My Software Crashed (And I Couldn't Stop It)
5 May 2026 · 20 min
Mike Andes describes a 20-minute outage of his HomeWorks platform that affected thousands of users. He breaks down a three-step crisis framework: control what you can (your emotions and communication), inform every stakeholder in real time, and conduct post-incident recon to find root causes. His metrics show a 90% year-over-year reduction in downtime after applying this loop.
Chapters
This morning, my software company, HomeWorks, crashed. It went down for 20 minutes and no one could access the software. In this video, I'm going to share with you how I deal with crises like this. And for you, it might be a customer leaving, leaving a bad review, someone threatening to sue you. These tragedies, these crises happen in our business all the time. I'm going to give you a 3-step framework that I use to mitigate the disaster and hopefully learn from the mistakes that you make And right now, like, HomeWorks processes over $1 million of revenue every single day in the software, hundreds of millions of dollars of revenue every year for the business owners that are using it. And so any time that it goes down, it has a big impact on the business owners. And all of my content is built to help home service business owners get rich. And so any time the software goes down, it has a negative impact on their business. It's probably the worst feeling ever. Immediately having thousands of business owners and all their team members and employees being dead in the water is probably the worst feeling I've ever experienced in my career. Now, you might not have a software company, but you are going to have crises in the business. This comes in the form of someone cussing you out and walking off the job. This comes in the form of someone threatening you on a Facebook post, you know, in a public group saying all sorts of negative things about your business. This comes in the form of an employee saying that they're going to steal from you or take away your customers. These crises happen all the time, and many times when they happen, it's like a gut punch to you. I still remember the very first negative response or negative review that I got from a customer, and it wasn't even public. It was just to me privately saying, "Hey, quality of your service hasn't been up to where it needs to be, and I really like doing service with you, but it's been a year and a half of us working with you and now it doesn't seem like it's so great." And that was right after I started hiring employees. And they got used to me doing the work myself as a solo operator and were now discouraged and unhappy with the fact that my employees did not do the same level of quality as me. When I went back to do the job, it looked really good. I was like, there's actually nothing here I would change. And what they actually had a problem with was how they conducted themselves while doing the work. They weren't moving as fast as me. They weren't as efficient. They weren't as tidy up. They had some holes in their pants. And it was like little tiny stuff, but nonetheless, when the original communication came to me, I still remember it was like a punch in the gut. I still remember exactly where I was driving up to a T-junction in Blaine, Washington, when I got the email that this customer was unsatisfied. And that feeling, only a business owner can really know when you are struggling financially and then you also have this huge cater, this thing just like barreling down on you, whether it be any of the crises I've mentioned. Now I'm going to talk about how I deal with it and hopefully use a step— 3-step program that you can kind of learn from these mistakes and make sure you don't burn the business down. First off, like, I'm doing this for the first time ever, sharing this the day of. Usually I wait 6 or 12 months to share this sort of stuff, um, but hopefully it being a little more raw is a little more impactful for you. So if you are going through a crisis now the emotion that I have now of just literally coming out of it a couple hours ago is felt by you that's going through the exact same feeling today. Let me talk about HomeWorks for a second. The hardest part about HomeWorks and any business that you have an online presence is all of your customers are literally in a Facebook group together, and anytime you make a mistake, it's going to be compounded. One of the things that has always sort of made me realize why I get so stressed out is think about all your customers that you care about, that you service on a weekly or biweekly basis, hundreds, maybe even thousands. Think about all of them in a Facebook group and every single time there's a mistake that you make, an employee runs over a sprinkler or someone backs into someone else or they mess up something on a bush or a lawn or a fence or whatever it is, every single time there's a mistake, people in your customer base just go to a public place where all of your other customers are at and just post there. That's essentially what HomeWorks and the franchise are for me. For me, I'm not the type of person where I see negative feedback about me. I'm like, oh yeah, like haters. Like, I'm not that person, especially with people I care about, people I love, and people that support me. They're my customers. I care deeply about them. And so I will first say this: if you are going through a crisis and it has to do with customers, realize that the reason you feel so horrible is because you care. And that's a really good sign. Secondly, if it has to do with your employees and you also feel really horrible and like me last night, didn't sleep, keep in mind it's because you're probably a decent enough person to constantly question yourself. Did