Bouncing Back From the Brink to make $1000 in My First Week

Tell us about yourself and what you're working on.

Hi, I'm Chris Chen. I created Instapainting.com, a photo to painting service that lets you turn a photo upload into a real hand-painted piece of artwork done by an artist in real life.

For the first two years I was operating purely from earned income, making about $30-$50k of profit to support both myself and the business (in expensive San Francisco, no less). Now the business is in its third year and doing over $400,000 in annual revenue. I'm still bootstrapped.

How did you get started with Instapainting?

I did YC in Winter 2011 as a solo founder working on a social music site that was a clone of Last.fm. I was fresh out of a six-month hiatus from college, which I've yet to go back and finish. (I'd only completed three semesters of a physics major.) I survived on the money I raised from YC and my various ideas and pivots for about three years.

Then I ran out of money.

Luckily, by this point I had already been testing more and more random ideas that deviated from "social music", and I had gotten pretty good at throwing up MVPs. I had a friend who bought paintings from China and sold them in the US, and she wanted me to build a website for her to sell art reproductions. Instead, I launched Instapainting in January 2014.

The initial prototype was a simple single-page website that allowed customers to upload any photo they wanted, collected their payment info via Stripe, and explained that a real hand-made oil painting of their photo would be mailed to them in two weeks. The hardest part was actually conveying that it wasn't just some print or photo filter, and this is something I still have difficulty with today.

At this point, I was about $4,000 in debt, so I had to put up the single page website to collect money. I created, tested, and launched it in just a couple of days. I put the page up on Reddit, and in two weeks I'd made over $2,000 in sales. It felt really refreshing to go from a stereotypical failed Valley startup to something that was revenue positive from day one.

After that, I started trying to secure the actual artists, but I failed to do so. I ultimately ended up hiring some local friends who were artists to help paint the orders. The total profit came out to be slightly negative, but it proved that the idea was sellable. The next steps were to increase the profit margin and to make a bigger splash. A few weeks later, I launched on TechCrunch and did over $30,000 in revenue in the next three months.

How'd you hire artists and get the business model working?

My friend had introduced me to the concept of hiring artists from China and selling the paintings in the US, and she even referred me to one of the suppliers she was using. While I did make my initial sales before finding any real artists, acquiring artists turned out to be far from the hardest part of the business.

Shortly after the launch I was contacted by many artists, mostly from China (they already saturate other marketplaces like eBay, Alibaba, and Etsy). The initial batch of orders were break even, but once the market-rate artists contacted us, I knew it was going to be profitable. The question was how to keep sales up and how to scale it.

Initially Instapainting started out as a single-page site with a Stripe Checkout form and with orders manually doled out to artists. Today it still appears that way to visitors for the most part, but behind the scenes it's a fully-automated marketplace system.

Artists and sellers on our platform directly communicate with customers and handle customer service issues, project updates, and shipping status tracking. Bi-directional feedback ratings are employed for automatic quality control, and bidding algorithms are used to select the ideal artists for each project. I've gradually adapted the software over time to automate every process, and generally I only have to step in in rare instances of high-level disputes.

What's your tech stack?

I've used PHP, Node.js, Mongo, and React.js. The stack consists largely of micro-services, which is crucial in allowing me (the only developer) to migrate the site to new technologies.

What have you done to grow and market Instapainting?

I think it took about 5-6 months for me to realize that the business was primarily sustained from SEO traffic. A friend had noticed what I was doing and gave me some SEO advice that had some pretty clear positive effects. (He later turned his consultancy into an SEO startup called Rank Science.) At that point I started focusing heavily on SEO.

I reaped the full benefits of this approach in my second year of business when I climbed to #1 spots on Google for the top search terms. I didn't have much experience in SEO beforehand, but I leveraged my experience on Hacker News to push out content marketing pieces that were crucial to SEO. For example:

• I built a robot in 2 weeks that could autonomously replicate artist paintings in black and white and, later, in color.
• I launched the very first free-to-use online demo of the algorithm in A Neural Algorithm of Artistic Style, I believe 3-4 days after the paper was released.
• I released an article detailing some of the things I saw when touring artists and art factories in China, which was read by Ev Williams, re-published on Backchannel, and totaled over 71k views.
• When the 2048 game came out, I quickly hacked up a 2 player version and placed it under the Instapainting.com domain for SEO purposes.
• Being lazy, I milked the robot I'd made earlier and hooked it up to Twitch.tv, where people could control it to collaboratively paint.

