140: Scary Support Stories with Todd Curtis

Charlotte Ward 0:13
Hello, and welcome to Episode 140 of the customer support leaders podcast. I’m Charlotte Ward. This week we’re feeling a little spooky. So stay tuned for five scary support stories. I’d like to welcome to the podcast today, Todd Curtis Todd, it’s lovely to have you join me on this chilly autumn evening around this glorious campfire as we gather together to tell some scary support stories. But first, this is your first time on the podcast Would you like to introduce yourself?

Todd Curtis 0:51
I’d be happy to thank Charlotte. It is a cold rainy day. So I’m happy for the campfire. My name is Todd Curtis, I’m the Chief Product officer at you need to budget. Or if you’re busy and short on time, you can just say hi NAB. And I used to be I used to run support for y NAB as well.And so my scary story that I think we’re going to talk about now comes from from five or so years ago. At that time. Yeah, now that still I would say that support is really still near and dear to my heart, I think of it as a part of our product. So

Charlotte Ward 1:35
it often is with subscription models, isn’t it? Yeah, I saw you tell this story a couple of years ago in Boston. And I was sat in the audience that day. And I have never seen so many support people raise their collective gasp of terror as you unfold this story, so I I’m, I was keen to have you back to tell it again. Because I pay. I mean, what what what is this week for if it’s not the putera and Chilling Tales? Yeah. Are you reliving it as we speak a little bit? A little bit a little bit? Would you like to kick off and tell us where it all begins? Sure.

Todd Curtis 2:27
where it begins Is that why NAB guest by the name you need a budget makes personal finance software, we teach people how to make a budget and take control of their money. And we’ve been around for a decade and a half or so. And up through a 2015 we had downloadable one time purchase software. And throughout 2014 and 15. We were working in the background on transitioning to web based software and transitioning to a SAS subscription business model. So that’s sort of the deep background for that, sir, as we did that, in support, support was always very important to us. Because not just technical support. But working with finances is really anxiety provoking. For a lot of people, there’s a good Halloween beam. And they need some hand holding a lot. So our support was always a place where people could ask a non technical support question about how to build a budget or the best way to go about managing their money. And so it was really, really important to us part of our support, also at the time has always been on teaching online workshops, for example. Anyway, we did that with a team of Gosh, when I first became responsible for support, I think there were two of us in 2014. Maybe. And by the time we got towards the end of 2015, we were probably in the middle of hiring our sixth and seventh Support Specialist. And we probably at that time we’re used to getting maybe 200 compensation for that. That’s kind of a deep background on where why NAB was before lineup support was before the story started. Our biggest issue was how to handle we had a sync algorithm between our our desktop software and our mobile clients that went through Dropbox and it was really tricky. And that was our biggest headache was how do we help people resync through Dropbox

Charlotte Ward 5:00
Yeah synching to Dropbox or as a headache 200 tickets, full time job for six ish support folk, right? And then you decided, Well, why not decided to move to a subscription model? That’s what I recorded the story. It changes,

Unknown Speaker 5:22
big changes and changes that we largely thought would reduce our support load. Because we were going to remove this Dropbox sync issue by becoming web based. And our long running issue we had was that there was people would talk about a steep learning curve for using line app. And we really felt like, you know, the new version wasn’t just taking the desktop version and putting it on web really felt like we had resolved a lot of the user experience issues that had made by NAB difficult to get started with. So we were pretty sure that our support load was going to go way down. With this launch.

Charlotte Ward 6:12
We reassuringly optimistic at this point, is quite optimistic. I do remember we had a retreat, probably in October before the launch. And one of our support specialists, Jen had said, What if it doesn’t go down? What are we going to do? And I remember saying something to the effect of we’ll get by like we’ll

I have I have visions of you on this retreat in October in another long distant since gone October. I imagine it was much like this, these comforting conversations around a campfire, your mushroom marshmallows the whole thing? Yeah, yeah. It’ll be fine. Yeah. Yeah. Nothing to worry, nothing to see here. And it’ll be totally fine. Excellent. It’s all good. What happened there? Yeah, what happened next. So

Unknown Speaker 7:05
we were in the hum stretches we went with, we did a kind of a soft launch paid beta, like pretty well. We had some things that weren’t quite done in the in the change to a new version. But it was also very important that we launch by New Year’s because New Year’s resolutions speaking of Halloween, I guess, New Year’s resolutions are a real thing, like ghosts. And people make them about their money. So January is very important to us. And very important to a lot of people out there as far as taking control of their money. So we really had to launch affordably. So we set December 30 2015 as our date. And I remember in the couple of days leading up to that making a really concerted effort to put the polish on all our snippets and make sure that q was sitting at zero so that when we kind of pulled that lever on noon on the 30th, we’d be starting just like we did, we were ready to go. Wow, wow.