I do something wrong? How could they be thinking that? Did I do something wrong? Like, you're constantly second-guessing yourself. And although that is not constructive or conducive to making progress in the business, it shows that you're a human. And so whenever I am going through these things, and I have cortisol pumping, I can't sleep, I'm agitated, I'm stressed out, I always go back to the fact, at least this means I'm human. Because I really do care about my customers. And if it ever happens with an employee, which thank goodness it doesn't happen as often, the same thing is true. At least I care enough about this human that I feel horrible and I second-guess myself. Because if I was just a robot that just operated based upon the book and I didn't care about everyone's emotions, then the crisis that you're currently going through, you wouldn't feel anything. You wouldn't feel the knot in your stomach, you wouldn't lose sleep over it. And so just as a side note, I would say you are human. I know what it's like to feel this way. And I think anyone that actually cares for customers or their team is going to feel this way, even if they're right, even if you fired someone. But on their way out, they said all sorts of horrible things about you, even though you're right, even though by the letter of the law and the employee handbook and the law in your state or your province, you are supposed to fire that person and you're okay legally. You still feel bad. I think that's a good thing. I think that is a sign of someone that is still a human and able to show empathy for the human, not necessarily the things that they did or the things that they did not do. At HomeWorks, we have reduced downtime by more than 90% year over year, meaning for every minute that we were down in 2025, we're down 6 seconds in 2026. So overall, we're doing really well. Here's the problem about crises, or I don't even know how to say it. I think maybe you just leave an I, I don't know. It's a crisis and then it becomes a crises. The thing about this, it is extremely acute when it is happening. In software, we call this an incident. And I think anytime there is deviation from a customer being satisfied, which is the whole goal and purpose for our business existing, is to solve problems for the customer. And if that is the goal and we do not do that for them, This creates pain. I would create this as an incident. An incident is when we do not meet the objectives of the business, which in this case is almost always to solve the problem for our customer. Now, I'm going to talk about the software and HomeWorks, but I think my hope is that you can relate this in your own business. When a crisis like this happens, I'm doing 3 things. First off, I'm asking myself, what can I actually control? Because as your organization grows, there's actually less and less that you can control without getting in the way of the people solving the problem. I'm going to use software and I'll bring it back to home services in a second. So I cannot solve the problem that this morning we had with huge indexes and queries hitting our database and slowing it down. I can't fix that. I do not have the technical ability. I'm not good at infrastructure, definitely on AWS and cloud infrastructure. I don't know how to solve the problem. I cannot control the problem. But the question becomes, what can I control? And the things that we can always control if we are going to take extreme ownership is we can control our emotions and we can control our actions. This is very difficult to do, especially when you can't solve the problem. This is probably the thing that I have had to figure out, or I've, I'm trying to learn and have had to learn over the past couple years the most. When there are crises that I cannot solve, That doubles or triples the amount of anxiety that I have and the amount of pain that it causes because I can't solve the problem, but I can control other things. How I communicate with my team, how I come across in terms of my emotions to customers, to employees, to everyone else involved, all parties. What can I control? Assessing this is most important. Now, if you're in home services, you might be able to go fix the problem, especially if you're very small. You could— let's say a customer calls in, they're super mad about something. Ideally, you solve the problem as fast as possible to reduce the crisis and the acute pain that you're going through. I would recommend this tremendously if you have 1, 2, 3 employees and you can just quickly go solve the problem. Because the longer the problem, the longer the crisis happens for the customer, the bigger the problem in their mind becomes. And so the longer you let it fester, the longer you let an online review or a negative response or customer being unsatisfied, the longer you let that go, it'll just double and triple and compound and they'll just get mad, more and more mad, and they'll catastrophize and now like everything in the world is collapsing down because you put a skid mark park in my lawn. We have to ask ourselves, what can we control? In a smaller business, many times you can go solve the problem. This is when you do drop everything and you do go fix the thing. As your business grows, you can't. When you have 20, 30, 40 employees, when you have multiple locations, when you have managers in place, you taking control does not mean solving the problem. You staying in control and controlling what you actually have the ability to control is many times the way you communicate and your emotions. And keeping those things in check is very difficult when you can't go solve the problem. If you're 4 hours removed from your business and a customer customer complains to you, you