The Support Automation Show: Episode 6

The Support Automation Show featuring David Apple

In this episode of The Support Automation Show, a podcast by Capacity, Justin Schmidt is joined by David Apple, Chief Revenue Officer at Zingtree. They discuss David’s journey from mechanical engineer to Industry thought leader, customer guide, and sales marketing professional, his leading business principles for customers’ success and support automation, and tips on boosting performance with the help of tech tools.

Listen now!

Justin: Welcome to The Support Automation Show, a podcast by Capacity. Join us for conversations with leaders in customer or employee support who are using technology to answer questions, automate processes, and build innovative solutions to any business challenge. I’m your host, Justin Schmidt. David, Good afternoon.

David: Good afternoon.

Justin: Thank you for coming on The Support Automation Show. Where does this podcast find you?

David: I’m in San Francisco and thanks for having me.

Justin: Absolutely. You recently joined Zingtree as their CRO. Why don’t you start by telling us a little bit about yourself in your journey to Zingtree and then ultimately a little bit about Zingtree and what it does?

David: With pleasure. A little bit about myself. I’m French-American. The first half of my career or a little bit less than half, I actually was a mechanical engineer and designed buildings, specifically green buildings, which is fun. After an MBA, I switched into tech and I’ve had the good fortune of joining two hyper-growth startups. First, in Typeform, I stayed for about five years. I was employee number fifteen there and built out the sales customer success, support, education, life cycle marketing, and operations teams and then moved to San Francisco with Typeform and then was then approached by Notion.

When they were very early stage, I was employee thirteen there and also responsible for filling out sales and customer success not support in that case. I was at Notion for just about two years before joining Zingtree, which happened under a month ago, as it’s still very new. Zingtree is a decision tree software tool where it’s very horizontal. You can use it for any type of decision but it’s true that we are focusing on the support area and specifically around agent scripting and knowledge-based/self-help solutions for the customers. My role as CRO, I’m responsible for sales, marketing, customer success, and support.

Justin: Very cool. Thank you. Here on The Support Automation Show, we like to start with the question of when you hear support automation, what does that mean to you?

David: My mind explodes in all the [crosstalk] that it could be. I think a lot of things that people talk about and that I was definitely very interested in exploring when I was running support at Typeform several years ago, was the whole AI, ML aspect of support automation. My impression at the time and still today is that, that’s still kind of in its infancy and people spend a lot of money there, but a lot of implementations may or may not actually be working as well as people hoped.

Now with Zingtree, my views a little bit biased but I think what’s most effective these days in support automation is automating or enhancing the performance, boosting the performance of the support agents [unintelligible 00:03:20] on calls, chat or email. That’s what I think about investing in when I’m thinking about support automation.

Justin: All right. That’s ultimately with the advent and proliferation of automation and knowledge worker for those on audio-only and making scare quotes, “White collar work.” The augmentation is really the big driver here, less so replacement. It’s– there’s freeing people up to do their best work and to be able to do the higher order tasks that humans are capable of, is the type of thing that you need to clear the runway to have that happen when people are bogged down with repetitive tasks and all that doesn’t happen.

Buried in my little spiel there is this notion that sometimes people find automations scary. Have you dealt with, in the change management, throughout your careers with growth and scale and having to bring automation in and now as someone working on bringing automation in other firms, have you come across any situations where you’ve learned how to manage the expectations of that fear to get to drive adoption and keep people from being quote-unquote afraid of automation?

David: Yes, I think that’s a great point and anytime there’s an initiative like that, I think the first thing to tell the team because ultimately, you do a POC or you experiment with the team or you do your research is, “Hey, our intention is not to let anybody go.” Nobody’s losing their job here. We just want to be even more effective in making the team more effective and also ultimately it creates a better experience for our customers. If we become better at first resolution time, if we’re able to solve more calls in the first interaction, customers will be a lot happier than if they have to escalation and deal with many different people. If we’re able to respond to tickets faster or to calls faster, people would be happier. I think framing it in that way and what our ultimate goal is, it takes a lot of the fear away.

Justin: Right. It’s interesting. The anecdote I always tell on this, I had a customer experience, customer support experience rather with Best Buy not too long ago. It was a very esoteric issue, like it’s extremely esoteric issue. It is one of those times where I thought the company and a business as large and as sophisticated as Best Buy, they do a lot to deflect my phone call ever happening in the first place. The fact that I had to call and I couldn’t get triaged by their voice assistant or whatever and I actually had to get to a person because the issue was esoteric. I had to get to an agent to actually speak.

I ended up having a really pleasant conversation with this woman while we were going back and forth on everything. It was just a good piece of human interaction and in a world where the person on the other end of the phone, where she has to constantly answer the same question a million times over and over again or do mindless things like lookup RMA numbers or whatever it is, you don’t have the bandwidth for that human connection. In addition to making people just more effective and efficient, you also get this benefit where the space is a little more clear to do what we humans do best and that’s connecting converse. I’m a hundred percent with you there.

