Building an Aggregator for Classifieds That Brings in $1k/mo

Tell us about yourself! What are you working on?

My name is David, and I created Value Forest, an aggregator that makes it easier for people to find the best deals on South African classifieds. Anyone buying secondhand cars, phones, antiques, and more, can use Value Forest to find what they're looking for, whether it's for their business or for personal use.

As of November 2016, Value Forest has 130k monthly unique visitors doing 200k searches a month, and it brings in around $1100 in monthly income.

What motivated you to get started with Value Forest? What were your initial goals? And how'd you come up with the idea?

I started it, I think the ideal way, with trying to scratch my own itch. Being a big fan of finding bargains on second cars or electronics, I thought it would be a nice project to create an aggregator for all the major South African classifieds and find a bargain.

After building a very simple version of it, I posted on local forums to share it and get some feedback. The response was generally quite positive. I was working full-time as a back-end developer for an established company. And still living at home, so my expenses were fairly low.

What did it take to build the initial website? How'd you find the time to work, and what was your tech stack like?

Because I enjoyed the project, I found the time mostly by working evenings after work and weekends, too. Thankfully I didn't really need any funding, because I was the only developer. My only expenses were server costs, which were very small.

I started with some Digital Ocean servers and a basic PHP/MySQL stack. I initially didn't think too much about the specific features, instead just focusing on the core functionality. But I listened to feedback from early users as well. I took a very iterative approach, improving the website over months.

The initial version was very slow for each search, as I was searching each classifieds site in real-time, which obviously wasn't very practical. Thankfully I formed a connection with a very knowledgable CTO of a big e-commerce website, and he gave me lots of great technical advice. One of those points was to use Apache Solr. It's basically a database and full text search engine in one.

I also used VPSDime, who provided a great $7/month high spec server to host it on. I then created Python based scrapers to crawl the websites or APIs where available, and ingested these results into Solr. The PHP code started getting quite messy, but a friend recommended the Laravel PHP Framework to me, and I rewrote a large portion of the website using Laravel to clean things up.

What have you done to grow and market Value Forest? And how does your business model work?

Initially I found users via the local forums I mentioned above (specifically, myBroadband and Carbonite) plus a bit of word of mouth. Since then, I've relied primarily on SEO long-tail traffic. At some stage I did join the FbStart program, and I used the credit provided to advertise through Facebook Ads.

Value Forest's income comes from primarily 2 sources: affiliate revenue and advertising via Google Adsense. Each accounts for roughly 50% of the income.

As of November 2016 revenue was close to $1100 USD. My earnings are in South African rands, so it's tricky to be accurate as the exchange rate has changed quite a bit in the past 2 years.

I see lots of founders working on "aggregator" businesses. Are there any challenges specific to this approach?

Yeah there are definitely challenges. "Scraping" often has negative connotations, and to keep things simple a lot of websites would prefer that you refrained from doing this at all. Some go as far as including this in their terms and conditions, and in some cases this can result in legal action. For example: Craigslist Inc. v 3Taps Inc.

What makes the relationship interesting is that a lot of websites in the classifieds industry make their revenue from advertising. If the aggregator becomes successful, it will be bringing a lot of free traffic to those websites. As the aggregator isn't the endpoint, the websites will still get their ad revenue, and possibly more of it. So I think some websites have mixed feelings about aggregators. Ideally the aggregator should get permission.

The other challenge with scraping is that as the websites change their HTML, the scraper has to be updated, so an API or data feed for each website is usually preferable if one is obtainable. Again, this can prove difficult in my experience.

This paper provides more insight: Legal Challenges and Strategies for Comparison Shopping and Data Reuse.

I'm inspired by some the of big flight aggregators such as Momondo. I really like the way they bring flight data together and make it useful. My goal remains the same as when I started: to help people find bargains on classifieds in a convenient and enjoyable way.

What are the biggest lessons you've learned so far? If you had to start over, what would you do differently?

I'm quite happy with the way I've progressed. I think each journey is slightly different, and making mistakes is a part of it. It's great hearing other people's stories for motivation, but I believe you still have to make the mistakes yourself to truly internalize them.

Arguably I could've started exploring other markets by now, which I've now started working on. Also quite often I've struggled with the conflict of whether I was making the best use of my time and focusing on the right priorities.

What's been most helpful to you on your journey? What do you think your biggest advantages have been?

Actually my boss at the time was very helpful to me. He was an entrepreneur, and he encouraged me to pursue the website. Also, there's definitely been an element of luck — I didn't expect the SEO to work as well as it has. When I started, I only had a pretty basic understanding of it.

I also went on a bootcamp which led to an accelerator program that taught lean startup principles and gave me constant advice and feedback. This was very helpful.

Finally, I've read some books with great lessons and advice that have helped me along the way. The 4-Hour Workweek was an interesting read, and I've enjoyed listening to quite a few of Tim's guests on his podcast. I also liked Rework by Jason Fried and DHH. And I have to include The Lean Startup by Eric Ries.

What's your advice for aspiring indie hackers?

My advice is probably similar to what most entrepreneurs who have been "successful" will say. Ship it! The logo doesn't need to be perfect, nor the name or domain name. Take a lean iterative process, and try not get stuck on paralysis analysis. Get a basic version out there, and get feedback.

Also, try not to get too attached to your preconceived ideas of the product/service. Easier said than done!

Where can we go to learn more?

Check out the Value Forest homepage or follow @ValueForest on Twitter. You can also leave me a question or comment below!