How I've Grown My Headless Browser Business to $1,000/mo

Hello! What's your background, and what are you working on?

Hey friends! I'm Joel Griffith, founder of browserless. I'm a full-time software engineer, married father of two, failed jazz trumpet player, and a soloprenuer on probably my sixth project. I'd go so far as to say that I'm a serial failure, as the prior five or so projects I've attempted generated poor or no revenue, which I think is important for folks to hear since this is sometimes like winning the lottery.

browserless is simply a headleass-browser service, full stop. You've probably heard about headless Chrome if you frequent Hacker News — if not, it's a way to run a web browser without any sort of graphical interface. This has quite a number of usages including web scraping, testing websites, generating screenshots, and printing PDFs. browserless makes it trivial to do all of these things without spending a lot of time on getting Chrome set up properly. If you search around popular headless libraries you'll no doubt see all the issues people face when trying to do just that.

Since I launched roughly three months ago, it's been steadily growing, and as of right now I'm seeing about $800 a month of recurring revenue from about 15 paying customers.

What motivated you to get started with Browserless?

It might come as a shock, but I actually never set out to build browserless. I was originally working on a wishlist app where users could just paste in links to different things they wanted from online and then share them with their friends and family. I had a Node program that would load the user-submitted sites and scrape all the appropriate details — pretty simple. Then I tested this with a big name-brand site which immediately failed because they were running what's called a single-page app that had no content whatsoever in their intial HTML. This was a pretty big showstopper, and required me to run a web browser to load their page in order to scrape content.

At about the same time Chrome released a headless mode, which I then wrote an open-source library for in order to use on my wishlist project. This library, called navalia, got some good traction on GitHub and eventually caught the eyes of chromeless and then puppeteer.

If you see something out there that's trending and that you're good at, jump on it while you can.

As I became active on those projects I noticed a trend of developers having frustrating experiences trying to get them to work in production environments. It wasn't easy or quick to get headless Chrome running on Heroku or AWS, and when you were successful, things like fonts didn't look good or work at all. I began searching for any sort of headless-browser provider that I could both personally use and recommend to these users — but none existed.

All of these different events culminated when I was talking with Matthew Mueller, founder of StandUp Jack, when he just flat out said, "Why don't you just make a headless service that does this?" This was the "aha" moment that propelled me into breaking ground.

What went into building the initial product?

Since I had a pretty good understanding of what was going on under the hood with these projects, I knew exactly what to build, and I wanted to be finished in a few months or so, since I didn't take any sort of funding.

I also didn't want to pay for any outside help as I had literally $0 for this effort. This type of mindset resonantes with me well because it forces me to move quickly and not churn on trivial things like framework choices and tech stack. I remember constantly saying to myself, "It doesn't have to be perfect — it just has to work." I think this is really important for engineers, since we all seem to be indoctrinated with "crafting only beautiful, elegant code." No way, man. If it works, is secure, and generates revenue, then that's all that really matters.

From that I decided to not try and learn anything new technology-wise for this project. I used tools that I could move as fast as possible with, which for me meant JavaScript all-the-way-down since I'm primarily a front-end dev. Scaling and account management were the only things that really remained, and thankfully there are some great resources for those.

I used Stripe for payment processing, DigitalOcean for all the hardware needs, GitHub for version-control, and Docker for hosting images and doing builds. The biggest unknowns for me were payments and hardware. Stripe made the payment processing dead-simple, especially with their test environments and subscriptions. DigitalOcean also became a big component as it allowed me to manage new instances every time someone signs up or cancels, which meant I wouldn't be left holding the bag for unused hardware.

However, the biggest struggle by far for me was time. I have a wife, two young kids, and a full-time job as a software engineer; so time was extremely scarce. For this you have to be incredibly scrappy and leverage everything you can: working after everyone's asleep, coding when you're on a flight, and organizing on your phone when you can't actually program. Even when I was commuting to work I'd listen to Indie Hackers interviews and try and absorb as much as possible when I couldn't actually program. I think I'm on my third listen through all the prior episodes, and I'm still picking stuff up.

I can recall several instances where I woke up on my laptop after having fallen asleep from working on browserless. This obviously isn't sustainable, but sometimes you have push with everything you've got.

How have you attracted users and grown Browserless?

The first batch of initial users were developers I had worked with on puppeteer, chromeless, and navalia. Having the first few customers already lined up helped me validate that the idea was sound and motivated me to launch. Once I launched I posted on most of the usual culprits: Hacker News, reddit, and on a few GitHub issues. I immediately found out that because the audience for this service was small, larger sites just didn't seem to care much. I made no front pages and didn't get featured anywhere. What did work was answering people's questions on StackOverflow and Github — even if it didn't mean a conversion right away, it started creating some backlinks into the site, which at least helps with SEO.

After about a few weeks I started seeing more and more users from those libraries I mentioned buying subscriptions. As a matter of fact the first customer bought the most expensive plan I had, so that kept me motivated to ship. The following days after I started getting inquiries from CTOs and developers, some of which bought a subscription right away, others I just answered questions for. I tried doing ads for a time, which did result in more traffic but was expensive compared to the value being generated. Since then, I've done almost no advertising whatsoever.

Once I had some "wind in my sails" with a few customers, I started to write more frequently about how to do certain tasks with headless Chrome on Medium. Occassionaly I'd showcase how browserless makes it easier to do these tasks, which resulted in a few more conversions and inquiries. I don't think this process is necessarily scalable in the long term, however my feeling is that word-of-mouth will spread a bit and headless Chrome will begin to gain more traction over older technologies.

