In Episode 592, Rob Walling is joined again by Cody Duval for a technical conversation about the dos and don’ts for amazing customer support.
The topics we cover
[2:00] Customer success vs customer support
[5:10] Response time
[8:59] Post-support surveys
[10:58] When to hire first customer support person
[13:02] Chat widgets
[17:09] Doing customer support early on as a founder
[18:01] Training customer support to ask a question
[19:00] Dealing with abusive customers
[21:10] Customer support tool
Links from the show
If you have questions about starting or scaling a software business that you’d like for us to cover, please submit your question for an upcoming episode. We’d love to hear from you!
Rob: Welcome back. It’s Startups for the Rest of Us. This is Episode 592 where we’re going to dive into nine tactics for amazing customer support. If you recall, this is an experimental format where Cody Duval, the founder/acquirer of Keeping, which is a help desk tool, came on the show last week. We dug into his story of coming to MicroConf in 2019 and later acquiring Keeping through FE International. He’s been running that ever since.
I’m bringing him on the show today purely for the tactics, the tips, the tricks, the do’s and don’ts of amazing customer support. Without further ado, let’s dive into my conversation with Cody Duval.
Cody, I think you are the first guest on Startups for the Rest of Us to have ever made his first and second appearance seven days apart.
Cody: All right, excited.
Cody: Thank you, Rob.
Rob: This is great. As people just heard in the intro, this is an experimental format. I realized that a lot of the founders that I interview have this subject matter expertise that you gain through the osmosis of running or building an app in a space.
I remember when I was running HitTail, which was an SEO keyword tool, early on, I acquired it, revamped it, and started marketing it. People started asking me, well, how do I do SEO? And what should I do with this? I was just like, it’s a tool, go use it. I don’t know, whatever. I know how to do SEO, but I’m not your coach.
I started realizing, oh, so much of the value is the knowledge that’s in our heads of whether it’s SEO, running an ESP and knowing best practices for email, running a support tool like Keeping and having a lot more knowledge and a lot more exposure to a lot of, I’d say, probably successful customer support experiences. You’ve probably seen some that aren’t great. That’s why we’re here.
Cody: Yup, that’s exactly right. I’m an operator of a tool that empowers teams to provide customer support to their customers, but I’m by no means a professional. I’d love to, as we have this conversation, have folks disagree, I’m @codee on Twitter to shout back because all the things we’re going to talk about here are things that I’ve just absorbed from talking to my own customers, having conversations with professionals. I think this will be fun.
Rob: I want to kick it off before we dive into these. It’s called nine tips for customer support or something like that. That’ll be the title of the episode. Maybe we wind up with 8 or 10. We kind of wander, but the interesting thing we were just talking about is the term customer success versus customer support.
I want to walk back and talk about my understanding of these definitions. Customer success was not even a term seven years ago. I think the first time I ever heard it was probably 2013, 2014. Customer support has been around forever because we know that you would call United. You called Pan American Airways and you’ll get customer support.
Startups have traditionally done it via email, sometimes chat widget. Customer success became this way a lot around onboarding and a lot around helping high value clients get value out of a product and retain those folks to cut churn. That became this new role with the subscription businesses.
You didn’t really need customer success if you made a one-time sale. They had implementation engineers and all that, but there was a relationship manager and AE or whatever. But really, SaaS, I think, brought about this need for customer success because we know that churn is the death knell. It’s the kryptonite of SaaS and customer success.
I think their overarching umbrella is to help people get onboarded, not churn, and stick around, but you and I were talking offline just before this that this definition may be changing or adjusting. Do you want to talk through how you see success versus support?
Cody: Yeah. I do think that this is a term or a job title that, like you said, has been birthed in the past couple of years. The way I see it, at least with my own customers, is that customer success envelops the full lifecycle from when someone comes on board as a customer to when they might churn on the other end. We might talk about customer support once they’re already a happy customer and they’re needing advice and/or a problem solved. I think in the industry, customer success now covers the whole lifecycle.
Even up another level, you have the customer experience or CX. There are conferences where these folks go. I think if you’re working in the industry, you probably call yourself a customer success professional, but you may be doing customer support as well. We’re going to talk about customer support stuff here, but this maybe is under the umbrella of customer success.