can't go solve the problem. You have to control your psychology, your emotions, your communication to make sure the people that are inside the organization, they go solve the problem and the incident is resolved as quickly as possible. And so asking yourself, what can you actually control? How am I coming across to my team? How am I coming across to the customer that is complaining? That's what I can control. And me being anxious, me have the knot in my stomach, me being mad, is not going to help the situation. And so reducing the amount of emotion that I add when I know I cannot solve the problem has served me well. Whether it be to the internal team, there's no reason to get mad at them during the incident. If you want to get mad at them later, fine. Maybe there's a reason to put pressure on people. That being said, during an incident, during a crisis for the customer or for your internal team, usually adding emotion makes the perception that you are out of control, and therefore they then when they try to solve the problem themselves, they come across with anxiety, emotion, and usually don't solve the problem as well. So the first thing I always think about when a problem happens and when it comes to software, it comes to franchise and it is affecting tons of people and it's very public, I have to ask myself, what can I control? In this case, when it came to the technology, I cannot solve the problem. It causes way more stress. I have to then realize the only thing I can control is how I communicate with my team and how I communicate with our customers. All right, second thing that I focus on is are all parties informed? As the leader of the organization, your goal during an incident when you do not have control is to make sure people are informed. The reason in a small business you can take control and solve the problem very quickly is because you are fully informed and you are therefore the most capable person to go solve the problem for the customer. As the leader, when you do not have control, like this morning, when engineers need to be jumping on it and they're swarming and I have to ask myself, are we informed? First off, at HomeWorks, for example, we have automatic responders and notifications that will wake people up in the middle of the night if necessary if certain things happen like what happened this morning. And so when this was happening, we were actually watching it because we had already been informed and notified that, hey, something's happening on the database and it's slowing down. So there was already engineers that were swarming before it actually crashed because we started seeing the degradation of the server and it's slowly getting slower and slower and slower. It actually kind of started yesterday, just a slight bit, and that triggered the notification and we've been watching it for quite some time. As a leader that I cannot actually fix the problem, I'm trying to stay in control of my emotions, I'm trying to then make sure that people are informed. Now this is all parties. This is your team, this is also customers, okay? So for, I'll use my example and I'll jump to Home Services. I was making sure that the customers that were having problems, I was immediately responding and just letting them know, here's the information that I have, and we're working on it. And a lot of times the biggest thing that your customers want to see is that you are problem aware. You are aware of the crisis and the problem that they are having because of the mistake you made is important to you. And so we— I've tried to make sure that the level of intensity and the level that the incident is affecting our customers is the same level of response and information that is conveyed to the team. So if it's like what we call a P1 incident, like the worst, like you can't use the software like this morning, that is like raising all sorts of bells, people are getting texts, people are getting calls, everyone's swarming, it's very focused. So that was the first step of making sure people are informed, that is when there's an incident and when there's a crisis, everyone that is correctly inside of the team is correctly informed. We're also working in the future to make sure that customers can actually see, they get the information, instead of like, is it down, is there a problem, is it slow? The ability for them to be able to see this down the road would be a way for me to make sure that they are informed of the problems and the crisis that's happening. During the actual crisis, making sure that the customer also has information and they're informed of what's going on. We're on it. Like, this is not something we're brushing under the carpet. We're all on it. Like, we have full attention. Here's the information I have from a non-technical perspective. Here's what's going on and we're diving in on this. That's the next thing I control. The last thing, and this is what I am working on today and the whole team is working on today, is a recon. That is after the acute pain, after the crisis, after the incident, after the problem has been solved, and after 20 minutes this morning, the software's up and running. We gotta do recon. Why did this happen? What systems do we need to put in place to make sure that this never happens again? Is it because there was lack of information? Was there someone lacked control? Was there a fact that we pushed an update or a feature and it caused a regression, which means we pushed something thinking it was positive and it made something else break. And so a recon is really valuable. In your home service business, the same thing's gonna happen. There's gonna be a customer complaint, there's gonna be a huge crisis, an employee walk off the job. The recon is why