Does this count network egress costs (for s3, lambda)? Those are hard to calculate but can add up. Also, you can uses sqs for queues, that has a pretty big free tier.

Finally, I think that bending the architecture of your app around what's free is not the best idea, especially if you are trying to get something out there to see if there is customer demand. (And doubly of you haven't built a cloud native app like is outlined here.)

In my mind the best way to build software for a startup is to build it using what you know as fast as you can. Avoid technical risk, because you have a boatload of business risk.

For me, that'd be using rails on heroku, which is still under $200/year for a fully functional dyno and database. For others it might be some varient of a mvc framework on a hosting provider. For others it might be WordPress (gasp!). For others it might be a cloud native app, as this post describes.

As long as you aren't spending extravagantly, time is more important than money when figuring out what your customers need.

This is true both in companies that have raised money and bootstrapped companies, for different reasons. For the first, you took money and need to figure out your product market for or scaling strategy ASAP. For the second, your time is super valuable because it is tied to your motivation. Doing work directly tied to customer value is a great motivator. (Doing other fun technical things that don't deliver customer value is a good way to learn things, but a bad way to run a bootstrapped business.)

I agree, for web apps with modest user requirements, Heroku is great. I am just running static web sites right now but if I start a new project I would probably start with Heroku. Their always on hobbyists dyno is $7/month and adequate.

If you ever grow your customer base it is not that difficult to move to directly using AWS, GCP, Azure, DO, etc.

In most cases it doesn't make a difference if you have costs of $6 or $60 annually, it's still marginal. I'd say speed it's way important - use the tools you know.

It's better to create a startup in 6 days than for $6/yr.

The vibe I'm getting in this thread is people are recommending services that are free for developers.. in which case, I recently discovered https://free-for.dev/ which is a mammoth list of such things.

Wow mammoth indeed ;)

Section I was most interested in "STUN, WebRTC, Web Socket Servers and Other Routers". I like the services like ngrok where you can just create a public ip for locally running web servers from the ide.

Some possible alternatives/improvements:

The AWS S3 free tier is 5GB for 12 months, with 20,000 GET requests and 2,000 PUT requests. [1]

Backblaze B2 (AWS S3 alternative) provides 10GB of free storage (no expiry date), and is part of the 'Bandwidth Alliance' [2] with CloudFlare. So aside from some pretty generous daily transaction fees (of which 2,500 of each type are free per day) [3], you could have your own free almost-cdn.

Porkbun currently have a 1st year (new domain registration) discount bringing the $6 .com domain cost down to $3,90. [4]

[1] https://aws.amazon.com/free/

[2] https://www.cloudflare.com/bandwidth-alliance/

[3] https://www.backblaze.com/b2/b2-transactions-price.html

[4] https://porkbun.com/tld/com

You can use Netlify DNS to simplify your stack. Cloudflare is not needed.

Also, Netlify provides Functions, which is Lambda abstraction. Faster to develop.

This is cool for simple ideas but really doesn't scale with complexity, not mentioning the 100% reliance on 3rd party services that could shut down at any time and bleeding potentially sensitive data at every step of the way...

I'd just use heroku and pay a bit more(or nothing if no users?), use plugins for what you've mentioned etc.

The cost of running is marginal and if I'm lucky enough to get a lot of users, I can easily switch to anything.

There is a lot of dev cost to wire up everything you've mentioned. Serverless is very devtime-wise consuming, it's state of art tooling is still slow as fuck if you compare writing an app using Rails/Django/etc.

IMHO dev time is what hurts more running a startup, this is what I would try to reduce

> 1. DynamoDB for database 2. AWS Lambda for backend 3. Netlify / Now / Surge for frontend

I'm evaluating FaunaDB with Zeit Now and so far it looks like a winning combo.

Zeit Now makes your dev much easier as all your application (front(s) + backend) can be in a single repo.

I don't have much experience with Dynamo but Fauna includes authentication + authorization out of the box and I think a more powerful query language. It also has first class support for GraphQL.

For marketing emails, Mailerlite has a generous and featureful free tier.

Zoho mail is damn good on the free tier.

Google forms / surveys.

GitHub of course!

Netlify is also a free CD / CI perhaps? You can run a script in a container on every deploy.

For seo, semrush has a good free tier for keyword research. There are other nice free tools like screamingfrog and keyword shitter.

Azure has some free forever tiers for functions and web apps but they are very basic.

Thanks for this reminder. My new mantra when currently developing an idea is "cheap MVP".

You will spend more while supporting such a stack. If you have enough experience just start with a production-ready stack, otherwise, youse cheap labor force (PHP, Python) to build initial prototype.

Do you own any startup that uses this setup? Could you please tell me?

Worth noting that the AWS services are typically free for only the first year. (though assuming you don't exceed those limits, would only be in the low dozens of dollars per year)

Lambda is free as long as you stay under the (I think monthly?) quota.

> IFTTT to webhook for cron

Could you explain this one a bit for me

I got the cheapest one, raspberry pi on my 100 mbps vdsl connection. Unlimited bandwidth/storage for less than 2.88 dollars per year.

Yes and then AWS free plan rolls off and you’re going to be paying quite a lot more!

Wait, how you can be free to use Cloudflare if you get 1M requests per month?

Are there time limits on these like free for a given time?

That's not a startup. That's a hobby.

It's not a business until you register with various levels of government and pay all of their fees, which instantly makes the $6 figure false.

awesome, but startups are not measured by requests per month but number of users.

i remember buying a domain on namecheap for 1$

As always, the devil is in the details. Was it a .com domain? I'm guessing $1 for the first year, is that correct? What is the cost from second year onwards?

Many companies use deceptive marketing selling stuff for such a low cost that it is too good to believe and it often is - you gotta read the fine print. I got my internet connection for $75 and two years later, I am paying 108$ for the exact same thing, with no additional benefit.

you can pick very cheap domains with some tlds.

For example, my .ovh is costing me around 4$ per year and there are probably even cheaper tlds.

what about sending and receiving email?

With Mailgun[0] you can send 5000 emails for 3 months for free.

The Amazon Simple Mail Service (SES)[1] lets you send 2000 emails per day with the Free Tier...

Sendgrid[2] gives you 40k emails for 30 days and then 100 per day for free forever...

[0] https://www.mailgun.com/

[1] https://aws.amazon.com/ses/

[2] https://sendgrid.com/

I still get 25K emails / month for free with SendGrid using Azure:

https://docs.microsoft.com/en-us/azure/sendgrid-dotnet-how-t...

+1 for sendgrid