The $1 start-up

Dear tech founder,

How goes it? Oh really? That well! Congrats on your billion-dollar idea. The next step is to get it in front of your users.

Before you spin up a Kubernetes cluster, there’s something I want to tell you. You can do this all for free until you get some paying customers. If you dare spend more than $1 per month, I will look for you. I will find you. I will glare at you. In this day and age there is absolutely no reason for a brand new startup to spend more than $1 per month until it has found product market fit. After you read this, your response to tools without a free plan will be simple and straightforward:

Ready to build a $1 startup? Awesome. Let’s grab our tools and get building.

Getting your tech online

Obviously, customers need a way to access your billion-dollar idea. And this is where your single cost will lie. You can grab a domain name for like R99 per year which is like $0.5 per month. Perhaps you want a more psychedelic name? Fine, go for it. It’s your idea. Risk it if you want. Just make sure you stay below $1 per month.

Hosting: Firebase

Firebase hosting is rock solid and most importantly, free. Obviously, you need to do the DNS thing and set up records and all that admin, then you’re golden. Now here’s the part you probably don’t want to hear. Give the single-page app a skip, in these early days you need all the SEO you can get. Making a SPA SEO friendly requires a PhD in “ripping your eyes out and wondering why you were born”, so just do a static path. If you insist on a SPA host on app.yourdomain.co.za then you can have your static marketing site on yourdomain.co.za . After doing this, set up Search Console so you can see how your site is performing in Google searches. Before you ask about Bing, please get serious.

Source control: GitHub

Go with GitHub. It’s straightforward and it works. With GitHub you also get CI/CD for free and once your build passes, you can deploy your code to Firebase . That way you can get your bugs straight from your machine all the way through to your customers without breaking a sweat.

I know writing tests is boring and that your code is perfect. Humour me here and just build two tests. One for signup and one for collecting payment. Other than that you can play fast and loose with your code.

Planning: Linear

As you develop your product and customers find the bugs you’re shipping to them, you’re going to need to organise your life. Linear is the way to go. They claim to be “a better way to build products”. It’s a bold claim. And a tad sensational too. But they’re a good choice. Let’s see if they die the hero or live long enough to become the Atlassian.

Understand your customers

Once you're live, you’ll probably want to learn a bit more about how your users are using your product. Sorry, sorry, what I mean is, since your idea is perfect you'll want to prove that your users are using your product exactly as you predicted…right? That’s where web analytics and product analytics come in. There are many products in this space. But let’s be real, if you’d already found a good one you’d have skipped this part of the article, so here you are…

Web Analytics: Microanalytics

The main aim of web analytics is to see how your users get to your site. Look, I’m sure Google Analytics is amazing and GA4 is a modern marvel. But the truth is nobody has time for all that complexity. Simple is elegant. Simple is good. And GA4 is not simple. Microanalytics , on the other hand, is small, fast, and simple to use.

You’re still loyal to Google? My friend, until your startup is printing money the only thing you’re loyal to is the cult of not being unserious.

Product Analytics: June

After seeing Microanalytics you're probably like: “But Lenny-san said I should track exactly how my users are using my product.” Firstly, we all listen to Lenny’s Podcast so you don’t need to flex. Secondly, that’s what June is for.

June is fantastic; to be honest I’m surprised people use anything else. It automatically tracks the key product features on top of generating reports for feature adoption. Look, I grew up on Mixpanel and Keen, and I’m never going back. June is the promised land and lemme tell you Lenny-san would be proud of you for using June; let’s be honest, he only plugs Atlassian because they pay him.

Customer Support: Crisp

I’m assuming—and it’s a big assumption—that you can’t walk over and talk to each of your users. If you can, this is not for you. While Discord, Slack, and WhatsApp are great for building communities, the folks who are not yet sold on your product are unlikely to join a community. This is where crisp comes in. You can live chat with your users. It’s simple. It’s elegant. And most importantly it’s… yip: free.

