Serverless Development Kicks ARSE: How to Build a Profitable SaaS Tool in Under 30 Days

In January I built and launched a serverless software startup using AI in 30 days

In the past, I’d wasted months building software products only to get a few free users and zero income

This time it was going to be different - 30 days to build and market a SaaS product and get paying users!

I’m going to show you exactly how I did it and the secret that made this project a success

Why build a SaaS Startup

For those of you who might not know, a SaaS product is just short for “software as a Service” and you probably use a few of them every day. From communication tools like slack, teams or zoom, to accounting software or CRMs like Salesforce or Zendesk.

I had wanted to build one for years. They're an amazing business where you can generate consistent monthly income as users subscribe to use your software.

But why 30 days?

To build a successful SaaS product you need to solve a problem for a customer. Finding the right idea can be tricky and I have a list of possible SaaS ideas. Any one of them I could spend 6 months working away at only to find out the idea isn't good enough.

Setting a limit of 30 days to get revenue meant 3 things:

• It forced me to get real user feedback as fast as I could
• If I couldn't get people to pay for something I built in 30 days, I could never scale the sales to the point where it would become a valuable business.
• If the idea wasn't good enough I could just move on and know I've not invested months

Finding the Idea

To build and market it in just 30 days I needed something that could provide real value without needing to be a massively complex system. I’d used ChatGPT a bit and knew that it could be incredibly powerful if used correctly.

One day I was browsing LinkedIn and saw a post saying that writing engaging and interesting comments is one of the best ways to grow on LinkedIn. They were suggesting spending 30 minutes a day just writing comments.

That’s when I had a lightbulb moment - Can I use ChatGPT to write a good comment for me?

Now we have an idea we enter the validation stage which has two parts:

1. Can I technically build this solution?

2. Are there users who would use it (and pay for it)?

Question 2 almost answered itself because I would pay for it. The original post was also from a LinkedIn Copywriter - someone who gets paid to write posts and comments. Plus in the comments, there were loads of people saying “I would love to but I can’t afford to spend 2.5 hours a week commenting”.

There are loads of other ways you can validate if an idea has customers, from just asking people who would the idea users, to creating a landing page and seeing if people join a waitlist

So now I needed to validate that the technology would work. The quickest way to do that was to copy the text from a LinkedIn post and paste it into chatGPT. I asked it to write an engaging comment and what came out was amazing.

How much you need to validate at this point depends a lot on your technical background. I knew I could create APIs and databases, but if you’re not a developer you could learn to build a little test app using a no-code platform like Webflow.

Building the Minumum Viable Product (MVP)

With basic validation done, I wanted to get real users testing this as soon as possible.

This is where you build your MVP - or Minimum Viable Product. As the name suggests this is the simplest version of your solution that still provides value to the user.

For me, that MVP was a chrome extension where you click a button on the post and the AI generates the comment and types it into the comment box for you.

Building and testing the chrome extension locally, as well as setting up the API took about 2 days. Now I just needed to add user login and usage monitoring and I was ready to get some real users.

Capturing Emails Early

But before I went any further, I did something that turned out to be massively beneficial later in the month. I recorded a quick demo of the extension generating a comment and posted the video on LinkedIn. I’ve got a decent following on LinkedIn and that post now has over 17,000 views and 52 comments which were a mix of support and hatred.

To capitalise on that I used a landing page builder to create a site that just had the demo video and a waitlist signup form. By the time I had all of the features for beta testing, I had 25 waitlist signups - another sign that there were people who wanted this product.

I sent out an email to all of the users on the waitlist telling them to sign in and start trying the tool.

Straight away I started getting really good feedback. This is why I wanted to get other people testing it as soon as possible.

From MVP to V1 Launch

At this point, I had two tasks

1. Handle feedback from the beta testers - some things are added to the ‘future features’ list, others you need to fix straight away, and some requests you decide you probably won’t ever support

2. Get the solution ready for a V1 launch

For me “Launch Ready” meant a few things

• A proper website with signup and account info
• Payment Handling with Stripe

I considered trying to build the website from scratch and then remembered that I was trying to do this as quickly as possible. Therefore I bought a React template, customised the styling and added all of my own copy. When you’re trying to launch in 30 days - good enough is the name of the game

The stripe integration I did have to build custom. Deploying and securing the webhook API and Database took a few minutes with Serverless, but getting all of the business logic right took a few days. The next time I do this, I’m going to find a website builder that has stripe payments integrated into it.

Launch Day

17 days after I started, I was ready to launch my full app. Over the last 2 weeks, I had continued to post progress updates on LinkedIn and this led to 69 waitlist users.

I sent out the Launch email and posted about the launch on LinkedIn.

And then I waited.

Would Anyone Pay for it?

Over the next 3 days, I continued to work on the app. I was fixing any bugs that popped up and adding new features like changing the tone of voice (from direct feedback from the beta testers).

I was also reaching out to LinkedIn Creators who I thought would be interested in the tool and might post about it. Anything to drive more users to the tool

I now had over 100 users but still no-one had paid anything. I was beginning to wonder if it was going to be another idea that people say they love, but wouldn’t actually pay for.

Then one night I was lying in bed and got a notification.

“LinkedIn Powertools - You’ve just received your first payment”

I was so excited - I was officially a SaaS Founder!

Growing and driving traffic

With this new wave of excitement, I spent the next 2 weeks writing LinkedIn articles, reaching out to more “LinkedIn Influencers” and adding even more features.

For me, one of the biggest feature releases was being able to customise the tone of voice on each comment - and even add a question or quote.

This was massive as I was finding the normal comments good - far better than other tools - but they didn’t continue the conversation or get a response from the original post creator. Just adding a quote or little question to the comment massively increased the number of times that the comment started a conversation.

End of the 30 Days

It’s now the end of the 30-day challenge and the tool has earnt a total of $140. We have 6 paying users with a total of 299 users signed up to the tool. When you think that this has only actually been live for 13 of those 30 days, that’s pretty insane!

Unfair Advantages

Now, what was the secret that I think made this project so successful?

I had two unfair advantages

1. I’ve been a Serverless Developer for years so the technical side is easier

2. Probably most importantly - I started with 14k followers on Linkedin

This meant that all of my posts got decent views. It also meant when reaching out to a “LinkedIn Influencer” they saw I was at least pretty good at using LinkedIn.

Finding your "Unfair Advantage"

Now you might say that you’re not a serverless developer and don’t even use LinkedIn, and that’s fine. You need to find out what “unfair advantages” you do have and make the most of them.

Are you prolific on Twitter or Reddit - use that as the platform to talk about your progress

Do you know a certain industry really well - Build your product in that area. You’ll know the right language to use and probably know a group of people who could become your first users.

It’s about finding the advantages you have and making the most of them.

What's Next?

But first, what am I going to do with LinkedIn Powertools?

I set the goal at the start of the month to generate revenue within 30 days. If I’d failed to do that I would have called it a great experiment but not worth continuing.

Since we got paying users and recurring revenue, I’m going to keep posting about it and looking for ways to get more people using it, but I’m going to slow down on the feature development. I’m also going to test some other marketing options to see what works.

If it continues to gain users and revenue then I’ll definitely be circling back to see how I can make this into the best tool possible.

If you’re interested in growing your LinkedIn or just want to try the tool, you can sign up here.

If you've liked this article, make sure to follow me on LinkedIn where I post daily about Serverless, AWS as well as my Startup Challenges.