Start With Why

Andrew Verboncouer:  You're listening to the Seaworthy Podcast, Start With Why. Seaworthy is a podcast about building successful software. Today we're talking about focusing on problems - how understanding the problem reveals the right solution.

Chris Schmitz:  I'm Chris Schmitz. I'm here with Andrew Verboncouer, welcome to Seaworthy. Today we're going to talk about problems. And why need to focus on the problem before the solution. So I think as entrepreneurs, it's very easy to jump straight to the solution, get excited about the things that you can build. But until you fully understand the problem, it's impossible to say whether any of that actually should even be built. So we have a couple exercises that you can go through with your teams to help understand problems a little better. And we want to share a couple examples here about some of the solutions that were presented to us, I guess, and how we were able to dig through to fully understand the problem, or at least understand the problem a little better. And that kind of made the, the solution that was initially presented just irrelevant once we got a few layers deeper and fully understood the problem. So did you actually dive into the first one?

Andrew Verboncouer:  Yeah, so the first example is a design sprint that we ran with a cable provider and their internal support and call center team, you know, they were asking us for some requests, and had a bunch of feedback of things they needed in the software. And, you know, at that point, we didn't really know a whole lot about the intricacies of what they did. So it was important that we kind of, you know, made sure we trusted the process and, you know, took a step back and didn't, didn't really take take their feedback at face value. So,

Chris Schmitz:  so I had a good chat here, but just to give a little more context, this is a cable provider, like, similar to like a Time Warner or something like that. And they had an in house like, customer support department, right? And this is software that the customer support department used,

Andrew Verboncouer:  right, so they used it to troubleshoot things when people call in trying to streamline that and increase, you know, the user experience, there is a call into decrease call times and increase your efficiencies.

Chris Schmitz:  Right. Okay. So, yeah, what did they present to us? That's kind of the problem there actual solution for it? How are you able to dig a little deeper to fully understand,

Andrew Verboncouer:  yeah, so the initial problem they said, is that there's a MAC address with every cable modem and it's kind of a unique identifier, so that there's a number and we were presenting it in the software as just a string of 12 characters. And really, it's, you know, anywhere you see it, it's four characters separated by dots, you know, three times over, right, or separated twice, you know, that a couple different ways they formatted depending on the manufacturer, but it's 12 numbers. And so they said, Hey, we really needed in this specific format with, you know, four characters separated by a.or a period for more separated by period and then the final four and, you know, we could have really easily just done that in the software that's a really quick you know, it's a one line piece of code that that you could use to give them that solution but you know, something didn't right so we started going down the path of the five wise and you know, started asking questions like, you know, why, why do you need it, format it that way, you know, and they would come back to us and they would say, oh, because we need to use it in that format. We said well hey, that's great. But you know, why do you Why do you need it in that format and they said really because they they were copying and pasting it into the console of the CMT s so the software that we created wasn't doing

Chris Schmitz:  CMTS is just the higher level networking equipment so like your further up your network, right from your from your home router, cause goes up to something by the end of your neighborhood to a router and then eventually makes its way to this device called the CMT s.

Andrew Verboncouer:  Right, which is a cable modem termination service. But yeah, so basically manages all the cable modems have a specific area and gets feedback and data from it like SNR, you know, a few different data types that signal to noise. So how, what's the clarity of the signal you're getting? What's the frequency, it gets all that data back. And some of those are key indicators of performance when you think about, you know, what the end users getting at their house. So

Chris Schmitz:  it's been really valuable data for these customer service people,

Andrew Verboncouer:  right? So they know that they know the right data, they can, you know, ultimately go down a tree diagram of decisions of what could possibly be wrong. And how quick can we get this problem solved,

Chris Schmitz:  right? So why were they logging in to the CMT s in the first place? So the feed Simon, you mentioned, they need to get it as data. Yeah, but I mean, there's also the software that was built for them, yeah, to troubleshoot problems. So what what were they, you know, why did they need to kind of circumvent the system?