Crisp is a great place for providing customer support. And you can download the app to provide support wherever you are. Your interactions here are very often the thing that will sell your customers on your billion-dollar idea. When they’re sold, and only then, can you add them on Slack and send them a link to your SoundCloud mix tape.

Email Automation: Loops

Once you’ve convinced the lambs to sign up for your product you’ll probably want to send them emails. Email is a nightmare: A records, B records, C records, deliverability, DNS settings, domain verification…you can walk this nightmare path or you could just use Loops . It’s simple, easy to use, renders on basically every mail client (if you know you know) and most importantly… it’s free. You can send marketing mails as well as transactional mails. And the founders are pretty helpful when you get stuck.

The other pretty nifty feature—which I guess is their flagship feature—is Loops. You can create event driven sequences of emails. If a user signs up do this, then after two weeks do that. It saves a ton of time and means you don't have to manage event queues.

Back-end infrastructure

This is probably the hardest part. A free backend is hard to wrangle, especially since Heroku clamped down on being "indie dev" first, which I’m pretty sure is completely unrelated to them being acquired by Salesforce.

The least dramatic approach is to create a virtual credit card with a $1 dollar limit. Then upgrade to the pay-as-you-go plan on Firebase. This will give you access to 2 million cloud function invocations per month. And let’s be honest if you have 2 million invocations and you can’t monetise your business you either need to read “Designing Data-Intensive Applications” or go back to focusing on your 9-to-5. It goes without saying that you should use Firebase for your database (Firestore or the amazing Realtime one) as well as Firebase Auth. They’ll have you covered for quite a while.

(To the Parse fan bois out there: looks like you backed the wrong horse, #justsaying)

Operating your biz

There is one inconvenient thing about a tech startup that you can’t really get away from: it is, in the final analysis, a business. That means documents and finances are part of the package. I know, it’s not ideal. But you only have yourself to blame, you could have just carried on with your 9-to-5.

Documents: Google Docs

At this point, you have a Gmail account and you’re running on Firebase, in for a penny, in for a pound. Just use Google Docs .

Payments: Paystack

Paystack is no frills payments. It just works. No drama. No backflips. No “contact sales”. Simple pricing. Simple API. Just sign up, add your compliance docs, and you’re good to go. Look their documentation isn’t perfect, eventually, Papa Stripe will come help them sort that out. What about all the other payment providers you ask? Okay, go look. I’ll wait.

Oh, you’re back. That’s what I thought. Let’s wrap this up.

Financial management: stub

Look at this point your business barely has a penny to its name. You gotta keep your costs down. What better way to do that than to use stub ? Obviously, I’m a little biased, but since you accepted my other 9 recommendations, why stop now? Also, stub is free.

But whatever you do, don’t put off doing your finances properly. The stat is something like 9 out of 10* billion dollar ideas fail because of poor financial management. I mean just look at FTX.

Launch and grow

You see you don’t need to spend lots of cash to launch your tech startup. The time for paying will come. Your job for the next while is to find product market fit. So as you chart ahead in your startup adventure, if there’s one thing you remember, let it be this:

“If there’s no free plan, it’s a scam.”

I mean, even life has a free plan. Good luck friend. And may your build always pass.

Regards

Alex

P.S. Since you’re going to do a multi-cloud kubernetes deployment anyway, why don’t you bookmark this page for when that first bill hits you. See you later. And sorry you had to learn the hard way.

P.P.S. Can't be bothered to read all this? Here’s the summary:

1. Hosting and tech stack: Firebase
2. Source Control: GitHub
3. Planning: Linear
4. Web Analytics: Microanalytics
5. Product Analytics: June
6. Email Automation: Loops
7. Chat Support: Crisp
8. Docs: Google Docs
9. Payments: Paystack
10. Finances: stub

* This stat is definitely made up. But what is true is that poor cash flow management is the number one killer of small businesses.

Original article on stub