Bootstrapping to $2M Without the Chaos: Lessons on Momentum and Focus

0:00So, as Rob said, I'm going to talk a little bit about our story, um, bootstrapping to over two million and how to build momentum in your startup as you continue to grow. We are almost there, I believe, on the last talk of the conference. Uh, and, uh, you'll be walking out of here soon, hopefully, uh, motivated and with some useful tips, but if not, at least you'll be walking out of here. So, who am I? Uh, for those who haven't, uh, met me yet, my name is

0:26James. I'm the co-founder of Estalty. I'm also uh husband to my amazing wife and she is at home looking after our three kids. And I did just want to give a quick shout out to all the partners and families that are probably not in this room because doing this is hard and it takes a toll on your family and I think they deserve a lot more

0:43recognition than what they probably get. [applause]

0:51when I'm not building uh when I'm not spending time with my family, I am feeding these two which is absolutely uh expensive. Probably the second biggest bill that I pay apart from AWS each month. And as you probably know, I am from Australia. Uh you might have heard of Sydney or Melbourne or even Brisbane. [snorts] Uh I'm not from any of those towns. I'm from a small country town called Dolby. Uh it's about 10,000 people there and our team is fully remote. We've been living there for about three and a half years and I

1:21absolutely love being out of the city. Uh so don't think that you necessarily need to be in a startup hub to actually build a startup. So why am I here? Well, Rob uh briefly mentioned that I was on the podcast, but I actually stumbled across the podcast in 2021 when I used to drive for about 2 and 1/2 hours to the airport to travel to uh inerson events and chat with my co-founder and things like that. And uh I love listening to podcasts and I stumbled across startups for the rest of us one way or another. And I was thinking to myself, what on earth is MR

1:54and LTV and retention and product market fit and all the other SAS terms? I actually had no idea, but I was building a SAS company. So I thought I better listen in a little bit more. And uh and obviously there must have been a pitch for Microcom. So uh ultimately I booked my first ticket to Micro in Atlanta 2024, excuse me, 2024. which was an incredible, incredible event. Uh, and I thought I'll really step out of my comfort zone and uh, and be a speaker at

2:23this one. So, on Friday, I got in a tin can and uh, traveled for about 25 hours all the way from Australia to to here. I think I'm finally over my jet lag. It's been an incredible two days. Thank you to Sonia and the whole Micro team. Um, it's an incredible event and uh, really

2:39honored to be here speaking. So, enough about me. Uh, and let's talk start talking about Astaltty and our journey. So, on the podcast, Rob actually said Asty is a bad acronym, but it is an acronym and it stands for a solution that actually listens to you. Uh, it's co-founded by myself and Jonno. And, uh, the best way to explain it is essentially it is a it's it's a HubSpot for the NDIS. So, it's a super specialized uh, CRM. And what on earth is the NDIS? So the NDIS stands for the National Disability Insurance Scheme and it's essentially a governmentf funded scheme that gives funding to people that

3:19have a disability to then go and uh access uh services provided by businesses to help them live more independently uh to help them at home to kind of navigate the NDIS in itself and uh and our software essentially helps those businesses delivering their services to run their whole business. And you can see some of the features

3:40that we uh have down the sidebar there. So it is quite a large broad um software but within a niche in Australia. So how on earth did I even come across building such a specialized tool. This is a chatbt image which is fantastic. But I want to go back to the origins of Estalti and Estalty kind of started around 2018 but in all honesty it started before then. Like all good startup founders, I dropped out of uni once. Not once, not twice, but three times. Uh the first time I failed, the

4:15second times I dropped out voluntarily. And I ended up actually running pubs for a few years. And in 2018, I said, "This is enough. I was working nights. It was not healthy for me, not healthy for my relationship. Uh and I decided to call it quits." So I was sitting at home and thinking, uh, I need to make some money. So, I went online and uh there's a platform called Air Tasker and it's similar to Upwork, Fiverr, kind of a gig um a gig, you know, marketplace. And I knew nothing about software, but I did know about spreadsheets and I did know about QuickBooks. And I found this job

4:49and I thought, great, I can make 50 bucks. Uh I must have done a good job cuz this says 100 when I went and search for the screenshot. So, uh I must have got a bonus for for doing a good job. And the spreadsheet was just a basic kind of transformation. But this is a very important day because this is how I actually met my co-founder Jono O who has a business [clears throat] that