Charlotte Ward 8:12
I sense that this story is about to take turn. It did.

Unknown Speaker 8:22
So our queue, as I said, but I’m going to say it again on noon, December 30 2015 to zero. And we were used to getting about 200 conversations,

Charlotte Ward 8:34
every support team stream every year 14 just waiting for someone to write in.

Yeah, yeah.

Todd Curtis 8:41
24 hours later, our queue was 5500. And they were still just six of us. Wow. So

Charlotte Ward 8:53
a little bit beyond capacity is a bit a bit. Yeah, we I don’t I don’t know how many of you must have gotten about 5500? Because I don’t know how many we could answer. That’s right. So that’s probably about how many? Well, first of all, wow.

So you’re sad desks, presumably looking at this queue? What’s going through your mind? I know what’s going through my mind. And what went through my mind back in 2018 when I heard this story, which was sheer terror.

Todd Curtis 9:31
Yeah, definitely some of that disbelief. Really. I mean, like, when I look back in retrospect, I’m almost surprised that I wasn’t more. I don’t know, panicked, or that I wasn’t like beating my head against the wall. And I think maybe some of that was disbelief like, okay, maybe I see that number right there. over on the side of helpscout, but maybe, I don’t know, tomorrow, it’ll be better almost, I think was an

Charlotte Ward 10:08
out of body experience of some. So

Todd Curtis 10:10
yes, so unreal, it can’t possibly continue almost right, like an apparition. I really think that that was that was part of my feeling at I think probably 330 in the morning, around the 30 December.

Charlotte Ward 10:33
Mm hmm. That was my initial sign. So deep breath. I can’t even begin to imagine a how you deal with that volume with six people. At that percentage over capacity? Yeah. What do you do?

Todd Curtis 10:55
And it is, um, my, my hopeful dream did not turn out to be true. It continued. I don’t I don’t know what our peak.

Unknown Speaker 11:07
I don’t know what our peak was. But I think it was probably higher. What do you do? I think for the first few hours, we just scrambled, you know, almost thinking like it’ll slow down.

Charlotte Ward 11:21
work fast enough to get on top of exactly,

Unknown Speaker 11:24
exactly no matter how few of you that are that good. Your first instinct, I think it became clear both quantitatively and qualitatively that that wasn’t going to work. Because what we uncovered, as we started, that process of just trying to power through was that there were all sorts of questions that we had never answered before. We didn’t have answers to this question, because it wasn’t just a spike in the queue. It was a brand new product with brand new features.

Todd Curtis 11:56
And one where we had not rushed, but where we had had to meet that deadline. So able to prepare as much as we would have could have for some of those questions. So we were discovering brand new types of questions in that it wasn’t just the volume.

Charlotte Ward 12:15
So you’re not, you’re not ready. You don’t have a stronger backbone of knowledge or macros, or I think snippets you said, yeah. So not none of that stuff that when you’re dealing at peak you often rely on when customers have the same question over and over again, that helps you get through normal day doesn’t right peak day.

Todd Curtis 12:35
Yeah. And I could think of an example where we still had this classic version. And so we also then had to apps companion app in the App Store. And

Unknown Speaker 12:52
we found that people were writing in basically with an issue of no longer being able to sync mobile, they were on the new web version, and they couldn’t think the mobile app, and it turned out that they had the wrong mobile app. But we couldn’t figure out we had a we had a screen on there that was supposed to take and we and so we’re live, you know, looking at this screen, what’s wrong with this? And we you know, figuring out that, oh, this button that we actually meant to de emphasise doesn’t even look like a button anymore. It looks like a an artistic reflection that we that we use. So like, you know, there we all are on just one continuous Google Hangout trying to figure this out live. Meanwhile, you figured it out. And then you know, 400 more people have written in the problem.

Charlotte Ward 13:42
Did you dream of the days of 400?