Rob: Yeah, and that’s the difference. When I left Drip, there were about 110, 120 employees and there was a 6-person customer success team. They very much were high value clients. We get them on board and keep them around, thinking about making onboarding better—all the stuff we could think of under customer success. Then there was a 12- or 14-person customer support team and they were very separate.
Someone talked about merging those two and I was like, there’s an overarching theme, but the skill sets of those two roles are actually quite different. If they move customer success up to this umbrella term over both of those, I think we need a new term for what those customer success people were doing at Drip. Whether it’s an account manager or relationship manager, there should be something that has to replace that so it’s not ambiguous. Does that make sense?
Cody: Yeah, I think those are AEs or account executives. They’re managing the customer at a strategic level, not just dealing with tickets.
Rob: Right, not reactive but proactive. All right, let’s dive into this first tip. If you pay attention to only one metric, make it response time.
Cody: Right. If you adopt a tool, there’s going to be a ton of metrics that you can look at, there are dashboards that are all dedicated to customer support. What we hear from our customers and what I’ve noticed is that the thing that matters the most to their own customers is response time, in particular, first response time. That specifically is how long does it take to get a reply from a real human, not an auto reply. The auto reply does not count.
Just to say, hey, we acknowledge that you have an issue or you have a question and that we’re on it or even you provide a solution. There are, I think, various metrics as to what is a good response time or a good first response time. We can talk about that too.
Actually, just before we started recording, I recently emailed a very well-known SaaS tool and I got an auto reply that said, our typical response time is three days. I just immediately was like, oh, are you kidding me? I get it, but a customer doesn’t care or know that you have thousands of folks emailing you. You are a special, and at Snowflake you are one and you want to reply. Generally, we can talk about what’s a good response time, but I think that this is the most important metric to pay attention to.
Rob: Yeah, what is a good response time?
Cody: I think that for teams that we’re talking about here, generally, around 12 hours or less is what you should aim for, which really means like a business day. I think that there are definitely tools and areas where maybe more than that would be acceptable. What I see from most of my own customers is that going anything longer than a day, customers start to feel ignored and they feel sad. That should be a nice rule of thumb to pay attention to is to try to get back to everybody within one day.
Rob: The second topic is this difference of interacting with a human, like, hey, this is just an email between two people. As cold as email might be, email always feels like I’m texting with a little bit of delay and I’m having a human interaction versus some systems like Zendesk is the one that comes to mind because, again, I hate to keep talking about merging with Leadpages and that whole thing. That’s why I have experience with a bunch of different systems because I’ve been on both sides of this.
We were on Help Scout. I liked it because it felt like a person-to-person interaction. It wasn’t a bunch of cruft around the email or it wasn’t this survey at the end. It wasn’t all this stuff. Then we moved to Zendesk because that’s what they were using and I hated it as an experience. Do you have thoughts knowing the space better than most about which is preferred by customers?
Cody: Yeah. This is near and dear to my heart because at Keeping, it’s one of the things that we like to say. What makes Keeping different is that you really should feel, as a customer, that you’re dealing with a human. I have to give Help Scout some props here because if you look in their marketing docs or even in their own support docs, they don’t like to use the word ticket. They call it a conversation.
This goes to what you’re saying, which is, when you start labeling, hey, here’s your ticket, it becomes anonymous. You’re this work product being pushed around a system. If you’re at a high value SaaS, again, you want to be talking to a real person that knows what they’re doing.
I really encourage folks to look at tools that don’t say, here’s your ticket number, click on this link to check your status, please reply to this, go to this portal. I think that today, in 2022, even though it’s an email, people want to know there’s another human on the other end of that email and not just a robot.
Rob: How do you feel about these surveys that go out after where it’s like, rate us one to five? How did Josh help you out? I have my own thoughts on it and I get it. I get that when I have a team of 5 or 500 customer support people, the way I’m going to evaluate them is based on response time and customer satisfaction.
If I’m not sending those surveys, I don’t have that second piece of data. I get it from a business perspective. From a user perspective, it pisses me off. I hate it, and I’m curious. You’re on both sides of it, what are your thoughts?
