Things nobody tells you before building a SaaS 🧵
1/ There are two types of people "building a SaaS." People building from scratch, and people just copying or acquiring something existing. The advice is completely different. I'll cover both.
1/ There are two types of people "building a SaaS." People building from scratch, and people just copying or acquiring something existing. The advice is completely different. I'll cover both.
2/ If you're building from scratch, ask yourself this: are you entering a validated market or an unvalidated one?
Unvalidated feels exciting. It's usually a trap. You'll spend 18 months building something nobody searched for, then wonder why CAC(Customer Acquisition Cost)..
Unvalidated feels exciting. It's usually a trap. You'll spend 18 months building something nobody searched for, then wonder why CAC(Customer Acquisition Cost)..
is destroying you.
Validated markets have competition. That's the point.
Validated markets have competition. That's the point.
3/ Hidden infrastructure costs will hit you like a tax bill you didn't see coming.
Stripe takes a cut. AWS/Digital bills spike. Postmark, Twilio, Pusher, S3; they all add up. I've seen founders underestimate monthly infra by 4x. Run the numbers before you price your tiers.
Stripe takes a cut. AWS/Digital bills spike. Postmark, Twilio, Pusher, S3; they all add up. I've seen founders underestimate monthly infra by 4x. Run the numbers before you price your tiers.
4/ Some platforms also have what I'd call "developer MUDs", third-party SDKs, API wrappers, or vendor lock-in you didn't fully account for.
Suddenly you're paying interest on a decision you made in week 2 for the life of the product. Read the pricing pages carefully. All of them
Suddenly you're paying interest on a decision you made in week 2 for the life of the product. Read the pricing pages carefully. All of them
5/ If you are building on an existing codebase, keep money in reserve for what's underneath.
Legacy tech debt is Murphy's Law. Something will break after you launch. The previous owner's "temporary fix" from 2021 is now your permanent problem.
Legacy tech debt is Murphy's Law. Something will break after you launch. The previous owner's "temporary fix" from 2021 is now your permanent problem.
6/ Utility costs catch first-timers off guard too.
Moving from a side project with 10 users to a real product with 500 is not a linear cost increase. Your email volume, storage, bandwidth, and support load can jump 5x while revenue is still finding its feet.
Moving from a side project with 10 users to a real product with 500 is not a linear cost increase. Your email volume, storage, bandwidth, and support load can jump 5x while revenue is still finding its feet.
7/ New build vs. acquired product is honestly the same tradeoff as new car vs. used car.
New build: higher upfront cost, slower start, but you own the architecture. No skeletons.
Acquired: cheaper to start, faster to revenue, but you will pay for someone else's decisions.
New build: higher upfront cost, slower start, but you own the architecture. No skeletons.
Acquired: cheaper to start, faster to revenue, but you will pay for someone else's decisions.
8/ One more thing on building new, if you miss a deadline with an integration partner, go outside their preferred payment processor, or fall out of API compliance, the penalties are real.
Platform risk is a thing. Build with one eye on your dependency list.
Platform risk is a thing. Build with one eye on your dependency list.
9/ None of this is meant to scare you. It's just stuff I wish someone had laid out clearly before I started.
The founders who survive are the ones who knew what they were walking into.
The founders who survive are the ones who knew what they were walking into.
@threadreaderapp unroll
• • •
Missing some Tweet in this thread? You can try to
force a refresh
