Ask IH: How do you scaffold a SaaS product quickly?
Hello, fellow indie hackers,
I have developed various SaaS products in the last 5 years, some for my own and some for companies I worked for, and one of the most common issues has always been to get going from the technical perspective so that the effort can be focused on developing the actual product value rather than dealing with tech-related stuff that is not directly helping the product value, at least not directly in the early steps. Stuff like authentication, figuring out development environment, deployment solutions, settings pages; essentially same things over and over again for every other project, I find it painful.
In order to get faster I have developed my own stack over the years:
- Laravel for the main application
- Docker & Docker Compose for development
- GitLab for CI/CD
- Kubernetes for live deployments
Having this setup with standardized files helped getting started much faster, but I still feel like I am developing the same stuff over and over again, with more up-to-date packages and libraries maybe, but they are still the same stuff. For every project I need to:
- develop authentication pages including login, register, password emails
- develop a basic dashboard
- put together a settings page for name, email, password reset etc.
- build landing page
- build payments integrations, plans pages, etc.
And all these are side works that don't directly benefit the core value proposition. My question is, what is your solution for quickly wrapping up an MVP of a SaaS product that needs coding, e.g. a no-code solution might be beneficial for some stuff but let's assume that it will be a coded solution because of custom needs.
Thanks!
For rails we also have https://jumpstartrails.com
Creator of Jumpstart Rails here. 👋Thanks for sharing us @shrikar!
We've got both a free version and paid version depending on what you need starting out. It turns out its a TON of work to maintain a SaaS template, so I charge for one version and include free updates on it. I built an video subscription product in under 10 minutes using Jumpstart Pro which is pretty cool. You can find the video on the homepage of how I built it.
Happy to answer any questions!
Amazing project. Totally worth sharing :)
+1 for Jumpstart! I have successfully built apps with both the free and Pro versions!
I second this! Jumpstart Pro is great value for money, it comes with a bunch of features, like payments, pricing page, etc. It's not too opinionated and feels really lightweight. Somtimes I feel like other similar products out there can easily feel bloated with too many decisions made for you which is hard to break free from.
I'm working on something that helps React devs skip all the boilerplate with project and layout setup (divjoy.com). I learned pretty quickly that the users that want this the most (and are willing to pay) are people building SaaS so I'm quickly iterating and adding the kind of things they need like Stripe payment integration, user admin dashboard, etc.
That said, I actually don't have much experience building SaaS products so I'm learning as I go here. I'd love to hear from you and others in this thread if there are any other features that are nice to have in a SaaS that you'd rather not build yourself.
I was having a conversation with a friend of mine and he was suggesting me to offer my stack as a product as well but I was skeptical about that, now you motivated me!
Divjoy seems very interesting, especially with authentication and deployment, great touches. I am not mainly a JS guy, but I will definitely give it a go as a learning project as soon as I have the time. Cheers!
Nice! I think there's room for different "SaaS in a box" type products that cover different stacks. You have https://bullettrain.co for Rails. I'm focused on React and bringing together all the things you need (which is a lot given how unopinionated React is). A Laravel one might do quite well.
Ah, I didn't know Bullet Train, thanks for the suggestion! btw, I think the link is going to be https://bullettrain.co/, as the other one seems to be a phone case brand. :)
While you feel that you are doing the same thing over and over again, my observation is that, in reality, it is always just a little bit different than previously. Changes in the onboarding process, the business model how you do the subscription exactly, etc. So for me, boilerplates are not that helpful besides learning/stealing some patterns, even though I feel the same pain. Nevertheless, there was this post on IH: https://www.indiehackers.com/post/17-saas-starter-kits-boilerplates-based-on-your-favourite-programming-language-framework-35387161e0 with a collection of various boilerplates.
I wrote this ;)
Hope this helps. There are 21 in the list after a recent update.
Yeah, very good point, I have the same confusion as well, sometimes I even enjoy those little changes, but at the same time it feels like I am moving slower than I can.
Early days but I am fascinated with all that can be done with aws amplify, react, gatsby, etc. The only issue for me is amplify doesn't look like it is coming to mainland China anytime soon (a lot of otherwise ordinary services are delayed or banned usually so domestic players can copy everything first). Need to find another way. Any ideas out there?
After a long time brainstorming on these, I have launched a product to take care of some aspects of getting started with a new idea: https://saasstarterkit.app/
Some of these tools and technologies might be foreign to some people; however, throughout the years, I have iterated through various setups and came to the conclusion that this is the best setup for increased productivity and keeping the quality high during the development. For makers, time is money. I don't want to lose more time trying to integrate payments into my product, or build settings pages, or pick a deployment solution, let alone automating it. With Nana, all of these come pre-made.