David: Part of the automation can also be just directing the right customer to the right person in the first instance, rather than having them deal with two people who either aren’t qualified or don’t have the answer to that question. To your point, I totally agree with that.

Justin: Yes. One thing that is particularly interesting about your career path, David, is two things. I’d love to unpack this with you a little bit because people in support, whether it’s in the customer support/success role or if it’s internal and in employee experience, that there’s effectively support and success agents too. You’ve had the good fortune to experience scale twice now, which is awesome.

I would love to just explore this with you a little bit and speak to some of the things that you maybe were able to replicate it notion based on your experience at Typeform or maybe wish you would’ve known prior to taking the rocket ship to the scale that businesses like Typeform and Notion have enjoyed. Would you love to speak to some of the things you learned about automation and scale in support and if you could go back in time to before you joined Typeform would have told yourself?

David: There’s so many learnings because, like many people, I’ve made so many mistakes [chuckles] along the way. I’ve certainly learned at times. I think one of the big things I learned about scale is– it’s A, focus and B, focused on what. I think it’s really important, A, to be focused and B, to be focused on the right things. When I say the right things, one of the issues I had that I lost sleep over probably for years, [chuckles] certainly for months was the return rate at Typeform. Our monthly return rate was very high by the certain benchmarks that we use, which was for SAS. SAS in general includes enterprise and SMB, already if you break it down between enterprise and SMB, I would’ve lost a little bit less sleep. Then furthermore, the use case is very different, even if Intercom is selling to SMBs, Intercom is a tool that you’re using all the time. Typeform is more a one-off use case that you can create a survey, get a ton of value out of it but then you’re done until you have your next survey or maybe you’re even done forever. Ultimately, several years into my role at Typeform, I came across the retention rates of SurveyMonkey, which is our biggest competitor and one of the companies we looked up to, and realized that our retention rates were significantly better than theirs, and actually, we’re doing a great job. That was one thing. I think relaying that to support, support is often considered a cost center, which I’d love to talk more about. I don’t like to think about it explicitly that way. We’re always looking at cost per ticket, cost per dollar, AR, stuff like that.

I think when we limit ourselves to only thinking about it that way, we miss the opportunities of all the other value of having a support team, which is not only making our customers happier but also learning from our customers. I think that in the early days, it’s good to track those ratios and our efficiency, but we should be okay with a lower efficiency because our goal in the early days is not to make every dollar, optimize the bang for the buck, but rather optimize the learning so that we are growing in a more thoughtful, strategic way.

Justin: That’s great stuff there because support isn’t a cost center. Unfairly, things like HR, support, IT, these things are labeled cost centers. It’s the revenue organization or sales and marketing, those drive the revenue. Everything else supports that. That’s not true. Is there any advice you would give to someone in a success or support role to help make the case that that function is actually a revenue driver and not just a cost center?

David: Absolutely. The first thing, so I actually wrote a blog post that was coincidentally published today. It’s about the ROI of a CSM. It’s published in the Startup Grind. I have two stories about that. The main thing is, nobody can realistically calculate the ROI of a customer success team or customer support team. There’s a few reasons for that. One is, it’s impossible to know what the actual baseline would be without those teams. If we removed support, how would that impact our business? Well, that’s not even a question. We need to have support.

Okay, but then how does incrementally each support agent increase that value, compared to that baseline of not having a support team? The other one is, there’s a lot of additional value beyond just helping a customer solve their problem with a support or success team. You also get more positive word of mouth. You get less negative reviews. You can offer premium support and make that additional value on your higher-tier plans, which makes it easier to sell higher-tier plans. There’s a number of others. Go to my blog post.

Justin: We’ll link in the show notes.

David: My point is thinking about calculating a hard ROI number, it’s a little bit easier to do in a sales perspective, but doing that for customer success or support is not the right way to think about it. The more tactical or tangible story is from my days at Typeform, made a great data team back then that helped me with this. We were debating whether or not to continue supporting our free users. Typeform is a freemium business. Something like 96% of our users never paid. So only a small percentage paid. Of our support tickets, roughly two-thirds were from free users, one-third from paying customers.

We know we need to support our paying customers, but two-thirds of our costs were supporting free users that weren’t paying us at the time. What we did is we did an analysis of how much does a support interaction increase the conversion rate from free to paid of those free users. We found that a free user that interacted with our support team was about five times more likely to convert than someone who hadn’t contacted our support team or had contacted our support team but just not received a response within a reasonable period of time.