Cody: You can start to see customer service tools are really segmented. We’re going to pick on Zendesk here. They are geared towards teams with hundreds, maybe thousands of folks on the other end who are replying to customer support requests. There’s a manager and a manager’s manager. They need data to know, oh, wait a minute, are we doing a good job? Unfortunately, you, as the customer, are being bullied into providing that data.
I think more often than not, the data isn’t very helpful if you’re at a small scale like most of us are. Number two, I think customers kind of resent it. I think a lot of times, they’re only going to click the frowny face and probably ignore the medium and the happy face. You’re going to get skewed.
Oftentimes, data folks were like, well, what does that frowny face mean? I took too long to reply? Was the person rude? It’s often not very actionable. Again, if you’re a 5000-person organization, maybe you can get some value out of it. But I think for most of the folks that are listening to this podcast, it’s not very useful.
Rob: Yeah. And I feel like the only times that I respond to it are when someone either is amazing and I want to give them a thumbs up or they’re terrible and I want to give them a thumbs down like they were overtly rude. Other than that, to me, it’s more work for me. My inbox is already clogged with stuff. I don’t want another email that I have to think about and sit there and may not even remember the experience.
A question I know that I get a lot when I’m mentoring and advising founders is, when do I hire my first employee? When do I hire my first developer? When do I hire my first customer support person? What are your thoughts on that?
Cody: Right. This is hard because I think that depending on the type of service that you’re offering really is going to […] the kind of person that you can hire. My own rule of thumb, just very generally without going too far into that, is that if you can’t get back to your customers within a day, so this goes back to that response time we’re talking about, that’s a signal that, wait a minute, maybe you need to hire somebody to be that first line to respond to your customers.
The big assumption here though is that those tickets are fixable by someone that you hire. If you’re early and you have a lot of technical bugs and everything requires a ticket in JIRA or your project management tool, that customer support hire is going to be not very helpful because they’re just going to say, well, yup, it’s broken. That’s it. But if you get a lot of requests that can be tackled resetting passwords or billing questions that can be hired by somebody who’s relatively trainable, I think a lot of founders find this as a huge weight off their shoulders that they’re not staring at that queue and they can focus on their business.
I also like to say that if you’re a technical SaaS that this is a great place to start a junior engineer, somebody who’s learning if your Rails and this is their first job. Opening bug tickets and tracking down bugs is a great place to start. They can bridge into a more senior role from there.
It really depends on your domain and the kind of tickets that you’re getting. But I think for the most part, if you’re feeling overwhelmed and you’re not able to get to your customers back in the day, then think about hiring somebody.
Rob: I think a lot of founders wait too long to hire a customer support rep. Oftentimes, it’s the same reason/excuse where it’s like, well, but I just know so much and as the founder, I can just do better support or our product is too technical, I hear this all the time. You don’t hire someone with no technical experience then. You hire someone who can either figure it out, or as you said, it’s a junior engineer.
I have known multiple founders who told me, oh, it just isn’t that much time. I’m not going to hire someone full time. I don’t know if I can hire someone part time. It’s only 30 to 60 minutes a day. A founder told me that and I was like, are you fucking kidding me that you’re spending up to five hours a week on this? This is the lowest value. Customer support is extremely valuable, but you can find someone for not very much money to do it well, not very much money compared to what your hours as a founder are worth.
Cody: It’s the thing that keeps people up at night. One thing as a solo founder, especially, you’ve got all these things circling you taking a little bit of your founder mana away. I think that not having to worry about that customer support queue, you’re just going to sleep better at night. I agree.
Rob: Yes. And the interruption, you don’t take that into account. The distraction and the interruption are the two biggest things, not the five minutes to respond to it. It’s that it’s off your plate.
Next thing is a chat widget. To chat widget or not is a pretty common question I get. I advised a company that had the chat widget for everybody including their free users and really inexpensive plans. They have plans running from $9 up to thousands a month and they have thousands of people coming in. They were trying to man the chat window or the chat widget and it was a huge amount of effort.
I talked them through, hey, maybe we should not have it available for free plans or maybe we should not have them available for $9 and it’s only 49%. We just talked through maybe only during a trial, blah, blah, blah. There are just ways to hide it. But I also know that chat can be great because it’s real time and people can get fast responses. What’s your thought on that?