Unknown Speaker 13:45
Oh, that was i was i Charlotte? I remember distinctly thinking like, well, if tomorrow only a couple thousand people write it. That’ll be better, huh? When we were used to a couple of hundred. So yeah, I was I was dreaming of 1500, let alone to life. Yeah, yeah.

Charlotte Ward 14:06
So you’re not getting on top of this pile then. And it’s shifting sands, I guess you never know that. You’re just constantly climbing and, and not getting to the top. What do you do?

Todd Curtis 14:21
We made a couple of hard choices. We very much pride and continue today to pride ourselves on very personalised support. And we have snippets, but we really, we use a snippet as a core of an answer. If we want. If there’s a five step process for someone to do something, we don’t need to type those five steps over. And better not anyway because we don’t want to risk typing it wrong and finding the person. So anyway, so we do use automation a little bit in that way, but only in the in the context of a personalised response but Obviously, that many new conversations person. So we did a couple of things. One, we didn’t even like to have an auto reply at the time, because we were going to reply fast enough that you didn’t need to set up an auto reply. It gradually got a longer and longer time inserted. You’ll hear back in 24 hours, you’ll hear back from us in 48. You’ll hear back from us And today, we kept extending that time, each time was a gut punch. Hmm. But it was direct, honest. So had to be done. We collected what we thought of was a list of maybe the five most common issues we were experiencing. And we made what was essentially a five questions FAQ around, hey, I’m having trouble with this launch, is it and we ended up putting that in the auto reply to say, we’re underwater, your issue is probably one of these, please check this page. If it’s not right back, you know, reply to this same message so that we will know that they were common issues. So that took care of a fair number. Then we get some things this I will almost always like, not regret because it needed to do it. But I feel like it was my lowest support where I made. So there’s those there were those five issues. And then maybe there were the next five mm that needed something a little bit more dialled in a response, something more specific. And I made snippets for those five things. But then this is the dark mode. I went through the queue 50 conversations at a time. And I would send what I thought was the right plot reply based only on reading the subject. If the user had some, wow, I would look at those 50 I would read the subject lines, and I would check off on that page of 50. These nine seemed like response number one, I literally had like a sticky on my desk that said workflow number one. And so I would check off those nine or 11. And I would run a workflow that would send that reply, never having read. brutal, horrible,

Charlotte Ward 17:40
the exact opposite of what I think of as failing all the terror again.

Todd Curtis 17:47
I’m glad I can bring it back for you.

Charlotte Ward 17:48
It’s it’s evoking such awful dreams of just hearing the story the first time, let alone experiencing it. Yeah.

Todd Curtis 17:57
So we did that. So I think those were the Those were the things that we did, too. Well, to try to move through conversations faster. The other dark thing that we did was kind of a queue bankruptcy. No conversations have been sitting here so long. The person has probably moved off. And we’re going to basically just send an email to them to say we couldn’t get to your email.

Unknown Speaker 18:30
We, you know, hopefully, here’s again, here’s the FAQ page, we hope you solved it. And we just sent that email that generic email to them. close those out, so that we could respond better if they wrote back. Or if someone Yeah, again, I don’t it hurts to say it out loud. But I already gave a couple.

Charlotte Ward 18:56
Yeah.

Todd Curtis 18:57
The other thing we did was we started hiring.

Charlotte Ward 19:03
And at this point, so just we’re just at this point, what’s the timeline? Like when you decided just queue bankruptcy, the ones that you could not even send a random, like not random but a best guess response to for a couple of weeks. Right? Yeah. Yeah, yeah. A couple of weeks of hell.

Todd Curtis 19:25
Yeah, about that time where it Did you know, the volume did dimin. It did go down to 1000 a day. But that was still five times what we were used to. And we were still trying to figure out some of the problems we had probably problems with technical problems like with migration from our previous version to the US.

So we still had to do you know, it was still trying to figure out the best way to help. And it was probably a couple of weeks in which we decided like, we need a lot more people because we wanted to get back to doing it in a year. really helpful way that felt true to our core values and true to our product? value?

Charlotte Ward 20:07
Yeah. And you’re still in January. So there’s still somewhat of a hope of rescue of the situation of your absolute peak. Right. So how long? How long would it ordinarily take to hire and onboard?