We were able to calculate, and I won’t bore you with all the details, but roughly that the ROI of our support team supporting those free users was above two to one. Therefore, it made sense for us financially, and obviously, we also know that for our brands, there’s a lot of value in anybody because Typeform grew mostly through word of mouth and virality. Just anybody, including free users getting access to support, was also valuable to us.

Justin: That’s really interesting. One thing you hear occasionally if you are listening to growth hacking or entrepreneurial podcasts, it’s the value of doing a cohort analysis on your customers. One of the common things is apps, pick on them for a little bit. You might get a big cohort of customers from some lifetime for $5, or whatever deal that you had on there, and you end up having this very expensive to support cohort that as time goes on, maybe creates a burden on smaller number of a higher dollar value future cohorts.

What you just touched on was something that I always wanted to hear on a podcast whenever people talk about that, which is like, “No, actually do the analysis on it and see if there is conversion opportunity with that large, thick cohort.” One of the things that old boss of mine back in my media days always used to say is that the clicks are the spice. It’s a Dune reference. Clicks are the spice. You got the clicks. You can start to do other things.

David: I play Dune. I hear you.

Justin: Trailer dropped today. It looks like it’s great, the new movie. When you have users, you can get data from those users. You can experiment with different conversion programs or whatever to move those users further down the lifecycle or into higher dollar value engagements or whatever it is. It’s great to hear that you all were able to leverage that into something, versus just saying like, “These are good brand advocates, but there are shackles around the ankles on this company. Let’s just stop supporting them or whatever.” Very, very, very interesting stuff.

David: Thanks. What was also interesting is around that time, we started ramping up the sales team, because for a while we didn’t have a sales team at Typeform. When we dug into those support tickets that free users submitted before they upgraded, they were absolutely not sales conversations. They were people coming in, trying to figure out if Typeform could solve their problems. It was very much a support-type interaction, not a sales-type interaction. A lot of companies will have a chatbot on their website or whatever, and it’s more like the sales team mans that, rather than the support team. That was another interesting observation from that whole experiment.

Justin: Very cool. When we’re thinking about bringing automation into organization, there’s a lot of options out there in the marketplace. There’s a lot of different approaches to whatever business case you have that you’re looking at automating. If you give any piece of advice to a business leader who’s looking at purchasing some sort of platform to bring automation into their function, what would it be?

David: I think it should start with what problem are you trying to solve because depending on the problem, you’ll have different solutions. Some solutions include outsourcing. It doesn’t even have to be software. It can be that’s for whatever. I think that’s the first thing. Are you trying to solve the problem of you have a huge turnover in your support team and you have a big problem ramping people? Are you trying to solve the problem of compliance?

You see that some of your support agents, like I recently read this blog post. It’s based on a book about decision making, and it says that problem with decisions are either that they’re biased, so the same agent will always be making the same mistake, there’s a bias towards making that mistake, or that there’s noise, and that depending on which agent you speak with, and what time of day, and if it’s before or after lunch or whatever, you’re going to get a different type of response. It’s that as opposed to bias. What is the problem that you’re trying to solve with your team? Based on that, you can find the right solution. Let’s start with the problem.

Justin: I like that. It reminds me of something that has come up a couple of times during the show that it bears repeating. Now you have to identify, really be comfortable in understanding what problem it is that you need to solve, but you also really need to understand the process you have in the first place. One of the things that we try to do, and I’m sure Zingtree is in the business of [unintelligible 00:20:11] very curious to hear how you guys approach this, too, is map out the process in the first place. You may find that there’s just good old-fashioned job design or org chart structure or just good old-fashioned process management that you don’t necessarily even need technology to solve. You just need to go back in time and just rebuild the thing in the first place. You might find a lot of benefit to that.

David: Yes. You just need a good [unintelligible 00:20:37] center, you actually need software.

Justin: Exactly.

David: The story I really like, a great problem that one of our customers at Zingtree solved with our tool is SpaceX, where their problem was, they have all these brilliant literally rocket scientists designing rockets, and they have hundreds of pages of code of how you’re supposed to build spaceships in the space written by NASA or something like that that that they have to follow, and they can’t expect their engineer to know the code by heart or to find a right place to search. Maybe they would make errors under searching, et cetera.

What they wanted to build was the Turbotax to the tax code of this code. That’s where they used a decision tree software that thankfully they chose Zingtree. A little bit of self-promotion, but their engineers now when they are looking for what the regulations for something, they can follow an automated decision tree that the compliance people built. They’ve saved thousands of hours just from that.

That’s a very clear problem of it takes a ton of time to read all these complicated regulations. You can imagine the same thing, that Typeform and Notion and Zingtree, the first approach to the product is very simple, but as you get deeper into it, there’s certain areas that go really deep. You can’t expect every support agent to know how to solve all of those problems. What you can do is give them a decision tree that helps them get to it, again as an example of shameless self-promotion. I’ll stop there.

