Building a new SaaS Business

[Sadik]

Ok, Here after a while. I want to use this thread as a log + bounce ideas.

Brief Intro

Am a freelance web dev. Overall been at this for over a decade. Have built maybe over a hundred client sites (haven't kept count! ). We are three co-founders. Each has a different area of expertise which complements for this new venture. I am the developer, Partner 2 (Let's call him P2) is a marketer plus product designer. P3 is a Finance guy with experience in buying and selling businesses. We are all in our mid 30s.

This first post is to essentially think out my SaaS infra plans. I am building a SaaS for the first time. Maybe you could call a few of my client sites as SaaS but this will my first own SaaS venture. I am hoping someone with more know how than me will add some interesting posts to this thread.

My experience is as a PHP web developer. So I am going with my strength and starting with a LAMP stack.

Task 1 is to select where I will host. The goal is that the host I select and infrastructure I design should work till around 10,000 users of the system. Yeah haven't written a single line of code and not a single user but one has to plan right!

A word on the SaaS is that this won't be a real time service kind of thing but more of a "back-end service" for a specific business type. So I don't really need high end real time processing. What I am looking to get out of the host is

1> Web Application server - Will probably be running NGINX, again due to existing expertise, though I may use OpenLitespeed. Not fully decided yet. Language is PHP.
2> Load Balancer - I have never implemented one but I will check them out. May just go with using separate subdomains for different things and not bother. Any recommendations?
3> Database server - MariaDB of course. Going to keep the Web server and DB server instances separate. Also am going to do a Master - Slave replication with writes on master and reads from Slave. Again this is the end goal and I might start just with a master server and implement slave replication after launch and when we reach a certain number of users.
4> CDN - Cloudflare's free CDN should work for now.
5> Backups and failsafe.
6> Long term data storage - The service I am building has payments data for our users which will need to be stored forever.

Places I have in mind are

AWS - I have used it for multiple clients but have never liked it. It's always a pain to get anything done in AWS. Everyone though mentions them as a solution which scales. I get it but I don't like them. Will probably not choose them

Google Cloud - I like them a little more than AWS or maybe I should say I dislike them less... Maybe there is a better option.

Digital Ocean - I used to manage a huge client social network on DO before the owner sold the business. They are easier to work with. Leaning towards them at the moment.

Linode - Maybe...

Anything that should be in my radar?

Since we are starting, we don't have a lot of money. And goal is to start to code as fast as possible with a infra design which will scale till at least 10K users without very high costs. Ideally less than $100 / month.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Kid]

Seems good.
Let's see how it will go.
Is there something more you can say about it
without disclosing important details of a business?

 

[TheSilverSpoon]

Good to see another PHP guy on here! I had a similar path, started freelance and now have a small portfolio of SaaS products.

My $0.02

- as long as you aren't doing anything dumb (N+1 queries and the like), scaling is usually non-issue. Focus on features and ship, ship, ship! As long as you have a basic CRUD-ish type of app, a basic LAMP / LEMP stack should get you to a few thousand users easy without breaking the bank. We have some services that are < $100 / mo in server costs and bring in silly amounts of revenue.
- find a framework you like and roll with it. It will save you and your team hours and hours of dev time and increase your feature velocity, which is priority #1. I <3 Laravel. The ORM alone saves our team a ton of time.
- get a good CI / CD pipeline in place early on. This will pay dividends. If you are on Laravel, Envoyer - Zero Downtime PHP Deployment is the defacto standard
- I'm a big fan of Digital Ocean. Cheap, predictable pricing. And usually better performance at the same specs compared to AWS / GCP. AWS always seemed like a clusterfuck of unnecessary stuff with very opaque "gotcha" pricing.
- for backups, if you are on Laravel - spatie/laravel-backup. You'll be up and running in under an hour with backups persisted to S3.

Best of luck!

 

[Sadik]

Good to see another PHP guy on here! I had a similar path, started freelance and now have a small portfolio of SaaS products.