5:08operates within the NDIS. So in March uh 2021 we registered the company and between 2018 and 2021 I had another software company where I was doing freelancing and essentially building custom tools for businesses. Uh but in March 2021, uh we decided that we would um or before then, sorry, we were joking about how bad software was in the NDIS and that perhaps we could build something. Uh and ultimately in March 21 is when we pulled the trigger, which was also a busy time. Uh similar to Kevin actually. Uh I literally signed the paperwork in a hotel room across from the hospital uh the same week that our

5:48first son was born. So between then and September uh I was winding down my other software company and essentially starting to write code for a stalty and in September the first commit was made to GitHub. Like all good software developers, I uh perfectly estimated how long taking it would build. Few technical founders in the room uh of course completely underestimated what was involved uh and and how long would really take. And we initially started to build you know everything that we needed and we started to realize that we were going to do a very poor job of that. So we scoped down within the NDIS uh to one

6:28of the 15 categories called support coordination. Uh which essentially uh meant that we were building a tool now for a segment of a segment. Uh so it was a niche of a niche if you like. But beyond that we still knew it was going to take too long and we wanted to start building our reputation and uh getting our name out there. So we thought uh why don't we build some free tools and free resources that allows uh customers to really you know to get our name in front of them and start trusting us and knowing who we are. Uh the NDIS has this catalog of of prices you can

7:02charge for. It's a big spreadsheet. It's not that important except for the fact that it's super painful to deal with. And uh I knew that customers that would ultimately use our software must be using this spreadsheet because I was using it building the software that they would use. And I had an idea of building a Chrome extension that kind of built this spreadsheet into your browser. You could search it really easily. Uh and this is the first, you know, product that uh that we ever built. We launched this even pre-AI. I think this took one or two days to build and we sent out a cold email and uh and it actually did

7:34quite well. A quick tip for anybody building uh free resources, make sure you get email addresses at least. To this day, we still do not capture any email addresses, which is a big regret from this extension. Um, in all honesty, we probably need to revisit it sometime soon. But if you're going to give free stuff away, at least capture something

7:51that can help you. This is our chart of usage. Uh, so in January 22, we launched the free tool. We got a couple hundred users. And then in February, we launched a Stalty Light V1.0.0. The V1.0.0 did not last very long. Uh, we got rid of that super quick. design is not my strong point, but it was a simple free solution for support coordinators. That's what we

8:14initially marketed it as. We had about 340 customers sign up to a Stalty Light. And of those, we had a few dozen, maybe 30 or 40 customers actually engaging with us, telling us what the platform needs, what it's lacking, what it's doing well. And even though I didn't know it at the time, cuz I didn't even know what the term was, this was an

8:33indication of product market fit for us. So between then and August 2022, we were listening to all of the feedback and we were building ultimately what would be the first version of our paid offering. And in August 2022, in true startup style, we launched a Stalty Professional the day before we were going to showcase it in an in-person event in a two-bedroom Airbnb. And this is a photo of Jonno and myself uh with a make make

8:59do desk and couple of laptops. So between then and uh October the next year, we were simply doing what didn't scale. We were going to the inerson events. We were literally building features on demo calls. Jonno would be talking sales. I'd be writing code. By the end of the demo, we would have a feature and ultimately close that customer. We were working anywhere between 15 to 18 hours a day. And for us, we were doing a lot of phone support because none of our competitors offered phone support. And we thought if we did

9:28that, uh we would be able to stand out. and I still get phone calls on my personal mobile from the earliest customers. It wasn't until in October 2023 that we hired our first full stack developer. Up until this point, I had written every single line of code and we had about a,000 users. Uh so this is roughly I don't know 40,000 a month um when we made our first our first hire. And this is just before we had a massive transition. Uh, I completely rebuilt the front end of a Stalty, but we're also going into the launch of a Stalty Premium, which is what I like to call a

10:03price rise without actually doing a price rise. So, what on earth does that mean? At the time, uh, Stalty Professional was actually $49 per user per month. This is an old screenshot, but it's the best I could find. And we wanted to increase prices, but, uh, I I hate conflict. I hate saying no. Um, so what we decided to do instead was we drew a line in the sand and we said every new feature from this point on is going to go into the highest tier because we knew that they were high

10:30value features. What [clears throat] we did is we then introduced a a third tier. We didn't launch with three tiers. We just had the free and professional. We launched with the stalty premium which was an extra $15 a month. So instead of increasing our prices by say 5 or 10% which is what is dictated by our terms, we were able to essentially get a 30% increase in revenue simply by building things we were going to build anyway but putting