Todd Curtis 20:21
Yeah, just the right question. Because up to this point, we ran, if we ran a really quick hiring process, it would be six weeks, it was often more like 10. We have always been a remote company, not just 2020. So we could never do that thing where you just set aside everything for three days and get people in a conference room and have candidates come in and go from group to group. It was always one stage one week, one stage the next week. And because we were remote, we also ran a very intensive process to make sure that we could really get to know this person remotely well enough to offer them, offer them a job. But that obviously wasn’t going to. So we set a goal that we would have, I think it was probably a Tuesday or a Wednesday, I remember sitting around talking about this, that we wanted to do job offers the following Friday. So yeah, we have we have 10 days to do what normally takes us up to 10 weeks. We happily, thankfully, had had an open posting on the web site, our marketing site. I don’t know if it was actually open at that time. But we had just finished hiring two people. And we felt like wait a minute, we have another, you know, x hundred people who had written in about that one.

Charlotte Ward 21:46
Yeah, you got to pull the start there. Yeah.

Todd Curtis 21:50
So we I think three of us went in and just pulled people who looked promising based on the letter they had written and we started them. We sent them a questionnaire, a long form questionnaire, which is usually probably the third or fourth step in our process. We did that first. And then after that questionnaire, we will usually have two or three calls. We had one call after that. And then we asked them if they would like to start basically on a trial basis for eight weeks. And that we can hopefully use that time to

obviously onboard but, you know, continue to get some dough with them. You know, that would normally we would still be in our hiring process. Yeah, yeah. Yeah. And, and yes, a bit the following Friday, we had 60. Wow.

Charlotte Ward 22:49
That’s a huge that’s a huge load for an already stretched to stressed overburdened team. I like onboarding into a team that six or eight or 10. One person is really felt if you do double the team at speed at stress, no more nightmares. Right? What did it go? Well, though, please tell me it went well,

Unknown Speaker 23:09
did I mean, again, one of those like, in retrospect, I mean, I don’t like I’ll come back to the more immediate matter of your question. But those people who were in that six group, I think it four of them occupy very significant positions at wine, I have no inside and outside support. It was a goldmine there. I mean, we, we did so well. So but yeah, more to the at that moment, onboarding a team to double the team. I think we did a good job being genuine, is one of our core values and saying, look, we, you know, we just need you to sink or swim and Gosh, hopefully swim. Yeah. And I suppose in one ironic way, there were enough conversations that had enough in common that you can train someone to answer one literally one type one question. And they could go in and get a whole lot of work done to answering that question. Or make sense. In a more personalised way. Yeah. Then we had been able to do before Yeah, so we and then we did we just did it again. We hired those six people, and we immediately did it.

Charlotte Ward 24:37
Wow. So triple the team to cope. I don’t know if we got six the second time we did it again as a goal, but I thought maybe we only got four the next.

That’s still some amazing growth at speed,stress. Wow. I don’t know what the final number was, but I feel like we might have we might have gotten gone from six to Low 26 bucks. Wow.

Yeah, yeah. Which is a success and success story wonderfulness. Yeah, yeah, yeah. And you live to tell the tale.

Todd Curtis 25:13
Yeah, we did just just barely. I’m Carrie, who was our director at the time and I took our first days off in February, including Saturdays. My favourite story from the story is when I was talking to my wife and she said, you know, you really should take a day off. And I said I just did yesterday, so Well, that was a Sunday. And I think that was Sunday, February 15, or whatever. That Wow, that was the first time we kind of put the keyboard down, huh? Wow, yeah, it’s very scary. And it’s, it’s a fun is the wrong word. It’s fun to think about in the sense of like, an experience that made you who you are a bit and it definitely made me it really it only reinforced what I believe about good support. Having had to do something that I thought wasn’t.

Charlotte Ward 26:12
Mm hmm. Yeah. live to tell the tale.

Todd Curtis 26:16
Yeah. To be better.

Charlotte Ward 26:18
Yeah. Yeah. Thank you so much for sharing that that was every bit as horrific as I recall it being and. And Dave, perhaps reinforced for the you know, the Dark Knight the flickering flames and the personal story. Thank you so much, Todd. Should we throw another log on the fire?

Todd Curtis 26:39
Absolutely. Let’s have some marshmallows.

Charlotte Ward 26:41
Thank you so much.

That’s it for today. Go to customer support leaders.com forward slash 140 for the show notes, and I’ll see you next time.