SaaS is hard. Few people are making money with it - even those who do usually don't get past the $10k MRR. Before the AI era, it was harder to build SaaS but easier to make money. Despite everything, I managed to take my SaaS Postiz to almost $50k MRR in one of the most competitive markets - thousands of competitors.
This is not a guru post that if you copy me, you can do the same (maybe you can, maybe you won't). You need to have a lot of resilience. What I did made me overwork, get fat, neglect everything in my life, and get burned out multiple times.
If you are ready for all those things, read this post (save it even); if not, maybe this post is not for you.
For the skeptics,here is my proof: https://profile.stripe.com/wickedguro/GNeymWms
Be different
When I say "be different," it doesn't mean build something no one else has ever built. Actually, that is usually a bad thing. If you build something different, people are most likely not looking for it, and it requires a lot of marketing education - something that is usually more familiar with VC-type companies.
The simplest example is the article's cover picture. You see a lot of X articles. Does it look different than most of the articles you see every day? probably. I can't say I am a "professional" marketer, but I am at least trying.
So we will get into the cover thing later, but what does it actually mean for you?
Think about this - if you can build something quickly with AI, everybody can. What are the small things that you can be different in?
I will give an example from my own experience, but you can try your own thing.
1. Designed by designers
If you have built a marketing website with AI, even if it looks good, it's most likely going to look like everybody else. Just another sloppy AI website that looks like another SaaS. So many people reached out to me and asked. How did you make the animation on your website? I always answer, "designers made it" - same thing for my dashboard.
2. Open Source
Indie hackers like us are building mid-tier SaaS, which usually means point-and-click startups with a few background jobs.
We basically wrap LLMs and add CRUD in different variations. We don't build deep tech or crazy hosting/payment infrastructure. We build things that are easy to replicate. It's just a matter of time and tokens. Being open-source gives you a bit of transparency: people can see your last commit(is it an active company), contribute code, and be part of your community.
Believe me, if somebody clones your SaaS, they won't beat you; it's so rare. In the two years I ran Postiz, nobody ever overthrew me with my own SaaS. Can you imagine somebody cloning OpenClaw and building an alternative that everybody will migrate to?
3. Find your blueocean
SaaS is stuffed, and if you copy-paste an existing SaaS, you won't succeed. I see so many social media schedulers out there. Almost none of them are passing the $1k-$3k MRR.
They suffer from high churn and money that goes to their competitors (me). Blue ocean doesn't mean you have different features or capabilities than your competitors. It means you are turning to a different market.
For me, when I started Postiz, I aimed for open-source, and later I aimed for people who are running agents. I disconnected myself from the competition.
Choose the right SaaS to build
As I said before, you don't need to be different, like "invent something new," you should actually pick something that can work. I would scout @TrustMRR to find repeated SaaS that are making money. And then I would think if there is an audience I can serve, they don't.
Just as an example, we are building now, agent-media It's an alternative to Arcades, makeugc, etc. What we are trying to do is make it fit AI agents. So instead of generating from the platform or using the API, we are giving people a special CLI so they can implement it with OpenClaw, Claude Code, Claude Cowork, etc.
Finding new markets is not easy. Last year, I saw a lot of hype around n8n, so I went ahead and created a node for it. Then I saw that MCP was booming, so again I tried to create an MCP as soon as possible and launch it on Product Hunt.
Now, for example, I also created my own OAuth for Postiz, so if you are building your own SaaS, you can invite Postiz users to connect their accounts to schedule posts for them.
I am also building an Enterprise mode for Postiz, so people who need social media scheduling don't have to implement everything from 0; they can use my approved apps and infrastructure and add it to their own SaaS.
I hope you can see what I am trying to do here. I FIND UNDERSERVED MARKETS.
Avoid making too many startups
I know there is a notion from Indiehackers to build as many startups as possible. I think it's wrong, and I always thought it was wrong. sure you can build many if none of them make money, but if they start to make money, invest everything you can in the ones that do. Here is why:
- We are entering the world of "brands", and direct marketing is not as strong as it was before. After being in the market for a while, people start to talk to you and tell others. Every time you start something new, you have zero brand.
- You don't have any market assets, and you are not mentioned anywhere. People can't even run a Google search on your brand because it just doesn't exist.
- If you really decide to build multiple brands, make sure one brand can convert to another brand. For example, I built Postiz, but now I am trying to send Postiz users to agent-media. Because I think those are two tools that can work wonderfully together, and Postiz can give a lot of credibility to the new tool on the block.
Market is smartly
There was a notion that you need to be super active on social media, write every day, "shit post". And it was true, that's how people who selected social media as a channel grew in the past.
And then came TikTok and said, listen, you don't need to be an old account to get views. You need to create good videos. It's even better if you are a new account; the algorithm might like you more.
And then the same thing happened on X: people with a huge mixed audience started to get a lot fewer views than people with 200-10k followers.
Just a week ago, I posted an article when I had 200 followers that got 500k views, another article that got 50k views:
Don't do what "everybody is doing",find where you can target your audience and can get a lot of traction. when I started out, open-source and Reddit / Lemmy was just insane:
And I got this amount of views like twice a month, because I released every new version on Reddit.
And right now if you haven't gotten it yet, TikTok and X are the easiest one to get tons of views.
You can see Oliver's article about Larry on getting a millions of views:
I also tried TikTok a bit:
Also X posts about money, work pretty well. Anyway, you get my point, find a channel that work a push on it like crazy. You don't have to "build in public" or "publish youtube videos". Just find what works.
Make people click
Sounds a bit stupid, but you always need to thing, how can you make people stop their doom scrolling, and just look at what you have done. Usually when people scroll, they don't see your content, they see a gist of it. You need to make this gist so good that they would have to click.
Of course the content of what you do must be good too, if people don't like it they'll churn and the algorithem won't like you.
So what do you think about the cover picture / title of this article? Remind you of something?
Ride hypes
I am a big hype person. Every time there is a new hype, I implement it in Postiz (if I can do a big marketing event around it)
When n8n was booming, guess who created a"n8n-nodes-postiz",when MCP was flying around -I already had a product hunt launch ready
VEO3 making awesome videos? Nevo was there.
I know that those hypes might die, so I don't build my SaaS as a hype, like I would not build an OpenClaw wrapper it's a bit risky.
But I would do things that can go really well with OpenClaw.
Stop making features, build a kick-ass product
I won't lie to you, Postiz first version was horrible and with SaaS being hard as it is, it might not have made it if I did it the same today.
Over my time in Postiz I literally stopped making new features. Instead, I refactor things in Postiz a million times.
I had tons of backend bugs and problems with the UI and UX. I registered to all my big competitors, buffer, metricool, later, etc. Checked what was good and what was bad, and implemented it. with designers!
Actually, when you create new features, you create new bugs and you get more churn (people leaving you).
Just an example:
Retention is the key
Most of the time inside of Postiz I lived on around 20%-30% churn, like literally losing tons of revenue every month. The key in the end to solve it was to switch the audience. It doesn't matter how many features I would add to Postiz, if a people stops to scheduling posts, they'll churn.
Once I understood I can serve people that are automating their posts / using agents. I drastically reduced my churn. because now the work was done by machines and not by people.
They go to sleep, and make money, simple.
If 100% of the people that are using Postiz stay with me, I would be a millionaire today.
I hope you like this one, if you do, follow me @wickedguro