Andrew Verboncouer:  Yep. So the software we had was giving them data, but based on the previous experience, they had with some offers, they didn't trust it, you know, they didn't trust the data was real time. And in order to really troubleshoot and make recommendations to the user and, you know, create a ticket, what have you, they really needed that, you know, up to date data to make that decision and say, Hey, this is a problem that I can solve while you're on the line, or this is a problem that I'm going to have to, you know, schedule a work order for someone to come out to your house and, you know, do some troubleshooting on site. So it was really important to them that they had the right data. So they would copy it and they would paste it into the console of the CMT s, which would essentially pull the real time data that we had, you know, so that was really key. But they were, they were going around it, because they didn't trust the software, there wasn't enough messaging around when, you know, when the data was from, and

Chris Schmitz:  welfare, right, there's another couple other common tasks that they had to do, like, reset a moto. Yep, I forget exactly. But I remember there was, there's a few few kind of trivial features that we added in order to let them do these workflows that they ended up doing, you know, 90% of calls that they were having to circumvent the system, and almost every call, which is

Andrew Verboncouer:  right, crazy, right. So after we, you know, figured out that, hey, they were, there were, you know, going around the software we built to get this data. And, you know, we've did some observation. So it's kind of, you know, a testament to not just doing what a user says, but also being able to observe what they're doing. So we scheduled I think, maybe 10 or 15 observation sessions with different technicians and call reps to really record and watch their workflow. And so after they would paste it into the CMT s, a common thing like you said, they would do is reset it or do a power cycle of the modem. And so that was more queries that they had to write in, they had to copy the MAC address again, and run these different queries into the system, you know, which took a lot of time. So the average call time was about seven minutes. And this specific process to power cycle and reset a modem was about almost three minutes. And so we hypothesize that we would be able to implement a nice little widget right in the software that would tell them data, that's real time, it would allow them to refresh it. And it also allow them to reset and power cycle the modem right from there and get that real time feedback. So we were able to reduce almost three minutes down to 30 seconds, which is really valuable. If you think about, you know, a seven minute call, and on these types of calls that happen pretty frequently, you know, there's a lot of time saved.

Chris Schmitz:  Yeah, if you have 50 people Manning phones, and you can make them six times more effective,

Andrew Verboncouer:  right? Yeah, you know, and we took it a step further from the areas, we said, Hey, this is great. We know, we know what the time value of this is. But let's, let's try to get this to dollar value. So we can realize the impact of this on, you know, the organization on the company, and just tie it back to the business value that, hey, if we become more efficient, and do things like this, we can actually increase the amount of subscribers that we can, you know, support with the same amount of staff, and we can, you know, make sure that customers are happier, because they're getting off the phone faster. When customers call in, you know, we know there's a problem instead of them telling us there's one,

Chris Schmitz:  proactive versus reactive,

Andrew Verboncouer:  right, so there's a, you know, a lot of things around it that really, you know, prove that just that one little request, you know, kind of lead us down a rabbit hole of a lot of value and insight into that portion of their business.

Chris Schmitz:  Yeah, absolutely. I remember one other thing we ended up implementing for them was an actual console built into the system. So they didn't need to log in manually and their machine anymore, they could just jump to this tab, which would let them execute these commands. And that way, we could record the commands that they're running against these devices. And we could find patterns and say, Oh, you know, 80% of the commands run on this x. So what are people trying to get from that, and then we're able to just give that that information when they needed it, rather than having it in the right context, rather than right in and make them either login or switch over this console tab premiere, that was something that provided a lot of interesting insights to Yeah,

Andrew Verboncouer:  yeah, and that's another, you know, another way to observe. And that's kind of that next step of, hey, they're doing this manually. Now, let's implement this in kind of, you know, the minimum minimally viable way where we can collect that data, we don't have to pull them about, you know, what, they're what they're using, how often they're using these, we can actually pull that data, you know, the other thing that we did in there as we can created some things within their stats that were essentially fake doors. So, you know, displaying stats we we assume they wanted based on their feedback, and then they would hit a more button, and then we'd ask them, Hey, what do you expect to see here, what do you want to see and collect data kind of writing context as they're looking for specific data, right. So building those kind of things into the app, you know, helps add that insight and capture data as it happens.