All in all, Nana is a boilerplate that gives access to all these things. It is not meant to be an updated product; it is a buy-and-fork type of product where one uses it as the basis and starts doing business changes. Nana follows the standard Laravel project layout, therefore all the Laravel upgrades can be done just like any other Laravel project, but Nana will not send updates because it is not the goal: Nana gives the basis, you build your product on top of it.
I am currently exploring Early Access options with more than a 65% discount, and I'll be giving direct personal support to early adopters through email and video calls.
I have also created a coupon for IndieHackers community for an additional 10% discount during the checkout for the Early Access period, use the code indiehackersearlyaccess during checkout and the additional 10% discount will be applied automatically.
Feel free to hit me if you have any questions or feedback!
https://saasstarterkit.app
I have built one in NodeJS and ReactJS - https://github.com/saas-developer/saas-starter . I have made a course explaining it - https://saasify.teachable.com/p/build-a-saas-app-with-nodejs-and-reactjs
You should really check out Graphile Starter.
https://github.com/graphile/starter
It’s super easy to start up on heroku or docker
Not your stack, but this is what I've done for just your problem: https://github.com/rails-boilerplate-code/base-template/blob/master/README.md
@karakanb We got you! What you described is exactly what we offer at Outseta (https://www.outseta.com/saas-scaffolding). This post details all the ways we can help you get your product to market faster by focusing on building just your core product. https://www.outseta.com/posts/how-we-launched-a-new-saas-start-up-50-faster-using-outseta
I go with React and Typescript for the frontend. For the backend, I used to go with Flask and Python but I'm really interested in trying things that would let me going a bit faster on the deployment side, like for example AWS lambdas and Amplify.
I would love to hear feedbacks of people having used these to bootstrap applications, especially the user login management part!
I built my own boilerplate (essentially what you've described) and then commercialised it for other people to use: https://usegravity.app after discovering there is a good market for these products.
For a Laravel developer have you looked into https://spark.laravel.com ?
I haven't used it before, though I have considered multiple times but never had the chance, but I am pretty sure it has tons of improvements over the regular flow; however, what I am skeptical about Spark is, it has the functionality as features, but you still need to implement it in your app even though it is a simple implementation. I have implemented subscription payments using Cashier a couple of times, it is already very simple too, but all that simple work piling up makes the whole process more dreading.
Maybe I am looking for something that is not possible, I don't know.
99$ per site? I guess there are many free starter projects in github, just a quick check: http://laravel-boilerplate.com/6.0/start.html for example.
$99/per site is severely underpriced. $1,400 is underpriced.
To outsource the work in a SaaS boilerplate would cost well over $10,000 to build properly.
For someone who is strapped for cash, it makes economical sense to do the work themselves. However, for those who have the money but would rather spend 100s of hours rather than $99, or even $999 is very poor business logic.
The time should be spent on building features that customers will pay for, not infrastructure that paying customers don't give a hoot about :-)
"For someone who is strapped for cash, it makes economical sense to do the work themselves. However, for those who have the money but would rather spend 100s of hours rather than $99, or even $999 is very poor business logic." That's absolutely true. And I just learned it today, because I've just entered this site and discovered that there's a SaaS ecosystem which people make profit from. And once I've started reading, I've realized that time is also a very important factor to catch the market hunger before it feeds and of course paying it to gain hours seemed more acceptable because you have fewer things to worry when you buy it.
$99 to save you 100+ hours of work sounds very cheap.
That's true, only if you can afford to pay $99 which is guaranteed to bring you more :)
This comment has been voted down. Click to show.
Ouch, it hurt.. I have a VPS hosting which I pay 5$ a month (60$ per year) for all of my projects, one domain costs 10-12$ per year max for a com, and for advertisement, you can use the forums, IH, facebook, twitter, laravelnews, reddit etc. And once you create one or two things, you would have a boilerplate to use on the later projects.
But for a single site, 99$ seemed a bit expensive, really.
You think you can host a SaaS on a $5 server?
Interesting.
Keep us updated
What, in your opinion, makes it impossible to host an SaaS on 5$ server?
I know some who host on shared hosting.
Depends on the product I guess and how much you optimise it.
1GB ram and 1CPU seems very low. How do you back up your database etc?
With conversion you can scale? I'm talking as someone who doesn't own a SaaS product online. If you gonna change the way you talk here, I'll just stop. Have a nice day.
https://bullettrain.co/pricing charges $1450 for a ruby on rails SaaS template and has 100s of customers.
I guess what market you target matters a lot.
I built https://versoly.com/ for the landing page, it allows you to build a well designed site in minutes. Sites are optimised for conversions (speed and design)
It also comes with blogs when you need to start producing content.
We also plan on adding a lot of other features that will make life much easier for SaaS founders and marketers
I see a lot of developers try to build a landing page and waste so much time. If the product is successful they will need to swap to a CMS anyways.