Cody: Yeah, it’s just not an easy answer. But I will say that when Intercom came onto the scene, they really I think were the ones that made the chat widget ubiquitous. My chat widget is amazing, especially if it’s manned and there’s somebody there. The one thing I will say is that once you offer a chat widget and you’re responsive on the chat widget, your customers will always go to the chat widget. Not everything needs an instant reply, but it’s very, very satisfying as a customer to not have to wait.
In talking to a lot of folks about what you just said, at some point, your business is going to be just too big to support a chat widget. I think if you’re less than 200 customers, that’s just a ballpark number, consider a chat widget. I think that it’s something that can give you really good feedback. It’s really, really useful when customers get stuck in their onboarding process and they just need a little bit of this or a little bit of that.
I think it’s worth it in those cases. But after 200 customers, I think it may be a feature that’s only offered on the higher plans. Premium support actually means something a lot of folks have on their higher plans and it doesn’t mean anything, but I think chat support is something that people will pay for.
The last thing is, you have to make sure that there’s someone there. I think that an empty chat widget is often worse than no chat widget. A lot of the chat like Intercom comes with bots that fake type, which is a little bit of a pet peeve of mine. We see a bot typing when we know a bot doesn’t type.
Those can be useful. I think chat widgets can get people to their answer like, hey, here’s a support article that you asked for. But if folks really want to talk to a human and they can’t get to a human in a chat widget, I think you’re, at this point, just have them email and then you’re going to meet their expectations. Chat is a tricky one. I think that just be careful because once it’s there, it becomes something that’s really hard to take away.
Rob: Cody, you have a note here in the outline. This is tip number five, by the way. It’s about how early in the pre-product market fit that customer success, sales, and product are almost the same role. I want you to dig into that in just a second.
I remember, I talked about hiring customer support out and how I always did with all my apps, but I didn’t do it at the start. When I acquired HitTail, I personally responded to every email ticket that came in for three months, it might have been a little more.
Drip was the same way. Three to six months, I did it until it started (a) I started hearing the same things over and over, (b) our product roadmap was set, and (c) it became enough of an interruption and a distraction and it made sense. But I found those early days extremely valuable for me as a founder to be doing it. Why don’t you tell us what you’re thinking about this one?
Cody: Early on, this really is about when you’re still trying to find the right product for the right customer. Your support channel is really just another way to talk to your customers. If your customers are taking the time to write you an email or get on your chat widget, they’ve just opened the door to being able to have a conversation about, hey, something’s not quite right, this product doesn’t quite meet my needs. But they clearly like you enough to take time out of their day to write you an email.
I think as far as product development goes, it doesn’t have to be like an outbound email. Hey, will you talk to me about this product that I’m offering? Someone’s already raised their hand and said they’re ready to have a conversation with you, even though it’s in a customer support ticket, and that can be turned into sales.
As soon as somebody has said, hey, this is great but.. Then you can say, oh, well, let’s do this instead of that. And the next thing you know, you’ve turned somebody that was maybe on the fence. This is during a trial into a customer.
I think in the beginning, it’s really, really critical, even though no one probably loves doing customer support, that as a founder that is your main channel to getting a dialogue with your customers. Getting feedback about the product, you’re seeing how they type about it, how they talk about it, you’re getting all that language, and that’s going to come out in all sorts of ways as you build your product.
Rob: Tip number six is training your customer success team or yourself to ask every customer a question. What’s that question?
Cody: This is just a very quick tip. Attribution is just a mess right now in marketing. If you have a lot of inbound sales, where do they come from? You have this opportunity to just say, hey, how did you find us? And they will almost always say, Google, but they may also say, oh, I saw you on a buddy at a bar and said, hey, you should check out this tool.
It’s just an opportunity to get a little bit of insight into your marketing, especially if you don’t know where they came from. Obviously, if you reached out, if this was an outbound lead that you pulled in, you don’t need to ask them. But if you have a customer that came to you and you don’t know where they came from, just ask them.
Rob: Another question that I get pretty frequently is, how should I deal with abusive customers that are pretty far out of line? What’s your tip for that?
Cody: This is really hard. I think there’s a line between rude and abusive. I think a lot of folks, especially in an impersonal medium like email, can come off a little rude. I think that’s okay. I think we need to be able to deal with rude customers. That just comes with the territory. They may not be nice, but they are your customer.
