SaaS Factory: A Giant Screwup and Looking To The Past For Inspiration
Time to dig myself out of the hole I put myself in
I fucked up.
There’s no other way to say it. I didn’t want to say that at all, which is exactly why I am. When I first realized how much I’d fucked up, I was almost sick.
To understand what I got wrong, we have to rewind a bit to July. I got an email.
It was letting me know that my AWS credits had run out. I had $5000 worth of credits that I got when I incorporated my first startup. I glanced at the bill and thought it would be about $70 for that month, and thought that I would need to figure that out at some point. Mostly looked to be on my database for Feather., which was also the same time that I was trying to replace a good chunk of the functionality of Feather, because Quirrel had been acquired and was shutting down.
So fast forward a little bit to when I got my bill for July. It was almost $400. I hadn’t budgeted for that at all and had nowhere near that amount of money available. Those AWS billing pages aren’t the clearest.
So I started looking at other options. Then I started getting notices about late payments and so on. I actually did contact them and ask for advice, letting them know that I expected it to be much lower, and they put me in contact with someone. I explained the situation and he recommended that I move my stuff over to a new AWS account that had fresh credits applied, because I can’t apply $5000 in credits to an account that had already done that. I would have had to find a $10K credit and apply that, which would then leave me with $5K because of the way they do it.
So I did start down the path of getting that account setup, the group I joined for getting 100 users in 100 days offered a package of AWS credits (and Render credits, which was great for my backend!). I continued to rack up charges on AWS though. And I did get a notification that my account would be suspended, which I was nervous about, but then the date of the threatened suspension came and went and they didn’t do anything. At this point, I’m about $800 in the hole, but I’m planning on moving things over and then paying it down as I could. Since they didn’t shut it off, I didn’t push as hard as I could have to move things over. I assumed it was too small for them to worry about. Which was fine.
Until it wasn’t.
On Friday, I woke up to the notification that my AWS account was suspended and they had taken away access from everything until I pay off the account. Which meant that not only was Feather offline (which I didn’t care about because it’s not actually doing much at this point), but Who Should I Unfollow? was also offline, because I was using the same account to deploy certain pieces of it.
And since that’s the one product that’s actually producing revenue, that’s kinda not great…
Looking back, I should have done things a lot differently. This is entirely my fault.
And that’s the worst part.
I thought about how I could spin it. How I could explain the site being down?
And then nobody even noticed (or at least no one noticed and bothered to tell me…). Which was an incredibly painful relief.
I tried to deploy the app to a different subdomain and transfer users who logged in there instead, which wouldn’t allow existing paid users to get the right functionality, but at least it would be something. Unfortunately, that isn’t working because I haven’t been able to get my backend to allow requests from the new url. Still not sure why, but I’ll figure it out eventually. I think I can rebuild most of the frontend’s DB based on the data I’ve got in the backend DB, but it’s going to take some time and effort.
Before I finish doing that though, I wanted to get the presale up for Social Media Gardens. That way, I’ll hopefully be able to make a few presales and that would let me just catch up on my AWS bill and I could avoid that extra effort to get back online.
Making Progress: Social Media Gardens
I had stared getting the presale set up last weekend. Wasn’t quite able to get the page finished, but it’s close. Just need to get the live pricing set up on Stripe.
But I spent this weekend working on the prototype. I really didn’t want to officially launch the presale without having at least something to show.
I was able to get it deployed and I have to admit, I’m pretty proud of it. I think it does a great job of showing what I want it to be. It’s a single garden that features a highlight video, recently published content, and a place to collect positive comments. It’s not perfect yet, but I really like where I’m going with it.
And the key piece is actually an old technology, which gave me the inspiration: RSS feeds. The first version of RSS was actually released in 1999, which blows my mind a bit.
But as it turns out, almost every major publishing channel supports them by default: blogging platforms, YouTube channels, podcasts, you name it. They all have RSS support built in.
In a nutshell, RSS feeds can help you keep track of new content being published and are super easy to parse. That let me populate the prototype super easily. I used a Python package called feedparser to grab the data, do some small manipulations to get the data I cared about out of the pile, and pasted that into the prototype. That will make it pretty easy to update in the future to pull the data in dynamically, because I’ve got an idea of what the data looks like.
This should allow me to get a working version out in the next week or two that can automatically refresh the data, and I’ll start by building profile pages manually. But as I do so, I’ll be automating more and more of it, as I figure out what people like and what they don’t like. I’ll use this as an opportunity to provide more white-glove type service to people who buy the presale, getting to know them and what they need, as that will be helpful in building out the product side of it. For regular gardens, it will be free. But the paid tier will offer a bunch of extra tools that will help users manage their accounts and probably some other things. I’ve got a number of ideas related to it, but I’m in the process of prioritizing the work. More details coming soon on that.
The extra benefit is that as I start getting people onboarded here, it provides a way for me to populate Friend Content Bot with data. (Remember Friend Content Bot? He was my entry for the Twitter Developer Challenge. Still waiting to hear back on those results…)
So overall, it’s been a bit of a tough couple of weeks, but I think I’m well positioned enough to get myself out of the mess I got stuck in.
Just going to trust my process, do what I can, and get everything pulled back together a piece at a time.
Factory Production Stats
Let’s take a look at where things stand now. I’ve got Who Should I Unfollow? generating revenue but is currently down. Fixed a couple of bugs and had a couple more sales before the app went down. I’m also calling Friend Content bot a product that’s live as well, although it’s not doing much right now. Will be more useful when I start getting data for it to use. Social Media Garden prototype is now live and just need to get the presale page up. Will hopefully be generating revenue from that soon too.
Current Products Live: 2
Total Revenue: $134
Recurring Revenue: $29 MRR