10:53him putting them into a higher tier. What we then did is we kept a salty professional but increased the price to $59 a month. So it's only an extra $5 a user and what we found is that 98% of all new customers were signing up to our highest tier anyway. So eventually we just got rid of a Stalty Professional and we had a 60% upgrade for existing customers. So 60% of all users that started actually paying us $29 a month because we had an introductory offer upgraded to paying $64 a month, almost triple what they were initially paying

11:25us. So during this time we really focused on these nine things. Number one, delivering the best customer experience. And this is going to look different for every single business, but for us, this meant getting back to people extremely quickly. Uh we didn't have any support platform. It was just email, but we were getting back to customers within 15 minutes. We were in Facebook groups. We weren't necessarily pushing our product, but our customers, they kind of have these Facebook groups. There's probably about a dozen of them that they hang out and talk to each other. And when we stumbled across this, we thought it was a fantastic resource. So, we were in

12:00these groups listening to what people hated about their existing software. And we would go and build those features, reach out to those customers. And now if you go into the some of these Facebook groups, customers will or potential customers will say, uh, what software do you use? And, you know, 40 out of 50

12:15comments are now recommending a Stalty. Uh, and I know you can achieve similar things in Reddit and other social platforms. We were going to all of the Inerson events, which does not scale. Uh, as a founder, it's extremely tiring, especially when you have a young family. Uh, but again, it's just one of those

12:29things that we felt we had to do. Uh, we built an affiliate program. We also built a partnership program and what that allowed us to do was stop being a place of no. We are a software company. We do not provide legal HR advice. We do not build your website for you. We don't help you set up your emails. But a no for us is quite, you know, there's a lot of friction there when a customer comes to us said we love your software uh but no, we can't do that. So instead, we now have a uh partnership program where instead of saying no, we can say unfortunately we

13:00don't, but here's a really great uh partner of ours that you can go and approach. And it just meant that customers now trust us. They can come to us uh and we, you know, we're very approachable and having really great customers that are happy to reach out to us for any of their issues is very

13:13important uh to the way that we work. We started to grow our uh our LinkedIn network by talking more about what we do. Uh and I'll talk about that in a little bit. We focused on getting customers for life. Uh and again, this is something that looks different for everybody. And I can give two examples of of what I mean by this. The first is uh we once built a feature in 7 minutes for a customer. I was on Facebook uh in our community and I was I think I was responding to a comment and I saw another customer comment about something that was super painful and I can't

13:46remember exactly what it was uh but I saw that comment. I had my editor open and I literally shipped a feature and replied to that comment and it all happened within 7 minutes. That customer then went into several Facebook groups and spoke about this experience. So by doing that one thing that cost me, you know, 5 10 minutes of time, I've now essentially got a viral loop because I have a customer now going and telling their network and then those people telling their network, so on and so

14:13forth. The other thing we did far too late was introduce annual pricing. And the way that this allows you to move quickly is it gives you an influx of cash that you can use to execute on expensive things like recruitment, consultants, compliance, infrastructure, and things like that. So by introducing annual pricing, we now have a cash influx twice a year that really helps us uh execute on all of those expensive things uh and spending money which most Bootstrap

14:40founders including myself hate to do. It wasn't until April 2024 when we had over 1500 active users that we actually hired our first customer uh success rep to help out with support. Uh, similar to Kevin, I probably wouldn't suggest doing this, but I do wonder if we weren't doing the support ourselves, would we have been successful if we weren't so close to the four, you know, to the feedback that we were getting from our customers. Uh, so I do think that it was important that we did the support, but perhaps we could have hired a lot earlier on because doing customer

15:13support can absolutely take a toll. So between then and now, we've really just been focusing on scaling the business. We're at almost 3 and a half thousand active users across 1300 businesses. Uh, and we're now solving other problems such as scaling the team, um, scaling communication, other things that isn't simply writing code. So, we're doing about 2 and a half

15:35million ARR. We have a small team of 10. That's made up of five engineers, two co-founders, one in sales, and uh, and two in customer success. We do 24 deployments a day to our production uh, application. And this is actually really important. There's a book out there called Accelerate by Jez Humble about software engineering. And there is a direct correlation between deployment frequency uh and your ability to respond to sorry like lead time of building new features, meantime to recovery and change failure rate. In short, the more you deploy per day, the faster you can build things and the less bugs you will have. We have just over,200 subscribed