Most of these things had orders of magnitude more traffic than the Instapainting home page ever received, and they usually maintained some traffic gains consistently afterwards. But most importantly, these efforts brought in massive amounts of inbound links from reputable sites like Intel.com, Wired.com, TechCrunch, Engadget, etc.

Pretty much all of these initiatives started by making front page on Hacker News, and then they'd get more coverage afterwards which would boost SEO, even if the coverage itself didn't convert to orders:

• Twitch paints — 109 upvotes on HN
• Inside The Fine Art Factories of China — 252 upvotes on HN
• Neural Algorithim Demo — 75 upvotes on HN
• Robot Painter — 73 upvotes on HN
• 2x2048 — 71 upvotes on HN

Those massive spikes were due to the content marketing initiatives. You can see after those spikes that traffic permanently improves. That was due to improved Google search rankings.

What's the story behind your revenue and expenses?

On day one, the website was just a Stripe checkout form embedded on a single page. I mentioned earlier that, after posting on Reddit, it immediately started doing $1k/week in sales.

The business became profitable about 2 months in, which is when the real launch on TechCrunch happened. By profitable, I mean it was able to cover my living expenses. Before that it had been about break-even as I worked things out. It couldn't afford to lose much money, because the whole point of the pivot was to stave off (personal) bankruptcy.

Today Instapainting is earning about $400,000 of revenue per year. As I'm fairly intertwined with the company, this is shared personal and business revenue. This is up from the first year, which was only about $89-$90k in revenue. The main driver of the increase is improved SEO from content marketing.

In tandem with increasing revenue from SEO, I cut costs by introducing a marketplace bidding system behind the scenes, and by on-boarding more and more artists. I also cut the amount of time I have to spend in day-to-day operations by automating all aspects of the business. Things started out with me manually doling out orders and managing the artists, but today I've got a completely automated system enforced by bi-directional reviews and direct communication between customers and artists.

No one except me is on payroll. The whole point of the platform is that many independent vendors and artists can come on the site and provide their services without much overhead. Hiring people would be overhead, and would only be necessary if we were growing much faster than we currently are.

If you could go back, what would you do differently?

Initially, my biggest fear was the expected "trough of sorrow" after the TechCrunch launch. The challenge immediately became how to permanently increase growth. As soon as the site started doing well, I would divert my attention and try to branch out into other things, often times too divergent from the main products. These turned out mostly to be a waste of time.

Another thing I'd do differently is that I'd pour my time and resources into SEO earlier. It took me over 6 months to realize why we were still making sales even after the TechCrunch article's shelf-life expired.

I would have also spent less time focusing on competitors. I wasted time and money testing Google ads, and also matching features and copy from my competitors' websites. In the end they were non-factors, as the only thing that mattered was our Google ranking.

What's been most crucial in helping you to succeed?

For me, SEO. I'll admit SEO won't apply for all types of companies, but it's probably responsible for most startups' successes. Unless you have no competitors, or you spread exclusively through some other channel like word-of-mouth or the App Store, then SEO is your main way of being discovered.

A litmus test is to ask yourself if you have more than 2 competitors. If you do, then it's a pretty good sign you should be focusing on SEO, as some absurdly high percentage of people go with the first result on Google by default. This doesn't just mean optimizing your website for search engines. It means gaining publicity so that Google recognizes that you should have the most mind-share and puts you at the top of results.

What's your advice for hackers setting out to be their own boss?

The most important thing for Instapainting was just doing it, and in the fastest and most testable way possible. And, as time and sales progressed, a big help was being able to engineer and automate the systems so that I had more free time to dedicate to new features and new sales strategies.

My advice to others is: Your idea doesn't have to have hundreds of thousands or millions in funding to be realized. If it does, then it's probably the wrong idea to be pursuing if you don't already have that kind of money. Your idea doesn't even have to take months or years to implement. It can be as straightforward as some simple proposition attached to a payment form. If you don't have a simple proposition to make, go seek out a non-technical revenue-generating business and help them out with their tech strategy.

Where can we go learn more?

Check out my posts on the Instapainting blog, or leave me a question in the comments below and I'll try to get back to you.