My heart started racing back in March when I saw a post from the Official Google Blog entitled A second spring of cleaning. Like many others, I wasn’t too happy about the loss of Google Reader. But that’s not what had me worried. The thought that crossed my mind was “this is finally the end of SportsLizard.” I opened the post (in Google Reader ironically) and confirmed my worst fears.
Before I finish the story of that day and explain what led to yesterday’s relaunch, for context I’d like to tell the story of SportsLizard from the start.
It’s the spring of 2004. I had just returned from a summer + semester co-op at Schick, where I’d eventually work after college. My previous two summers I had also had internships. As I approached the summer I wanted nothing to do with the effort involved in finding yet another. Most college kids are lucky to get one, I had three plus an extra summer because the co-op pushed back my graduation to December. A class called Information Systems had piqued my interest. Industrial engineering students like myself were paired with computer science majors to work on semester-long projects with local businesses. My group worked on the flow of patient information in the Alzheimer’s unit of a local hospital. Quickly it became apparent that the programmers in the group were able to understand the IE workflow better than we were able to understand their code. As the summer approached I got the idea: what if I taught myself how to code?
The idea sounded better and better the more I thought about it. I had caught the entrepreneurial bug at Schick when another student and I invented a new quality testing device that we eventually failed at trying to commercialize. We had also failed to win business plan competitions on campus to get funding for another idea. With the web blooming, my pockets low on cash, and my partner now off at a full-time job, the timing seemed to be right to go at it myself. As an avid sports card collector at the time, I had long envisioned a marketplace better than eBay geared strictly towards collectors and dealers.
That summer, I purchased a new desktop, a few books on PHP and MySQL, a few more books on starting your first business, and set off to launch my venture. Here’s what the site looked like when I launched in July of 2004:
Despite some cool features for the time, the marketplace idea failed. I didn’t have enough know-how or resources to solve the chicken and egg problem of getting enough buyers and sellers simultaneously. I did however catch the entrepreneurship bug…hard. In the years that followed the site became the original location of this blog, a much more popular (and more edgy) sports collectibles blog, a free-to-use listings site (think Craigslist for card dealers), one of the most popular resources for learning how to customize on the web, among other things.
None of those were a financial success either, but we really struck gold in April of 2007 after having formed Pure Adapt when we launched our Price Guide Tool, a freemium service that we charged $4.99/month for. At the time I wrote extensively about how we accomplished it. The key was using the Google Base API to pull in the data in real-time when someone searched. Everyone else who attempted this tried to piece together a database of an impossible number of cards/collectibles. Instead, we skipped all that and relied on the API and our marketplace for data. That API that eventually powered Google Shopping contained almost every collectible for sale on the web – eBay, Amazon, sports card marketplaces, individual online sellers.
We knew we were on to something when some of the big players in the business called us up asking how we accomplished it. We then spiked the growth with the most successful AdWords campaign I’ve ever seen – we had a click-through-rate of over 30% and only spent 11 cents per click! Also, because of the success of the sports collectibles blog, it was easy for me to do the PR rounds in the sports blogging world to help promote the launch.
Here’s what a pricing report from back then looked like:
I had forgotten about that tag line “Once you go lizard, you never go back”. I also had a shirt that said that – I’d wear it to card shows. Makes me chuckle to think about it.
Cruise Control – Every Entrepreneurs Dream
After that, we very successfully did absolutely nothing. Aside from a redesign in 2008 and some API stuff mentioned below, from 2007 until that blog post by Google we did next to nothing. Subscriptions continued to build, and while there was subscriber churn, it was less than the rate at which we were adding new customers. It sat there slowly building itself into a really profitable business without us lifting a finger. Aside from answering a handful of customer service questions per week there was no maintenance, and aside from PayPal fees there were no costs involved. Revenue-wise it paled in comparison to Detailed Image, but profit-wise it was not insignificant. There was no cost-of-goods-sold, no warehouse, no employees, no shipping costs. The cash from SL was an important part of what has led to Pure Adapt’s overall success.
Because of how easy it was I’ll probably be chasing that success with ever venture I attempt for the rest of my life…which is either good or bad, depending on how you look at it. At least I know it’s possible!
That Damn API
Back to my original story. That day when I opened up that blog post I knew they were shutting down the Google Shopping API (what was originally the Google Base API). I just had that feeling where you know something is about to happen. Then I saw it. Towards the bottom of the post it read:
We’re deprecating our Search API for Shopping, which has enabled developers to create shopping apps based on Google’s Product Search data. While we believe in the value this offering provided, we’re shifting our focus to concentrate on creating a better shopping experience for users through Google Shopping. We’ll shut the API down completely on September 16, 2013.