16:14businesses and we have a uh net monthly negative churn of minus 2.4%. Which is uh you know I think is a uh key indicator of our customer experience and our customers being able to grow with us. So what's different now? Net revenue retention. Yes, I do mean retention. Sorry, that would be very scary. Yes, I think I had a negative churn and then I swapped the number and uh I I didn't change the churn, but um what is different now than it was 18 months ago. And for us, the reality is not too much. Yes, we have more revenue, but more revenue doesn't necessarily mean less problems. It just means

17:01different problems and you can solve them a different way. So, not much is different. Uh we continue to do what works. We continue to not do what doesn't work, deliver the great customer experience, and build really great software. And for that, we're rewarded with an MR chart that looks like this. And I wanted to take a screenshot of this because I wanted to indicate that there's no hockey stick growth. There's no massive spikes. It's just nice, steady, linear growth, which can get you into a great spot. However, there are some days that still look like

17:33this. It is still hard. There are still problems to solve. Um, but it's been a great journey and I don't think I'd be doing anything else. So, how do you actually build momentum and continue to move quickly? First, I wanted to touch on the definition of momentum, which is the quantity of motion of a moving body measured as a product of its mass and velocity. Or in business terms, I think it's the capacity for sustained growth measured as a product of a company's size and the speed at which it can execute. So you can have a very small company executing extremely quickly and we're seeing this so much now with AI

18:12and I think there's teams like Cursor and Lovable and all these teams of like 10 20 or 30 people doing tens of millions of dollars ARR but small companies that move extremely quickly uh can really punch above their weight. But if you have a massive company that starts to move really slowly, it's startups that are in this room that will start eating them for breakfast. And what I'm focusing on now is making sure that we don't become one

18:33of these companies. momentum compounds. You only need to be 1% faster every day and you need to do that for weeks and months and years. And at times it doesn't feel like you're moving any faster. But if you continue to stick with it, you will absolutely see the benefits in years to come. And when your startup is successful, you will thank yourself for doing the small

18:57things early on. So, one thing uh that I pointed out before was deployment frequency. You need to be deploying often. The only way you can do this is by using feature flags in my opinion. Uh if you're not using feature flags, you should introduce this into your codebase as soon as possible. Even if you think you won't need it, uh when I get to point four about shipping straight to production, feature flags, feature flags become extremely

19:24important. Keep your infrastructure simple. So many people jump into Docker and Kubernetes and you know all these other complicated things. Uh we ran on a single EC2 instance for an embarrassing embarrassingly long time to be honest and it was only until we had compliance requirements to have redundancies that we actually even made our infrastructure more complicated but it is still

19:47extremely simple. We have a few servers. We deploy using you know um tools that are provided. Nothing complicated. Lean on managed uh pro providers where you can like Laravel cloud, Heroku, Netlefi, um Versel platforms like that. Have very good seed data. And this is extremely important. You can use and when I say seed data, I mean you should be able to spin up an example of your application within 2 or 3 minutes that looks

20:13completely real just with fake data. This helps you move quickly because your sales team can use it in calls, your support team can use it to build out your documentation, and you're no longer slowed down by having to manually enter things into your software to get something that looks good. This might be a spicy one, but ship straight to production. And a lot of people that start out probably do this, but as you grow, you will feel compelled

20:38to introduce a staging or a dev server. You do not need them. you can ship straight into production and you can put processes in place to ensure that code goes from your local environment and is pushed straight into your production environment using feature flags, automated testing and things like that. Uh and again that book uh Accelerate by Jez Humble talks about this and why it's

20:57important. Take action. This is simple advice but so many people get stuck analyzing things and not moving fast. You need to be okay with being wrong. When you're making decisions quickly, you will naturally make more decisions. The more decisions you make, the more likely you are going to be wrong. And that is completely fine. The only way you're going to know you're wrong is if you do the thing and somebody tells you because as a startup founder, you just don't have all the answers. Anybody that says

21:25it does, they're lying. Limit the number of one-way doors. Uh I think Jeff Bezos talks about this, but everything that you do should ideally be a two-way door decision. If you make that decision, you should be able to reverse it and do it the other way. And when you limit one-way doors, again, it's being it's easier to be wrong because you can just go and do the other