Chris Schmitz:  So that's a great example of digging deeper, fully understand the problem coming up with a much more effective solution. And really, you know, like we mentioned, the beginning solution just became completely obvious once once, we're able to do a little deeper and understand the problem a little better. So I want to share one more example about experience I had doing some consulting for a company who's trying to revamp their onboarding process. So they mentioned that it was really Congress, some time boarded new customer. So this was right at the beginning of my engagement with them, and where they had identified a screen in their onboarding process where they wanted to revamp it, and build this drag and drop interface for customers to kind of select from a menu of items with their package was going to be. So here's other things, you can have drag over the options for your for your subscription, and continue through through the onboarding process. So I didn't really understand how that was really optimizing any other workflows. And they can also mentioned that they were doing this internally right now. So they would kind of hand hold their customers as they as they were on boarded and fill out all this information for them. 

So the hypothesis was that providing this drag and drop interface would make it make it easier and allow them to number more customers and not have to do it manually for them. So I asked them, you know, how long does it take time work customer? About four hours? Okay, so how, how long does it take to get through this portion of the onboarding? Five or 10 minutes? Okay, so that that felt a little weird there, how much time do you think this would even save you. And, well, us, probably not much, because, you know, we're, we're comfortable with the existing interface or whatever, but I think our customers will. So again, based completely off of assumptions, right, this new interface would even be better for, for their customers that they were about to have started having a designer redesign this page, and, you know, probably would have resulted in quite a few wasted resources going towards that project. So, and then she said, I wanted to know a little bit more about the onboarding process, and by four hours to to bring on a customer. And turns out, one step of the onboarding process was designing these custom like, kind of like a name plate, these things that would go up in the brick and mortar locations where their technology was being used. And they would have a designer design each one of them, and they'd have to go and deliver them. And obviously, I mean, this is just not a very scalable way to do it. But their hypothesis there was they're not going to want something with our branding in their store, attracting and just distracting I guess, their customers. But again, now is even hypothesis and it turned out, you know, they, they hadn't hadn't tested that at all. And people actually didn't mind that it was that they, you know, if it was their logo and their colors, the company's Yeah, and some, I mean, they were missing out on building some brand recognition through that they were creating all kinds of additional, you know, expenses with the designers to design these spoke pieces, and then all the extra time of delivering them and making sure people use them. And so, yeah, I mean, that whole process was able to go from, you know, four hours down to half an hour, just by eliminating that, that one step.

Andrew Verboncouer:  yeah, right. Yeah, know that. Yeah, that's super interesting. And that's, you know, here, we do, you know, customer journey mapping, which is essentially that, so, trying to understand end to end, you know, what the customer goes through where our product, you know, where our customers product fits in their life, and then, you know, doing an internal audit and external on it. So, how do we think this this specific thing like a nameplate, right, do you think this adds a ton of value to them, you know, and obviously, internally, they thought they did, but, you know, what we would do is go to talk to their customers, and, you know, you would say that, oh, they really don't care. But you can see something that we place a lot of value on, because we think the customer wants it, you know, native, it really doesn't matter to them. You know, that's where a lot of kind of that waste comes in, like you said, where we do things that, you know, aren't, aren't necessarily delightful, you know, we tried to do try to delight our users and design with intent that, hey, they're gonna like this, because they get a custom thing. But if it doesn't matter to them, and it doesn't add value to them, and they don't really enjoy it, right? You know, it's not something that sticks out to them about your product, then it's something that you could probably cut back.