Justin: No, you’re more than welcome to self-promote. David, this has been a fantastic conversation. I think some of the stuff we’ve talked about today, I’m certain it’s going to spark the imaginations and minds of people listening here. If someone does a cohort analysis on their free users inside their application and figures out a way to optimize those into a higher-value user base, then we’ve done our work for the day. Excellent conversation. I really appreciate you coming on. We’re going to end with our quickfire round Famous Five Fab- for I haven’t branded this yet. We’ll get there.

David: [chuckles] You should.

Justin: Let’s send out some quick ones here. What’s the book you most often recommend to people?

David: It’s a book called The High Output Management by Andy Grove. I literally just recommended it yesterday.

Justin: Yes, I read it. It’s a fantastic one.

David: Yes. As I became a manager and then a manager of managers, it’s really what helped me navigate that.

Justin: That is an all-time classic. What is the best productivity hack or productivity practice that you’ve implemented for yourself?

David: I use Notion for everything. I’m a little bit biased, I worked there. [laughs] I agree. In there, I have a way to organize my days that I really like. The way I organize my to-do list, I have my ad hoc list of things that just come up periodically and I have a place to put them so I don’t have to organize them as they come up, I can just put them there.

For the week, I have a Monday, Tuesday, Wednesday, Thursday, Friday, and I have these columns and I have one for next week where I put my tasks for next week. Then at the beginning of each week, I drag the task into each day. Then lower down in the page, I have tasks that I want to do some time and then I have other.

The point is that I really like organizing my week that way because I make sure that I don’t drag more than two or three items for each day. It allows me to plan my week and to block off time on my calendar if I need to to-do things. Since I have that system, I just feel a little bit more productive than I was before.

Justin: It’s very interesting. It reminds me somewhat of the Eisenhower Matrix. I don’t know if you’ve ever looked at that, where you have stuff you have to do, stuff you can delegate, stuff you can put off, and stuff you can just–

David: Exactly. Important versus urgent.

Justin: Yes, exactly.

David: Before that, I always had to-do lists, but having it those days, what I used to do is just drag two or three things every day, but then sometimes they weren’t planned far enough ahead and I would feel like the week ahead isn’t good. Anyway-

Justin: No, I love it. That’s a really good one.

David: Thanks.

Justin: That might be another blog post for Startup Grind.

David: I do advise the two companies and I’ve given that presentation a few times of my productivity model.

Justin: Very cool. If you could recommend one site, blog/community, LinkedIn group, et cetera, for someone in the support space, what would it be?

David: The companies that seem to create great blog content, I don’t actually follow any blog, but I gravitate often to the same places. I think Gainsight is great-

Justin: Yes, Gainsight’s got a great one.

David: -thought leader in this space.

Justin: They do a great job.

David: I really love their content. I think Zapier is great and Intercom would be my top three for learning how startups operate. The last two are learning how startups operate. The first one is more thought leadership and customer success.

Justin: Yes. Totally agree. If there’s one person in the world of support or even revenue teams, let’s say in the revenue team organization, there’s one person in the revenue org world that you could take out for coffee or cocktails depending on the time of day and vibe, who would it be?

David: I guess if I could only choose one, the leader I admire the most is Nick Mehta, the CEO of Gainsight. I admire him both because I happen to know a lot of people at Gainsight and they all speak very highly of him, extremely consistently. What I love about his leadership style, which he shows even when he’s speaking publicly at his events, the pulse events, is he’s very vulnerable. He’s not what I– and I think a lot of people dislike about Silicon Valley arrogance. He’s the opposite of that. I hope that I follow in his footsteps and I hope a lot of other leaders become more like him.

Justin: That’s great.

David: I had the pleasure of meeting him, but I’d take him out again. [laughs]

Justin: That’s really important. It’s a little anecdote, we had a bit of a fire drill at the website this week. One of the things that I kept telling myself the whole time was humility, the buck stops with me, this is my fault. The team is doing their best. You get really far in this world by not being a jerk. It’s great when people who lead businesses as successful in this space, something like Gainsight, still have that humility and goodness about them. That’s fantastic. David, thank you very much for joining us. Where can people find you if they want to learn more about you or Zingtree?

David: I’m on LinkedIn under my name, David Apple, and same on Twitter. I have to admit, I’m a little bit less on social media now that I’m a dad as of 14 months ago. I think I’m prioritizing the right things.

Justin: You are.

David: Please, follow or reach out to me in any of those channels. I love this type of conversation. Thanks again for having me.

Justin: It’s my pleasure. Thank you.

Justin: The Support Automation Show was brought to you by Capacity. Visit to find everything you need for automating support and business processes in one powerful platform. You can find this show by searching for Support Automation in your favorite podcast app. Please subscribe so you don’t miss any future episodes. On behalf of the team here at Capacity, thanks for listening.