you can ask your customers for feedback at those points.
Act. Improve, re-analyze, repeat.
I could go on (SLA!, Tagging!, Sentiment analysis!), but should sum-up instead: Do use data. Don’t use it blindly. Think, Plan, Measure, Act.
The end (until I improve this entry based on data).