My $0.02

- as long as you aren't doing anything dumb (N+1 queries and the like), scaling is usually non-issue. Focus on features and ship, ship, ship! As long as you have a basic CRUD-ish type of app, a basic LAMP / LEMP stack should get you to a few thousand users easy without breaking the bank. We have some services that are < $100 / mo in server costs and bring in silly amounts of revenue.
- find a framework you like and roll with it. It will save you and your team hours and hours of dev time and increase your feature velocity, which is priority #1. I <3 Laravel. The ORM alone saves our team a ton of time.
- get a good CI / CD pipeline in place early on. This will pay dividends. If you are on Laravel, Envoyer - Zero Downtime PHP Deployment is the defacto standard
- I'm a big fan of Digital Ocean. Cheap, predictable pricing. And usually better performance at the same specs compared to AWS / GCP. AWS always seemed like a clusterfuck of unnecessary stuff with very opaque "gotcha" pricing.
- for backups, if you are on Laravel - spatie/laravel-backup. You'll be up and running in under an hour with backups persisted to S3.

Best of luck!

Thanks. This is terrific advice. I will check out each of the links and give them serious thought. I intentionally didn't post much about the application in the first post. I will talk about which framework and why in a next post.

Seems good.
Let's see how it will go.
Is there something more you can say about it
without disclosing important details of a business?

I can't go into much detail for obvious reasons. It's closely related to payments online and there is a specific need which we feel we can build a product for. We did a quick validation exercise amongst our network and through some reddit ads which convinced us to move ahead.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WestCoast]

I don't know what any of this means... but, sounds pretty cool.

I wish I knew what any of this meant though

 

[Deleted50669]

Node guy here, I feel you on ducking AWS. AWS has the cloud industry by the balls, but it's only because corporations are risk-averse and AWS presents the lowest risk profile (unless you factor in their Kinesis streaming service went down just before black friday...). I've been working on a few CRUDish apps over the past two years, would be interesting to share ideas.

 

[Kid]

It's closely related to payments online and there is a specific need which we feel we can build a product for.

Ok, so payments, good niche.
Also glad to read that you've (semi)validated it first.
Many people go to code without validation and get burned.
Fingers crossed.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Sadik]

Ok update.

We have our team call on Tuesdays. Using Notion for keeping track of everything. About my decision to go with DO vs a more "managed" server, P2 asked a relevant question. "you ok waking up at 3am to fix a server?". In all my years working on many client sites, a good host going down has been very rare. And when it happens, it's usually been a developer or admin's fault vs the system misbehaving. It's a valid point though.

On application shoutout to @TheSilverSpoon. I was torn between Laravel and Phalcon. I love Phalcon due to it's speed and how easy it is for me to get started with it. BUT... I am not building for "me". Can't be emotional about it. So we are indeed going with Laravel due to availability of integrations and the community. And IF this thing takes off, I will be hiring. Easier finding other developers with laravel experience. I don't have a lot of experience with laravel but I can learn...

P3 is in charge of admin stuff. We need a service provider for a very specific user activity which is core to our product. Can't share what it is, as it will give away the main business idea. I did some research and we have a few leads. P3 is calling / contacting them. We need this finalized.

We are now brainstorming a product name. Interesting thing that happened I thought a name and googled it. And would you believe I found another company in the process of building a 90% similar product with the exact same name! I requested early access on their landing page.

Also earlier this week, I saw that the biggest player in this industry is launching a similar service for it's user base. I am actually excited from these finds. If others are building and launching a similar / same service, including big players, it undoubtedly means demand!

We are moving slowly as all three of us have other work / businesses which are paying the bills. Keeping it slow but steady.

 

[Sadik]

Things are moving a little slowly. Haven't started any coding yet. I have a HUGE freelancing project ongoing to go live in the next 2 weeks. It's eating away all my time. I have refused new projects but I have to see this one through.

I am grateful for having partners. Both Partner 2 and Partner 3 have slowly moved things forward. P2 is working out use cases and workflows. We have a new Company Name, registered a domain and and P3 is in the process of incorporating a new company. We need a legal entity before we can enter into a contract with a service provider. Ideally I would have liked to do this later.

Target is to start coding before the end of this month. Have to set realistic targets to prevent guilt trap.