First days after landing page launched
Last week I launched a landing page for Docker Deploy (more info here).
I did little to promote the page, basically:
- Answered an X thread
- Answered two reddit posts
- Answered a dev.to post
The X post appeared in my timeline after I tweeted something with the #buildinublic hashtag, so the algorithm seems to work. For reddit and dev.to I just made a couple queries in google as my potential customers would, and commented the posts promoting Docker Deploy.
I got around 45 visitors, most of them (~35) after I commented or reddit.
Making people click
The first days, with only the X reply and 9 visitors, I got no clicks in my Join Waitinglist button. These visitors might have bounced because a variety of reasons, but my feeling is that it might have something to do with the Join Waitinglist button itself. If you need to deploy something now, you want a solution, not a promise. And even if you just want to play around, I guess joining a random waiting list is not very attractive.
To get a better results, I changed Join Waitinglist to Create Account.
Besides that, I realized that I kind of miscalculated the prices for the bigger plans, 1 to 4 CPUs, so I fixed that in the pricing table. The biggest drop was for the 4 CPU option, which went from US$ 21.60 to US$ 16.80. I don’t know how sensible my potential customers are to such a difference, but in any case I don’t want to lose leads for that difference.
After posting on reddit I got around 35 visitors. Unfortunately, most of them appear in Plausible as “Direct” traffic (because they come from the app maybe?), but I attibute them to reddit because the X visitors appear wth the right source and it just spiked the day I commented on reddit.
I consider that a whole lot of traffic if it really came from a single source. For speakerhouse we could only achieve that kind of traffic in two days only by paying.
Point for reddit. I’ll try posting more there to check if I see similar results.
First interested person
This week I had the first person creating an account. Even tough he is very interested in the service, he is not willing to try it until the app is more or less working, and he can use it without me doing manual work. If he would, I could take his docker files and deploy them by hand.
Again, in less that one week, I have an interested person, apparently willing to pay. For my previous project, we needed around two years from the idea to first paying customer.
Progress on the app
I launched the landing page without a real implementation. In the past days I put a some thought in how I’m going to implement the solution if it seems that there is a market for it and I will likely have to do some stuff that I never did before (like send commands to a remote machine programatically), but nothing too crazy.
First step was to make the Create account button functional, so I just created an account with Auth0 and set it up. I’ve used it before, so it didn’t took long.
After the first (and single) user registered, I started implementing the app. I don’t know how many people is willing to pay for this service until the app works automatically (without me intervening), but I think I can hack an MVP in some days. So I started doing that.
For the frontend I’m using the same Astro app as for the Landing Page. I tried to avoid using react and keep everything in plain JS, but I quickly realized I’m much much faster in React than in vanilla JS. This is because the declarative approach of react vs imperative of js. Nevertheless, I’m using Astro as wrapper and router, this way, each page is an astro page that contains a single screen in React, preventing the whole app to be bundled and sent together to the user. It’s quite fast actually.
For the backend I created a .NET project using EF Core and SQLite. Why this choice?
- It works
- I know the stack
This two points translate into speed, exactly what I need right now.
Of course I’m writing no tests and I’m not building any complex framework. I need to deliver to see if there is a market, that is the goal. And being focused on that helps me decide better. If I’m stuck with some silly problem that makes no difference for the users, I can carelessly hack an ugly solution that works. I can make it nice when I have paying customers.
Next steps
I’ll push the MVP until users can create and edit a Deployment in the app. From there, I’ll do more marketing and check if someone is willing to pay for the service with that implementation. This means, users only create the deployment and I have to do the rest of the work by hand.