Chris Schmitz:  Yeah, and this reminds me a lot of running lean ash Mari

Andrew Verboncouer:  Ash Mario

Chris Schmitz:  Yeah, he mentioned that you can be fast, and you can be focus, but you have to be it's it's learn, or I forget, learn might not be the exact word. But where you really achieve great results is it's a Venn diagram. And there's an overlap of those three. And that case, they were being fast, and they were focused, they were executing quickly, they would have gotten this project done quickly. And they would have had a nice drag and drop UI. But they weren't learning from it. And from what their their customers could have been telling them, right, and the example and running lean in that sticks out to me, and see only when you're doing those two things, you're really like a dog chasing your tail, you just you just working in circles, and you feel productive, but you're not gonna get any kind of traction, right. It's interesting. Um, so to wrap up here, we, you mentioned a little bit about this technique that we use called the five wise, I'd also like to talk a little bit about problem interviews, I kind of mentioned a little bit of that, too, with your story, some of the interviews you're doing, yeah, I'd like to give some sort really actionable specific things that people could take away from this and start using with their teams.

Andrew Verboncouer:  Yeah, definitely. And will throw this in the show notes will have a little download kind of a template we use for some customer problem interviews, and you can customize it to, you know, your situation, it depends on the customer, depends on if you're enterprise or consumer, but kind of a boilerplate of, you know, kind of where we start just to try and get open feedback. So, you know, the first one is the five whys, and Eric Ries is a big proponent of that just because, you know, taking things at face value. And what users say, and potential users say is, it's pretty dangerous, as we saw in that first example, if you would have done that, you know, we would have just really, you know, taking the pain off the user of putting into periods, right, you know, so we're talking about one or two seconds, something that they did so often that it was just repetitive that they said, Hey, we could just do this, right. And we would have never saw the value of, you know, optimizing that entire process,

Chris Schmitz:  right? is a band aid over something, whereas you could have just cured The, the actual right ailment or whatever, you know, right. So, I think that stuff.

Andrew Verboncouer:  Yeah, so, so going into the five wise, that was an example where we just, you know, keep asking why. And there's a lot of stuff out there, and we'll, we'll throw some links in there for you that are helpful to, you know, how they're using other practices, and then just a model for the five wise, but just making sure that you don't take things you know, as their said, but you dive deeper, and you figure out why this specific outcome is is valuable to them? And what are the kind of pieces along that

Chris Schmitz:  Yeah, and that this technique is amazing to me, because it sounds very childish, the front of it, but it's extremely valuable. And if you do need to do a little more than just, you know, sit there and ask why it can't be just like my two year old son, Isaac. But you do have to put a little bit of time here. But really, the core of it is you need to know why. But sometimes you need to provide a little bit more context to tease out exactly what you're looking for. But yeah, with a little practice, I think you get the hang of it really quickly,

Andrew Verboncouer:  right? One example that comes to mind that, you know, where it was used is Starbucks, for example, you know, where they had people who were dropping off in the morning because dropped by dropping off, I mean, not, you know, not coming in to buy coffee anymore. They saw their numbers in the morning, we're going down, and the line outside and the drive thru was full, you know, the line inside was full of people that were staying in and trying to drop in. So they kept asking why. So, you know, why aren't you coming to Starbucks anymore? Well, because the, the lines too long, you know what wise line to obviously their people around the way to work, they understand that. But diving deeper, you know, they ultimately implemented a solution where you're able to order on your phone and just walk in and pick it up. And it's already prepaid. And, you know, the friction of them getting out of their car and picking it up was worth the value in them not having to stand in line or wait in line in their car. And they just walk in and walk out, right. And so they're gaining a lot of traction on that. But understanding that problem of, you know, people don't want to wait in the car. And obviously, you know, when you're on your way to work, or you're, you're running late, that's the last thing you want to do. But if you can type in your order, just reorder before you leave home. And it's ready when you drive by Starbucks. It's pretty low friction,

Chris Schmitz:  right? That is interesting, the bottleneck of even just taking the orders, you know, the solution for getting around that. And I mean, you've even seen it some other places,

Andrew Verboncouer:  right?