21:43thing. This is one of my favorite ones from 37 signals, but on the side of do I really struggled with this as an early founder about being an analysis paralysis. Should I use Vue or React, PHP or JavaScript? And at the end of the day, use whatever your gut says in terms of building code. Use whatever you can build with fastest, but don't get stuck in an analysis paralysis. Just do the thing and if it's wrong, do the other

22:07thing. It's no big deal. And when you are in that stage of kind of choosing whether you should do option A or option B, just choose a simpler option. This is a big one. Uh and this is something that I struggled with for a long time, probably until the last 12 months. But you should talk about your work. And I don't mean you need to post your MR charts on, you know, Twitter or X or LinkedIn or or whatever you want to use. Um but you should talk about what you're building. It builds trust with your potential customer base and it allows you to give an insight into your

22:37actual journey and people resonate with that. Building trust makes it a lot easier to actually close deals. Uh involve your family. I uh obviously your family is kind of along for the ride, but for me personally, I didn't actually talk about work with my family for a very long time until about the last kind of 6 to 12 months. What I realized by actually talking to my family, my wife actually, you know, she doesn't really know anything about SAS or or or technical, but sometimes you get the best advice from people that have a completely different perspective. Talk to your family. Talk to your friends about what

23:11you're working on. Even if they don't understand it, sometimes they do have the best advice. Keep your channels open and public. And when I say public, I don't mean public to everybody. Just public within your workspace. If you use Slack, have all of your channels open. There's no need to have siloed information. It just slows down the decision-making. If everybody can access everything, it means that your engineering team can see the pain points that customer uh the customer support team are seeing. Customer support team can see the things that the engineering team are fixing and it actually gives insight and motivates people to move quickly because the

23:45success and support team can see how quickly the engineering team are shipping and in turn that makes them move quickly. Record as much as you can. This is extremely easy now with AI transcription tools and things like that. And I wish we did more of this earlier. I can't tell you how many times I've thought about a decision that we made two and a half years ago. And I thought, I really wish we had the transcript or the recording for that. Uh because we could just use that to then make the next decision rather than sitting down and being confused about why a decision was made, which ultimately just slows you

24:13down. If you can, you should. And I mean this with regard to your customers. If you can do something, you should do it. And the best example is refunding. Uh in our terms, we don't refund annual deals, but I can't think of a time where we've actually declined a refund. If somebody's asking for a refund, in my opinion, they've already churned. The least the the best thing you can do is try and have a positive outcome by refunding them. We've actually had examples where we've had a team of one come to us, they sign up, they close down their business, we refund them, they then went to work for another

24:47business, a team of seven, that then came and used our software anyway. if they had a bad experience leaving us. I do wonder if that ever would have happened. Fix bugs extremely quickly. Again, this builds trust with your customers and it also means that you don't have your engineering team worrying about how important is a bug and then they again slow in making

25:06decisions whether they work on or not. Every bug for us goes to the top of the list. We have a zero bug policy. It becomes the most important thing. We build that then we get back to whatever we were working on. Call your customers. I challenge everybody here today, within the next week when you're at the grocery store, on the way to pick up your kids from daycare, whatever it is, pick up a customer and just call them. It's those phone calls where you'll get some of the best feedback because at the end of the call, they'll say, "Everything's going great, but and then they'll tell you

25:34about something that's super painful that you can fix for them that you never would have heard of otherwise. And lastly, be honest." Again, this just makes it faster to communicate with your team and with your people. When you're always telling the truth, it makes it [snorts] a lot easier just to communicate. If you want to work, work. Again, this is one of my struggles and I'm not sure how many people fall into this bucket, but for so long, I felt guilty for working weekends and working nights and taking my laptop on holidays. And eventually, I just gave into that. And I thought, well, it's going to be less

26:05anxiety for me to check my emails, for me to actually do the work than sit there and feel bad for, you know, not whipping out the laptop. So, if you do feel like working, just work. Don't worry about the productivity, you know, the work life balance police or the LinkedIn police coming to uh to jump on

26:20you for working too much. And finally, the goalpost move. Everybody thinks that entrepreneurship is going up this step ladder from the bottom to the top. You reach the top, your life's great. Uh the reality is goalpost move. For us, it was 300 users. Then it was 500 users, then it was 80K a month, then 100 grand a month. You will always be stuck somewhere between where you were and where you want to be. And those layers will continue to move as you do. So my biggest piece of advice is build for the

26:54journey and not for the destination. Thank you. [applause]