How do you determine customer success? Do you need a group of metrics specifically chosen for your business? Or maybe there’s one perfect metric that tells us everything we need to know?
The Data and Research team from LiveChat got together on Slack to look for metrics that can determine customer success. How did the chat go, and what did they find out?
[Editor’s note: Lightly edited for clarity and correctness]
Jacob: Hello everyone. Before we start looking for the perfect metric, let’s decide on a use case. For example, customer service in a small ecommerce company.
Marek: I would extend it a little more to give us more space. Let’s have a use case of a small and medium ecommerce company.
Radek: Right. I would also specify a group of metrics and call them must-have metrics rather than the perfect one.
Jacob: Makes sense. So, how does our example company communicate with its customers? Because that should also impact the metrics. I guess the most popular options include chat (on-site and via Facebook), email, and phone.
So, let me fire the first shot. Having these methods of contact available, the first metric we could have could be the number of “successful” contacts. This can be anything from a chat that scored a good rating (if you have ratings available) to emails that end up with a customer purchasing something.
Marek: Shouldn’t we then unify a metric for successful communication? Is it more of a conversation that ends with a good rating or with a purchase?
Jacob: I think it could be both. Customer actions (purchase) are more telling than what the customer says (rating). If they vote with their wallet, it’s a pretty strong vote of confidence in the service. This is one of the few things I remember from “The Invincible Company.” :)
Radek: Also, if the metric will be used to assess the effectiveness of the given communication channel among customers, then we should also take both into consideration.
Marek: Definitely. We can consider both, but there is also a third option. Interactions with customers that don't end up with a purchase, rating, or any trackable feedback. They can still be considered a success because a customer’s question was simply answered. So, what about the number of answered questions as a metric?
Jacob: I think that would be a good metric for a customer service team performance report, e.g., to measure how many cases they can take on, but alone, it doesn’t say too much about the quality of customer service. I think it could be modified a bit to tell a better story. For example, how many customers left without getting an answer. I think that would be a pretty good metric for cases where bad customer service happened.
When I think about it, it’s hard to find one metric that would be enough to describe the CS situation accurately. I guess it’s as Radek said at the beginning. We probably should look for a group of metrics and pairs/combinations within this group, too.
Ania: I’m afraid calculating the “left without an answer” cases is not always possible and depends highly on what communication channel we’re talking about. But, regardless of the channel, what we should look at is the proportion between incoming vs. solved cases, an average response time (even if the acceptable thresholds differ across channels), and the overall customer satisfaction (rating, NPS, or whatever you use to ask for feedback).
Marek: It sounds to me like we are getting to the subject of the famous North Star Metric (NSM) - a single metric that directs company growth. Now, do we want to have a company-wide metric that correlates to its values, mission, etc., or do we need a reportable set of metrics showing a company’s status? When thinking about customer success I would go for the first one and that’s NSM.
Something I’ve found recently that gives a little bit more visual perspective about where to place the above.
Radek: I’m wondering if ecommerce is the right place for NSM to be applicable. Let’s say in streaming services, customers are there for one reason: to listen and watch content. In ecommerce, we have many more choices. Some customers might be there for warranty reasons, and some will want to purchase new products. And we also have those who are seeking information but will buy a product in different places (in the end, they’re not customers). But there might be a way to connect those actions under one metric.
Jacob: NSM is meant to improve the biggest goal of a company for some period of time. Once that’s done, it can change, shifting the focus to something else. So, we’re talking here about a perfect metric with an expiration date.
Radek: Yeah, I guess in today’s world everything has its own expiration date.
Marek: I believe, if well defined, it can stick as the metric for a long time, but not forever, as Jacob said. Overall, do we all agree that we define customer success from a customer’s perspective, which is bringing value to the customer? Because, you know, we could also define customer success from a company’s perspective and count it in cash.
Jacob: Yeah, I’d go for the customer perspective.
Radek: Customer perspective should be the one to focus on. If customers are happy, they bring revenue for the company.
Ania: Isn’t it a bit idealistic to assume we can choose just one point of view? I don’t think any company works on improving CS only to have happy customers and is not interested to see how it influences cash flow. And I wouldn't say it’s so obvious and always true that happy customers will bring revenue. Let’s not forget that, besides voting with their wallets, what is pretty easy to measure is that customers can also shape the company’s image through reviews and recommendations. Although the impact of word of mouth on revenue is harder to calculate and should be tracked in the long run, it should never be neglected.
Jacob: With NSM, you’re still looking at a bunch of different metrics. You just focus on the one you think will have the biggest impact first. You can use all the other metrics to validate your idea further and to get more details. As you said, no company tries to satisfy customers without looking at cash flow, but I guess it’s easier to increase revenue while focusing on customer success than the other way around.
Marek: Exactly, looking at cash metrics only and focusing the whole team around it can become a very nasty trap. Of course, there are other metrics that are more short-term. Those are called OMTM which stands for One (Only) Metric That Matters. Those become outdated much faster and their optimization are targets for single teams in the company.
Going back to the metric itself. I went through what we were talking about again, and what if our metric is counter to what Jacob said? We’d focus on successful interactions: questions with answers. Making it a ratio would be the first step: successful interactions/all questions.
Then we could adjust it to the interaction effect by doing the following: NPS, e.g., 1-10 (10 being the best), and purchase, let’s say, 0 or 10 (10 accounting for ending with a purchase). This could give us a pretty precise score. What do you think?
Jacob: Makes sense. In cases where there was no follow up (NPS/purchase) even though the contact was successful, we’d see a lesser impact. In cases where there was a purchase or high NPS rate, the impact would be greater.
One additional thing may be worth considering: repeated success. We could keep track of an additional metric showing how often, if ever, a customer returns. That would show us the success in the long run because continued success > one-time success.
Marek: Great, let’s wrap it up then. When trying to measure customer success, you should think about the value your company is bringing to the audience, not the other way around. In most cases, it won't be a standalone metric, like NSM, that shows how well your company or a specific team is performing. And also, remember that the customer success metric is not one-metric-fits-all. You need to be aware of what customer interaction means and how you want to score it.
Radek: Definitely. Great chat guys!
Ania: Yes, thanks!
Jacob: Thanks, and let’s do it again soon!