Chris Schmitz:  Don't know why I can't think of any right now. But I know, I've seen that, you know, order online, up in the store. Yeah, all kinds of places. Now,

Andrew Verboncouer:  when you see it to where people be in line, and they'll be looking at their watch, and then just ultimately leave because it's taking, you know, and coffee shops other than Starbucks, it's just, you know, when your people are ordering specialty drinks, it takes time, you might just want a coffee. And, you know, you have to wait 20 minutes to get it right to pretty bad experience in the users. I so,

Chris Schmitz:  yeah, absolutely. So the other technique I wanted to talk about was problem interviews.

Andrew Verboncouer:  Yeah, I'm sure the goal with the problem interview is just to be able to find out enough data from from the users. And at this time, you're not think we mentioned it on our last podcast, but, you know, you're not supposed to present a solution to them, you know, so you're trying to gather open feedback on as you're trying to get the user to talk, and you kind of, to just ask questions and step out of the way, you know, and then provide a little insight or maybe contact as you go along. But the confirming there is a problem that's kind of, you know, the, the step one or the, the number one goal of the problem interviews, the next goal is to, you know, make sure that the customer sees enough value in it, that they think the problem should be solved. So, you know, a good a good indicator that is really the next goal is looking for ways that they've solved a solved it already themselves. So, have they done it manually, what other substitutes or competition to the US and really trying to understand that to make sure that, hey, this is valuable, and people are trying to solve it through these other means. So, that kind of, is an indicator that the problem should be solved. Yeah, so, the, you know, the last part about seeing, if the customers actively tried to solve it, is just making sure there's no barriers. So, if they haven't tried to solve it, you know, what are the things in the market in, you know, any circumstances that prevent them from solving it, and that a lot of times, you know, advanced, you're learning in the process, and maybe identify some other risks you didn't know were there that you can try and solve. And the last thing to, to remember is that, you know, you can really learn about 80% of what you need to know, on five, five interviews, you know, so talking to five people, I'd recommend five to 10 just to be safe. But you can really start to identify those patterns between the people that you interview and talk to. And so it's really a really valuable exercise that takes, you know, not a lot of time initially to, to really get get out of the building and start talking to people and see if this this problem exists. And people are, are willing to, you know, investor or pay, right, for something like this,

Chris Schmitz:  do you have a kind of a template that you use with the main questions for a problem interview? Or are they custom for for each circumstance?

Andrew Verboncouer:  Yeah, so we have kind of the base template that will use and it's like, five or six questions. And so will will make that available, as I mentioned before, but, you know, it really depends on your circumstance, if you need to ask any other questions. And a lot of times as people answer, you know, kind of sit back and get out of the way. But then also, you know, dive in deeper if something kind of strikes your, your interest or is very, you know, intriguing to you to learn more about that. And it's about having a conversation, but, you know, not trying to be too scripted, where, you know, it's, you're just rattling off questions, just try and listen. And yeah, and you'll get a lot of lot of valuable insights

Chris Schmitz:  create to do you find that you actually use the five lies within your problem interviews?

Andrew Verboncouer:  Yeah, absolutely. So five wise is really, yeah, as you do your problem interview, someone will say, Oh, well, I use let's say, if you're if you're working on a, you know, photo sharing app, or a memory sharing app, and people are saying, Oh, well, I use in order to share memories. I use text, I use email mail, I use Facebook, you know, why do you do that? Oh, because different people need it. And you keep asking why. And you get down to the end. And it's because grandma and grandpa use email your long long distance brother and sister and texts and they're not on Facebook. And so you, you can figure out, you know, why they're using different alternatives to solve the same problem. 

Chris Schmitz:  Right. Well, thanks, Andrew. That's some awesome techniques that you can start using immediately. And yeah, we'll have some more resources. If you want to learn more about the five wise problem interviews. And we'll definitely be talking about more things around this idea of understanding and focusing on problems first on future episodes. So I think that's all we have for today. Again, I'm Chris Schmitz. Keep up with me. @ccschmitz on Twitter. 

Andrew Verboncouer:  I'm Andrew Verboncouer. You can find me @ averbs. Thanks for listening to Seaworthy. Connect with us on Twitter, @seaworthyfm and make sure to subscribe, ask questions and leave feedback on the remarks app. We'll see you again in two weeks.

‍