I think when something drips into abuse, and I think it’s important as your company that you have already decided what is abusive with your team. Especially if you have employees and you need to train your team to very, very quickly just push that ticket if you’re dealing with a support ticket system to you, the founder, and let you deal with it so that your employee isn’t in the firing line of somebody who’s really stepped over the line.
After that, it’s up to you as a founder how to deal with it. You can fire them as a customer. I certainly have done that. I think I’ve talked to other founders who said, you know what, we just don’t need or want your business.
Obviously, that comes with some risk, especially in the world of social media, product review sites, and all the things that come with it. A customer support tool or a customer support platform gives you, as a founder, visibility into what’s going on. I think it’s something that just happens and it comes with the territory.
Rob: I had to fire a few customers for sure over stuff like this. My phrase was always, it sounds like we aren’t a fit for your needs. I would apologize. I’m sorry that we’ve disappointed you. It sounds like we’re not a fit for your needs. We need to part ways.
Sometimes the person then just went ballistic. Other times, they apologize and start begging for me to let them stay. This wasn’t one email that was rude, like you said. This was way over the top stuff—threats and this kind of stuff. It’s like, no, you’ve shown me your true colors. There’s just no way that, unfortunately, I can let you stick around.
Cody: Yeah. Again, this is after you reach a certain point. It’s worth spending some time. We actually have a blog post on our site about typing up a policy about what we consider abusive.
You can just outline things ahead of time that you can share with your team, anyone that’s on the front line, that they know they tripped over into a new character. Then it goes up to you as the founder or if you’re big enough into their boss and then you can handle it, so someone’s not stuck there in that really bad space.
Rob: Got it. We’ll link that blog post up in the show notes. It’s on the keeping.com blog.
Cody: That’s right, yup.
Rob: All right, we have two more here. This one seems obvious to me but maybe it won’t be to some people. It’s about deciding when to pay for a customer support tool. I know some people in the early days tried to do it directly in Gmail with labels. I think that not having tooling around that, whether you do go for a full-blown Help Scout or whatever or using something like Keeping that makes that possible within your Gmail. What’s your thought process around that?
Cody: I think some folks don’t know or, hey, what do I get with the customer service tool? What does it bring? I actually think that if you’re really small, if you’re still in the product development stage, these tools are expensive, so it’s okay to wait. You can get really far with just an email email@example.com. Especially if you’re a solo founder, you’re not assigning tickets to anybody, you’re just dealing with them.
On the high end, some of these full-service customer support platforms are $30 or $40 a month per user. That’s not a lot if you’re a big, well-funded SaaS, but if you’re bootstrapping, that’s a pretty big expense.
The killer features that may make you want to pay for a tool are (some call these) canned answers, shared templates, response templates. You’re going to find that you get the same questions over and over again. Not having to type the same email over and over again is a huge time saver if you can take the time. Basically, cut and paste that question and answer into a system that lets you one-click, insert. You can even customize it, of course, to make it feel like it’s being typed. That’s just a huge time saver.
All these tools give you basic analytics like we talked about with response time. The more expensive tools go really crazy here. I don’t think you need every metric under the sun. But being able to calculate something like response time with a little dashboard, I think, is really important after a certain size.
I think most of these tools allow you to communicate on the side about a support ticket, call it a note instead of forwarding an email around, is a huge thing you don’t know you need until you need it. A customer emails in and says, I have this question. So having able to have a chat about that ticket inside of your tool is really great.
One thing that is different for each domain is, we’re living in this omnichannel world and some folks get customer service requests only through email, then you should choose a tool that’s optimized for that. If you’re in the more prosumer consumer side of the equation, you’re going to get a lot of your customer support requests on Twitter and Facebook. And so you need a tool that can plug into those platforms.
In fact, I think for a lot of consumers today, that’s their primary channel. They’ve realized that the visibility of complaining on Twitter or Facebook gets a faster response. You got to have a tool that plugs in there.
Rob: Aside from spammers and IP blacklists, that right there, going to social media because my thing is so urgent and you should just pay attention to it. Those were the top three things that I hated about running Drip.
