In this interview, our star-studded cast of speakers discuss the balance between automation and human interaction in customer service. They each weigh in on how to decide when to have a human engage with a customer versus when to fully automate the interaction.
Of course, there's no one right answer for every business—but segmenting customers into high touch, low-touch, and tech touch categories can be a great place for businesses to start with their engagement strategy. The other thing they also agree on? The importance of product research and data-driven solutions in proactively solving customer problems before they arise.
Give the full video interview a watch to learn how to best balance people vs. product in your engagement strategy, featuring:
- Amanda Kleha: Chief Customer Office, Figma
- David Apple: Head of Customer Success & Sales, Notion
- Tanya Littlefield: Director of Lifecycle Marketing, Litmus
- Joel Stevenson: CEO, Yesware
- Jonathan Kim: Cofounder, Appcues
Amanda Kleha:
Well, one example I can think of that we haven't tried to automate is helping people set up single sign-on because you have a lot of different ways, different products for single sign-on, and some of them involve custom solutions. So that's just hard to automate into a self-service experience. And then where there might be a major decision point, you want to make sure you have all the information that people want at that point in time, but not more because then it gets too much.
So if that particular moment can't be figured out in an online way, then I think maybe first having the ability to quickly talk to someone like via chat could be a solution for sure. I like to do customer journey mapping and really get inside the head of the customer from the moment they touch your brand for the first time and all the experiences in between. What do you like about that journey and what do you think could be better? And doing that on a regular basis, getting into a habit of that will help identify where you might want to either automate things or improve things with humans.
David Apple:
So in terms of deciding when to have a human engage with a customer versus having it be fully automated, I have a model that I use, which is a simple segmentation model of putting customers into buckets that are either high touch, low-touch or tech touch. The high-touch are people you want to be proactively engaging with and creating long-term relationships with. The tech touch is what I call just in time where you're just trying to help a customer achieve a specific outcome at a specific moment in time rather than building an ongoing relationship.
And the tech touch is fully automated. So once you've created that segmentation, each time you want to engage with a customer, you have to decide which bucket does that customer fall into and therefore how am I going to engage with them? So in some cases, for some customers, you always want it to be a human touch because they're your largest strategic customers and for some users you never want to engage with them because in the case of Typeform and Notion, there's millions of users and some of them are very unlikely to ever pay, and we can't afford to invest resources in engaging with them.
Initially the easiest way to do it is based on firmographics. It would be the company size, their potential to spend, stuff like that. What we've also done at Typeform is once we aligned as a company on which persona was the one we were going to focus on, both from a product marketing and sales and customer success perspective, we started practically engaging more with that persona so that we could have more of a feedback loop of what's working for that persona and what's not working for that persona. So even if they didn't fall in all the other characteristics that we normally care about company size, we would still engage with them to get that learning.
Tanya Littlefield:
Products can help solve problems before they happen. So ultimately, if you're doing the right amount of research and you are very data driven, you have an idea of where people are having trouble, you're fixing that in the product proactively and regularly, or you're adding in-app messaging and prompts to get people back on their way. I think the personal approach is tricky, especially when you have a large base, but you should always be doing that on the side as well. So I'd say for every 50 pieces of survey response that you get from somebody, you should be having one or two phone calls as well.
Human interaction helps qualify the interactions or the feedback that you're getting from users. So you might get a sentence or two, or maybe you just see a trend when you're doing something that is data driven and in app. But when you actually get on the phone with that customer or really dig in, have a conversation, you are going to learn a lot more about what that problem is and how you can get them to the right solution.
Joel Stevenson:
We try to think of it as self-service first. So wherever possible, we would like the product to do the job of educating the customer, only because it's just so hard to get somebody's attention otherwise, whether it's via email or a phone call, it's just very hard. So our approach tends to be let's invest in making sure that onboarding is good, that people can find the information that they need to solve their problem. If they can't and they do want to call us, let's make it real easy for them to call us and contact us, versus having to fight around it to find a number or not having a properly staffed call center so that people can't get through to us. We're not always great at all those things, but those are our goals is to be able to solve it first in the product and then be ready to help people in whatever way that we can, treating them as real humans is part of that process.
Jonathan Kim:
I really like this concept of having people be the unscalable version of what you can then eventually scale. So we tell customers who are really early stage startups to do their onboarding manually. Onboard every single customer, get to know what works, get to know what doesn't work, and then put that messaging, put those repeatable best practices into your product to help them scale. Using the one two combination helps you understand where you can continue to improve. And then once you know that you've mastered that, put it back into the product in a way that's scalable.