There are a few featuers coming up that I think will help new users get started a lot more quickly and have better developer ergonomics, which I believe will help spread the word. The benefit of this type of business is that there's almost no competition out there, so I haven't seen a lot churn in users. However, the downside is that browserless isn't something you'll generally know of or need until you find the problem it solves. Because of this, I'm really trying to focus on a highly polished product and documentation so that it continues to stand out.

One thing I'd like to stress to others is that if the niche for your product is small, don't expect to get a lot of big attention, at least initially. Even now browserless only sees roughly 50 users a day, which is nothing, quite frankly. I think what's important is that it solves a problem these people are facing, so those 50 users are already far into the conversion funnel. Because of this, you don't need 100k users a day to make your product successful (though it does help), you just need to know the problem well and where to find the frustrated users.

What's your business model, and how have you grown your revenue?

I pretty much stole the model used by projects like Ghost and Sidekiq when it comes to monetization. Essentially, browserless makes money in one of two ways:

1. Selling a corporate licenses so users can use their own infrastructure.
2. Selling hosted plans in various tiers.

Hosted plans account for the majority of my income from browserless. My hypothesis is that most developers don't really want to spend large amounts of time pulling images and getting infrastructure ready for this type of work. They've run into a problem and just need something that works quickly, as it's not their core business or a problem they're solving. With hosted plans it takes around five minutes to get everything running, which is entirely automated, and then they're good to go.

Hosted plans obviously have a smaller margin, since browserless needs to lease infrastructure. Whenever a new user signs up I use the DigitalOcean API to launch a droplet and provision it with all the software required. This ensures there's enough scale to meet demand, and it helps me better predict costs and long-term profit. Because of this strategy I probably over-lease hardware, since not every user will utilize 100% of their account, but it ensures that there's availability and helps keep sessions isolated, which is always a concern.

Initially I struggled with charging users, as I'm sure many people here can understand. Because of this inhibition I ran a closed beta that ran for nearly three months… when I only intended it to be for one. I wish I had started charging sooner, and I can recall one user even mentioning to "Let me know so I can pay you for your awesome service." Engineers, myself being guilty as well, seem to focus so much time and effort on solving the problem that they forget all the other things about running a business — namely making money. It's one of those lessons I think most people have to learn, even though they've been told many times. I definitely had to learn the hard way.

What are your goals for the future?

The best and worst thing about browserless is how new and unknown it is. Since this is likely the first headless-browser service, there's no outline of what to do. Because of that there's a good deal of education that goes into how the service works, and initially I didn't expect that to be a hiccup, so I'm hoping things like the new serverless workflow will help to alleviate that.

I also want to continue growing the service to larger consumers that use headless browsers for their infrastructure. I can't count how many hours I've spent in my day job trying to debug or fix headless workflows, which is something that costs companies a lot more money than any of the browserless plans.

Forecasting growth month-over-month, it'd be nice to get to 100 paid users by the end of the year. I'm doubling down on my documentation and blog/content-marketing to help achieve this plus more cold emails to companies I know could benefit from the software.

I think the final thing is that allowing 3rd-party integration from different platforms would be fantastic, and is something that serverless would help with. Zapier and even IFTT will easily be able to integrate once browserless is more externally accessible, which I feel will open up the doors to more users.

What are the biggest challenges you've faced and obstacles you've overcome?

Pretty much what everyone else has said here: "Charge your users." It was quite difficult to keep pushing on nights and weekends when money wasn't coming in, though it could've been if I'd done things properly. There's no better feeling than seeing that first deposit hit your bank, which is the biggest form of validation I think there is.

The biggest challenge I think I'm facing now is finding users. I know the problem is there, and I know folks are looking for answers, but it's a matter of education and getting the message just right. I'm trying to make some lemonade with this lemon by documenting best practices and common gotchas. Creating a community around browserless will likely be crucial to its success, and is something I'm trying to find an outlet for.

Have you found anything particularly helpful or advantageous?

If you're reading this: you've already found a helpful resource! The advice from folks on Indie Hackers has been priceless, and I can't imagine the amount of time I've saved because of it. There's also a wealth of great technology out there that can springboard your efforts quite quickly. I'd definitely recommend using Stripe, DigitalOcean, Docker, Github, and Netlify to help you achieve your projects goal. They're dirt cheap for paid plans and worth every penny for what you get.

I also feel that there's definitely an element of chance in getting something that generates income. I just happened to have a need for a library, wrote one, found users, and had the foresight to make it a business. This doesn't happen often, so if you see something out there that's trending and that you're good at, jump on it while you can.

What's your advice for indie hackers who are just starting out?

Sort your favorite open-source projects' issues by comments or reactions, and see what's missing that you could build and monetize. As an example, what if the Ghost blogging platform was just an open-source blog? You could have easily created a hosted version and started monetizing it yourself, and those guys are killing it. I'm sure there's a ton of other databases, frameworks, and other cool things that could be made into proper hosted services — it's just that no one is doing it.

The prior five or so projects I've attempted generated poor or no revenue, which I think is important for folks to hear.

The other piece of advice I can give — and it's been been given numerous times here — is to just go and do it. Build something that you wish existed but does not, and build it well. Not only will you have passion, but you'll also walk in knowing what exactly to build. For me I've always found headless browsers to be really interesting, and knew exactly what I myself would want as a developer, so I built just that.

Where can we go to learn more?

You can check out more browserless updates on Twitter at @browserless. Or you can follow me directly at @griffith_joel! Thanks for having me.