When it comes to big data, everyone wants the same thing at the end of the day: insights that can have a positive impact on the business.
More and more, that has come to mean insights that are immediate in addition to being actionable.
This column will take you through the thought process necessary to make the best use of the myriad sources of data about your business. Let’s start with some definitions, and discuss “immediate” and “actionable.”
What does “immediate” mean to you? This is going to depend on two things:
- The type of business you are running. Are you making screws or running a hotel?
- Your desire to affect an individual experience, such as an upset guest, versus your desire to improve the overall experience for everyone. Are you responding to an individual review/Tweet/Facebook comment right now or are you analyzing all of the data on a monthly basis so as to make systemic changes to your business?
It’s critical not to confuse the two goals (immediate versus systemic) and never try to make systemic changes based on a single comment. In other words, don’t get caught in the trap of “nothing matters but the last person I talked to.” I’m looking at you, Sales Team.
Next, what is “actionable?” Actionable has two important components:
- It must be something you can actually do something about. It’s not actionable if you can’t act.
- It must have a measurable outcome. It is just as important to be able to tell if you were successful as it is to actually do something.
Last, what is “customer satisfaction?” This all comes down to the question of measurability. If you don’t have a strong method of measuring customer satisfaction, you will not be able to tell if you have more satisfied customers, and you won’t know if your use of “immediate, actionable data insights” made any difference at all. Some common examples of how to measure customer satisfaction are Net Promoter Score, Review Star Rating, Social Buzz, Returning Customer, and Dollars Spent.
Now that you know the timeframe, what is actionable, and how to measure whether or not your actions made any difference at all, let’s go through two separate cases, each with a few options as to how you can handle the situation.
Case 1: Responding to the Complainer
One of the most interesting things about social media is how the balance of power is reversed, or at least companies seem to think that it is. Anytime someone complains on Twitter, writes a bad review, or leaves a nasty Facebook message, there’s a knee-jerk response to do something about it.
Let’s consider the case of @DavidTheComplainer. David, not surprisingly, is a verbose Twitter user with a large following who enjoy his never-ending supply of pitchforks, tar, and feathers for any company that dares to make a mistake. @DavidTheComplainer is having a bad experience with your brand and tweets right @You.
What do you do?
Option 1. Respond quickly with a direct fix to his problem. The obvious advantage of this approach is that, if you can fix his immediate issue, you look really fantastic and even a professional hater like @DavidTheComplainer will be happy with you.
But you better have the resources to do this for everyone, 24/7/365, because by doing this, you will be setting expectations that you are going to respond this quickly to everyone. If you don’t, you are going to upset someone else. In addition, you may not have an immediate fix, and may not even want to solve his particular problem.
Option 2. Respond that you have heard him and will get back to him with a fix. The advantages of this approach are that you buy yourself some time to understand the issue and what your options are. But keep in mind that the amount of time is limited by the sort of poor experience he’s having. If it’s something like a flight, you have very limited time. If he’s having trouble assembling something (unless it’s Christmas Eve), you have a bit more time.
The disadvantages of this approach are that it might not be enough for David, as he’s right in the middle of a bad experience. And if you say you are going to get back to him, you have to follow through or you look bad.
Option 3. Do not respond, but make note and see if there is a consistent pattern in need of a fix. If you have limited resources, you can’t fix everyone’s problem immediately. This is one of the great issues of the social age, this expectation that, if you complain, the company will drop everything to fix your problem. In addition, this approach gives you time to see if David is just a serial complainer and if you have other complaints in the same vein. However, this approach does involve the risk that David will feel ignored and may go on to complain even more.
Choose your plan ahead of time. You have to weigh your resource constraints against the necessity of being high-touch with individual customers. Decide now which approach you are going to take and evaluate the change in customer satisfaction over time to see if you need to shift your strategy. If you are the Ritz-Carlton, you are famously in touch with every customer demand – you have to be right on top of every possible area of dissatisfaction. If you are making electrical staples, you may not need to be quite so quick to respond.
Case 2: Evaluating Customer Feedback over Time
“Immediacy” can have many different timelines. In this case, the meaning of “immediate” is driven by the demands of your business. Hotels have maintenance spend, but they are not going to bring in a truckload of beds based on a single day’s worth of reviews and customer comment cards. “Timely” is probably a better word here, enough data over enough time to show a trend, while being a short enough timeframe to act on complaints with a minimal number of people being upset.
I suggest a daily or weekly report of anything negative. This allows you to respond to individuals relatively quickly if there’s a big problem and to avoid getting sucked into making a big decision based on a short timeframe. Then, do a quarterly report of all trends to line up with your planning schedule. If you have multiple weeks of the same negative things happening, you can tighten up this timeframe.
In order for the reporting to be actionable, you can’t stop at how customers feel. You need to dig into why they feel this way. Star reviews, net promoter scores, and even sentiment analysis are all just clues. You have to use other techniques (that are beyond the scope of this article) to find the root cause and decide if this something you can actually fix or ignore.
Now that you have the information on what is making people unhappy, here comes the hardest part: actually acting on this information. Now you have to take all of the other factors into consideration – like competition, cost, time, and revenue impact – and make a decision about whether to actually reprint all your menus, order those new beds, or put a warning on your staples reading, “Do not staple hand to wall.”
You have determined your metric for customer satisfaction. You have decided on your strategy for both immediate complaints and for acting on long-term trends. Now, make sure that your metric for customer satisfaction is going up. If it isn’t, then you need to re-evaluate your process. Are you missing a critical data source? Did you pick the wrong way to respond to immediate complaints? Maybe that customer-service touch is more important than you thought. Is it worth the additional expense to try to improve customer satisfaction? Sometimes, in certain markets with emerging competition or disrupting factors, just staying the same is a “win.” Whatever the case, you certainly have better visibility now, and it’s just a matter of tweaking the process until you are happy with the direction your business is moving.
Seth Redmore is CMO at Lexalytics. He is currently 3-for-4 in startups, including co-founding and selling a networking company to Cisco Systems. Seth was into data before it was big, and has been dealing with unstructured data problems for over 10 years now. Seth is responsible for all marketing inside of Lexalytics, from arguing about taglines to filling sales’ insatiable need for leads. Seth has a degree in Chemistry from Carnegie Mellon University, which occasionally makes him dream of being back at a lab bench when he’s debating whether the logo should be teal, cobalt, or more of a robin’s-egg blue.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise, plus get instant access to more than 20 eBooks.