Damn. While that explanation makes no sense to me, there’s nothing we could do. We’d have to shut down our nice little profitable side venture.
I can’t say I didn’t see this coming. In July of 2011 they started capping the API calls, forcing us to adjust the number of free calls available and forcing me to create a more advanced caching system. When I appealed for more calls, as they said we could, they never gave us an answer, just that it could happen sometime soon. When it never happened, and the forums were filled with frustrated people like myself, I saw the writing on the wall. In some ways I’m surprised it lasted as long as it did.
The timing couldn’t have been worse. We were in the middle of working on the responsive redesign for Detailed Image that we launched in April. My wife and I were in the process of buying a house. I was just generally exhausted from the busyness of work and getting married in 2012. At least they gave us until September to figure things out.
We stalled until after that launch in late April. We had three options: shut it down, create a less-good free service, or try to license the data we were aggregating from Google from the individual companies to keep the thing alive. We decided to pursue the last option to the fullest extent because we’d never again be able to carry over a large paying subscriber base. Once you cancel their subscriptions you can’t get them back.
We thought about it some more and decided that if we were going to pay for the data we needed, that we might as well relaunch the website, and if we were going to do that we should also consider pursuing other verticals that need price guides like action figures, stamps, coins, comic books, etc. We decided to go for the jugular and turn this thing into a large business in and of itself. That week after the DI launch I was fixing bugs, closing on my house, and trying to license the data we needed. It was crazy exhausting.
Unfortunately it didn’t work out. The cost and terms to license the data we needed would be so cost-prohibitive that we’d more than cancel out all of the recurring $4.99 subscriptions we’d be carrying over. At that point you’re starting a business in the red, negating any benefit of having that large subscriber base.
But we do have a large loyal user base who just want to value their cards and collectibles. We could still piece something together with freely available APIs. We wouldn’t be able to charge for it because of the data drop off, but we could still use affiliate links when referencing products, which accounted for roughly 30% of the revenue on the old site. If we made a useful site, maybe we could grow that number and continue at least pulling in that 30%, if not more.
Build It FAST
The new objective became this: build the best possible site using free API’s without spending much time or money, and make sure that after it launches it requires no customer service and very very little maintenance. We want to focus our time and money on DI while still hopefully providing a useful enough SportsLizard that it generates some side revenue.
Lucky for me we have built a pretty nice back-end PHP/MySQL framework that powers DI and LockerPulse. For the LP ad platform, I wrote classes to tap into the eBay and Amazon API’s. What we were missing was a way to do a fast, clean, responsive front-end without tying up Mike. Enter Bootstrap. I was able to do the entire site myself. Combining those things made my development soooo fast. I’ll definitely be using Boostrap again on future sites. All told I probably spent two weeks worth of time on the new site from start to finish. I also experimented with using HTML5 PushState to load new pages faster and smoother. The URL changes, the content inside the page changes, the title of the page changes, but the template header/footer don’t change and the browser never loads a new page. It gracefully falls back to the links functioning traditionally in older browsers. It’s too buggy for a site like DI yet, but I wanted to experiment with it on an entire site and SL was the perfect test case. If there’s interest and no one else does it, maybe I’ll release the code as an add-on for Bootstrap.
I kept all of the popular articles that received steady traffic, but got rid of the rest of the site. What’s left is simply a search box for shopping for and valuing cards/collectibles:
The resulting report can be used to shop or to value (or both) depending on your query:
Search results are sorted by the basic-but-solid natural language sorting I built for the LP search engine and then improved upon for the DI guide and blog. eBay doesn’t allow us to use their listings in aggregate to create a pricing report, so that’s why the value is based upon the Amazon results only. The eBay Completed Listings displayed below are really the key for valuing so it was important to get those in there, along with whether or not the item sold.
Back on Cruise Control
Without subscriptions to manage, emails should come to a halt. We created a new email address as well and will be closing out the old ones from 2004 that get flooded with SPAM after a little time passes. Hopefully with the responsive design and smoother functionality the result will be that collectors sit around entering in their collections for hours. That’s what happened on the old site, and early early indications from day 1 of the new site seem to show that people are doing the same with the new site. And if it turns out they do, hopefully we’ll be able to make a few bucks on the affiliate sales with almost no maintenance.
Now it’s time to get back to focusing on growing Detailed Image as the holiday shopping season approaches. Ultimately we decided that the money spent to license the volumes of data we’d need would be better spent on a more sure thing – Detailed Image.