did this happen? What do we need to change about the business systems? What systems do we need to change in the business to prevent it from happening? And if it's an employee-related issue and someone freaking out at you and you're firing them, the recon is can we be clear about training? Can we be clear in onboarding? Can we be clear in the hiring process? Can we be clear in the employee handbook? These are things that in the recon, I then make sure to mitigate the disaster. So, the only insanity in business is constantly staying in this loop, which is crisis and you're out of control and then you're just like spreading information. This is the constant crisis. This will drive you crazy. It'll also drive your customers crazy. The recon is the unlock to make sure this this loop is prevented and we prevent or we mitigate against this disaster. How can this disaster or this crisis be avoided? Whether that be cutting out a service that you constantly are making mistakes on and it's causing crises. For me, it's a matter of, okay, do we need to just increase the server load? That was the first question for Recon was like, is this a matter of us just getting in a spring rush and we, like, the servers just need to be doubled or tripled? And like, no, this is a matter of, the indexes being incorrect, and it doesn't matter if we 20x our AWS bill, it would just create hundreds of thousands of these and immediately fill it up no matter what. Okay, bad idea. That's not going to fix this crisis from happening again. Okay, is it a way of informing our customers better, informing our team better, faster, more efficiently? Were the right people informed during the crisis? Did I get out of step and did I miscommunicate or lose control by having the inability to be able to keep my emotions in check. Doing this process, this 3-step process, and doing a recon allows you to reduce the amount of times you end up in crisis. And we make mistakes, a lot of them. We are not by no means perfect. We are missing features for larger operators. We're trying to work hard on a lot of chemical application and snow features, a lot of stuff. We are by no means perfect, but I have to rely on the fact that we have reduced our downtime by 90% year over year. And when you're going through a crisis, it is by no means consoling, but afterwards you have to step back and look at the data. And that is, are you improving? And so tracking the churn of your customers, tracking how many complaints you do have, tracking how many 1-star reviews you do have, and then asking yourself, if I do this loop enough, that I go through the crisis, I make sure people are informed, and I recon and fix the systems inside the business to prevent this crisis from happening again, Does it actually yield results? And I believe that it does. If I can take from what in the past was 10 minutes of downtime, it's now averaging 1 minute of downtime, which is still 1 minute too long. If I can do that year after year, eventually we'll get to a better place, and we will serve our customers better, and they will solve more of their problems, and crises will happen less. And because, because in crisis, no one is served. It's all disaster, it's all horrible. Like the team, like they should be working on features. They should be working on improving the software, making it faster, making improvements. The same thing is your business. Like all of a sudden you're pulling crews off of jobs and throwing them at some random job because someone complained or someone freaked out or there's a damage case. It's not ideal case for the business. You're losing money. The team is frustrated because they're— you're pulling them off of jobs. It's not good for the customer because they're mad. If they go online, start sharing it publicly, that could lead to a 10x impact on your business. Because now not only is existing customers or that one customer, if they go share online, now 10 customers that would have otherwise joined your service don't even show up. And so this actually affects your bottom line. This affects whether or not you even are able to attract customers. This affects your customer acquisition costs. This affects your close rate because now there's negativity about your business. All of this is flowing through your veins, all of the cortisol flowing through your veins when you're in a crisis. This basically just says you're not alone, and if you can use this process, you will at least reduce it over the course of time. And if you do this enough, If you do this for 5, 10 years, eventually you get to the place where this is really, really good. And the only hope that I have for, like, Homeworks, for example, I'm less than 3 years into it. For Augusta Lawn Care, we're 6 years into it. For Augusta, our NPS, Net Promoter Score, was 80. That's really, really good. If you look up Net Promoter Score, 80 is really good. At Homeworks, we're only at 30. And the only way to improve that Net Promoter Score be able to improve my customer acquisition costs, improve the experience that my customers have, improve the size of business they can grow inside the software, improve their efficiencies, is to constantly feed this loop over and over and over again. And so often at times, the reason you want to quit is because we are caught in this loop. But if you actually complete it, you stay in control, you figure out what you can actually do in the moment, you keep everyone informed, all parties including your team and the customer, and then you recon and make changes to the business and the systems, Over the course of the long term, the long term, the business will work out.
More from Mike


I Sent My Head of Sales to Hormozi's First-Ever Offers Workshop | C.Suite Unfiltered 117