Don’t fix what aint broke

When I was 13 I created a site called 22Pixels, a forum where teenagers like myself could share techniques for Photoshop. It was a simple PHPBB forum mostly focused around making “sigs,” signatures or content that would appear below every post a user made.

I got a young start as my dad gave me the tools and knowledge to put up a basic web page with FTP. He’s an old school hacker who runs Bootdisk.com. He would always always say “Don’t fix what ain’t broke.” It was his recipe for success.

Success

As I learned to program a bit better, I created a list of the best photoshop downloads. It was a simple robot that would crawl other photoshop sites, copy some text, transload an image, and add the link to the index. People could vote reddit style on what the best downloads were.

It worked very well. I became the top result for “Photoshop CS4 Brushes”. I dominated the market for people searching specifically with ts version in their query.

Then CS5 is released. I change my SEO to match CS5. This worked even better. My traffic grew steadily as I continued to scrape downloads from other websites.

Failure

In late 2011 I started Hacker League with two friends from college. Everything else got put on hold, and Hacker League took off. It had our full attention.

The story ends well for Hacker League, but not so much for 22Pixels.

See, as I pointed my attention toward Hacker League, I removed it from 22Pixels. I made two awful decisions because I was not giving 22Pixels enough attention.

I did not watch site growth closely and I was not aware of what was driving site traffic.

The Big Picture

When I logged into Google Analytics, this is what I saw:

This is a monthly view of traffic. For a site like 22Pixels this kind of information was worthless.

The site runs autonomously and since I wasn’t doing anything active to promote it, the monthly view wasn’t very helpful. I would check and make sure that the overall traffic was increasing, but never looked at it relative to the past 6 months or a year.

This was my worst mistake, I wasn’t even aware that the site was growing. This led me to make some idiotic decisions in 2012.

Things Break

The site had a”tags” feature. It was an awful hackjob that slowed down every page load considerably. As the site grew, it had a larger and larger impact on the site performance.

I wrote this off to being a problem with the growing database, not traffic. I decided to remove it because it was raising my hosting costs and I didn’t think it could possibly provide anybody with any real value.

But it turns out the search engines must have loved the unique page titles it provided. When it was removed, my traffic started slipping.

I wasn’t drilling into the stats deep enough to even know that this was providing decent traffic. I wasn’t checking them often enough to know that it had even changed anything.

The correct answer here would have been to rewrite the feature. To watch the stats. To understand the market.

Fixing What Ain’t Broke

When I changed my SEO for Photoshop CS5 I had great results.

CS6 was due to come out in May 2012, so I decided it was time to play my trick again. I updated one variable in my configuration and “CS5” was now “CS6”.

The graph above is the direct impact changing this variable had. It’s about 5,000 views a week difference. You can see, it took me 2 months to notice.

And when I changed it back, the stats pop back up. But my SEO already took the hit and it was too late.

What Aint Broke

The above is a graph for “Photoshop CS[x] brushes” where x is the photoshop version from CS1 to CS6.

It’s easy to see why this happened now. CS5 is the red line and CS6 is the yellow. It turns out that people had been searching for CS5 the entire time. I was 8 months early to changing the keyword.

Take a close look at the end of the blue graph and the beginning of the red (CS4 vs CS5). CS5 was almost immediately adopted and CS4 took huge hits.

Now look at the end of the red line versus the beginning of the yellow (CS5 vs CS6). CS5 slowly transitions into CS6.

Additionally, each release of Photoshop dilutes the market as I can only choose one version to stick with. At the end of 2012 you can see that people are still using all the versions of Photoshop CS!

Don’t fix what ain’t broke

The larger morals here are that

When you figure things out, don’t change them.

If something breaks, repair it so it’s back to how it was

You need to understand the difference between fixing, updating, and improving. You need to recognize the reasons for doing so and the risks involved.

You especially need to understand your market and how it will change, because it will.

Make something successful and keep it successful.. Don’t fix what ain’t broke. Stay the same and change what’s only nessesay. Do the research to find out what’s needed.

And remember, you can only focus your attention on so much. If you let something start to slip, it’s going to slide.