Actually, I […] despise them like it ground on me. Because then it was always like, cool, well… oh, I already have. And it’s like, oh, yeah, I see you’re third in the queue, we’ll get back to you. Our response times were very fast. It was especially in the days when I owned it. It was a few hours tops for an initial response.
It just always felt so entitled. It was never from the awesome customers who were singing our praises and who legitimately had an urgent issue. It was always the demandy customers paying us the least and who complained the most. It just inevitably was that.
Rob: All right, last one, our ninth tip. This is around tooling and software. You have some specific thoughts for folks about deciding maybe which type of tool to look at.
Cody: Yeah. What toy to pick? This is a hugely crowded space. If you were to go into Google right now and search for a help desk or customer service tool, you would get page upon page upon page of options.
Obviously, I’m going to plug Keeping if you’re primarily dealing with customer support through email and you need a tool that does everything we just talked about. Perhaps, you don’t have a 10- or 15-person customer service team, Keeping may be a great fit for you.
My own advice is that buy the tool that’s the right stage for your company. Help Scout is a great tool, but maybe not if you have 10 customers. The segments are now big enough that really, these tools are oriented towards specific use cases.
There’s a tool called Gorgeous, which is all about ecommerce. I think it’s worth spending a little bit of time and investigating which tool is really oriented towards your industry. My churn is very low. I’d like to think that’s because we have a great product, but it’s also because there’s a real investment that you make with these tools by pouring in your knowledge base, all your customer support data gets in there.
Once you choose a tool, you’re going to be on it probably for a while. Just take some time to choose the tool that’s right for you. Don’t just grab the first one. Know that most of these bigger customer support tools are built for customer support teams. If that’s where you’re going, then choose those tools.
If you’re a founder with a couple of folks, just be aware that you’re not falling into a really complicated and expensive tool that is way more than you need. I think those are just some things to think about before you put your money up on the table.
Rob: In terms of the knowledge base thing, I was always resistant to putting my knowledge base into a support tool because then it tied me into it. We either looked at third-party knowledge base tools or, at one point, we used WordPress with a theme. There are just other options, but it does seem like that’s the way people are going now. Is that accurate, where the KB is combined with the tool?
Cody: Yeah, the public knowledge base I think is a business idea for anyone out there. I think that it’s a really underserved market. I think if you want to build a knowledge base right now, all of these tools come with a not very good knowledge base, I have to say.
It’s good enough. It’s nice. It’s integrated with your tools so you can easily link to it from a support ticket. But from a customer side, they all just don’t feel very good. This is true with Intercoms, it’s true with Help Scout, it’s true with Zendesk. I think they all just feel like they’re not very well—there’s not a lot of folks paying attention to those products inside those companies. They feel like kind of an add-on.
I really encourage folks to unbundle if possible. I just don’t think there is a lot of great knowledge base that’s specifically oriented towards SaaS tools right now. A lot of folks just do, like you said, something in WordPress, a Gatsby static site, but it doesn’t feel like there are a lot of great options for knowledge bases right now.
Rob: I think the bummer is, given the fact that there are mediocre free versions—free in essence—attached to these customer support tools, it would make me resistant personally to going into the knowledge base phase because I know that a bunch of the sales objections are going to be, well, I already have a half-assed one in Zendesk, so why would I pay what I want to charge for it? Hopefully, someone out there is gaining traction.
Frankly, if you are getting traction, I’d love to have you on the show and talk through similarly. Do the interview about your product and then do an episode like this where it’s like things to do in your knowledge base, nine tips for founders.
Cody, thanks so much for coming back on the show. You are @codee on Twitter and, of course, keeping.com.
Cody: Thank you, Rob. Great talking.
Rob: Let me know what you think of this new experimental format where I have a founder come on and tell their story, and then the next week come back with tactics, strategies, or thoughts on how to do things better, something they’re a subject matter expert on. You can hit me up questions at startupsfortherestofus.com.
You can click the Ask a Question link at the top of Startups for the Rest of Us or you can head to Twitter. I’m @robwalling and we’re @startupspod. Let me know if you liked it. If you didn’t like it, you want more, or what have you. Thanks so much for listening this week and every week. I’ll be back in your ears again next Tuesday morning.