Carl here. You signed up to receive updates about Ymir, the WordPress serverless DevOps platform that I’m building.
I wrote my previous report thinking this cycle would just feel more or less the same. I'd continue working on the product and trying to market it. It really didn't play out that way.
As I was doing work with customers helping with performance issues, I stumbled on a significant scaling problem. One so serious I thought I thought it might make Ymir unviable as a product. This, combined with the continuing growth struggles, really bummed me out during the whole cycle.
I spent marketing week working on the pricing changes I'd talked about in the last report. They went live at the end of the week. I'll get more into it in the business section.
You can always view the history of Ymir's product development at https://ymirapp.com/changelog.
While it might sound dramatic to say a scaling issue can make a product unviable, I really felt like a fraud. You can't sell the promise of horizontal scaling and not deliver on that promise. It also made me realize that clearly no hosting company was seriously scaling WordPress horizontally (despite their claims) because they would have hit this issue.
So because I felt things were so dire, I stopped all product work except some bug fixes and tried to find a solution. In situations like these, I get really really obsessive. I basically spent 9 days straight working non-stop trying to find a solution.
Towards the end of the week, I'd tried almost every solution I could think of. But as I narrowed things down, I became convinced there was only one way to solve this issue. I didn't know how I could pull it off. It was way beyond my area of expertise.
But then I remembered that Till Krüss was working on RelayCache and it actually did the thing I was trying to do. There might be hope!
Current status: Cursing the gods at Redis scaling issues. Praying @tillkruss's @RelayCache is the savior I need. pic.twitter.com/A2wSuoetz7
September 23rd 2021
If you go to the RelayCache site, it might not make sense why this would solve a scaling issue on Lambda. And that's because I uncovered an important use case for Till's product that he and his co-founder hadn't considered. So the marketing site doesn't really discuss it. I plan on writing a case study about it where I'll discuss the issue in less cryptic terms than I am right now lol.
You might have also noticed that I've been using the past tense. That's because after some trial and error, I got RelayCache to work on Lambda a day ago. Preliminary results are really really good.
RelayCache is still in development, so this is still very experimental. But I don't see why I wouldn't Ymir to use it in the long run. I also think this is a fantastic opportunity for Till to put his product through serious scaling scenarios. He's also super helpful and knowledgeable.
For marketing, the goal was to tweak Ymir's pricing. It's not a huge marketing exercise, but with everything else going on, that's all I had the energy to work on.
The first change was adding a 14-day free trial. I decided that, at this point, I just had to do it and see if it helped. I'm not super convinced it'll help convert, but I should try it at this point.
The second pricing change was adding a new agency/enterprise tier. I did a small thread on it. I'll discuss the gist of it in the business section.
I also added a chat bot on the marketing page. Not sure if it'll help, but I'm just adding standard marketing site things. I can always remove if it doesn't see any use.
You can always view Ymir's up-to-date business metrics at ymirapp.com/open. They're updated every 10 minutes.
September was another flat month. Gained some new customers. Lost some. It's not disheartening, but I'm in the long, slow, SaaS ramp of death and it shows.
Besides this scaling issue where I feel like a fraud, I still think Ymir is a genuine product that solves a real problem. It's pretty clear at this point that, even if Vapor has existed for over two years now, I'm still pretty early in the serverless trend. There isn't the same level of interest in the WordPress community as in the Laravel community. But there is interest and some of the customers are really excited about it and putting Ymir to the test.
Last week was a pretty rough week for me and @YmirApp. Found a significant scaling issue with Redis. But I got this earlier in my emails earlier and it gives me hope I can get this to work. pic.twitter.com/QFGHbcyxBG
September 27th 2021
So my thought is "If I'm early, what can I do to survive this period?" I've accepted that Ymir probably won't pay my rent this year if I don't do a change. That's why I added a new agency/enterprise tier.
One common complaint I've had from friends and other people who understand the product is that I don't charge enough. But I also don't feel there's a way to tier Ymir's offering by removing features without compromising the product. WordPress developers work in different ways and no usage based pricing made sense. I also like the idea of leaving value on the table to create goodwill.
I also didn't want to charge more than $39. It's expensive for some with the exchange rate, but, since you also have to pay your AWS bill, it feels fair. AWS is already going to cost more than you're used to if you want a $5/month VPS.
So I settled on the idea of the value add that I'm doing for certain customers already. Things like doing consulting calls about scaling, performance and platform integration. These are things an agency or company want, but doesn't matter so much for people who do it themselves. Forthem, asking in the Ymir Discord server or discussion forum is good enough.
That's the idea with the agency/enterprise tier. Based on the feedback I got, it's an insane value at $999/month. If I get 1-2 customer, it'll give me some much needed financial security. Although it'll be stressful to rely on those 1-2 customers.
But let's wait and see if anyone signs up first before worrying about that lol