Software Bugs Me; Numbers Make It All Better

Everything we ever do comes back to us someday. My education in writing has never earned me much money, but it pays off daily in my website text, my newsletter, and this blog. When I realized that my English degree was useless, I washed up in the bookstore business and learned how retail works. Hatred of shopping malls eventually steered me into a science museum gift shop where I learned a few things about selling non-books. I dulled the boredom of that dead-end career with an obsessive computer gaming hobby; in DOS days, that meant learning the technology necessary to make them run. The death throes of the bookstore industry propelled me into PC game development. Now I’m doing something that I cobbled together out of those writing, retail, and computer skills.

Lately I especially appreciate the unhappy year that I spent testing broken games in Quality Assurance, because most of the programs that I use today are crippled to some degree. Take Simplaris Blogcast, the fragile little app that spews the first few lines of these posts to my Facebook fans. Last Friday it crapped out with this message:

Error: Failed to get posts
SQLSTATE[HY000]: General error: 145 Table './Blogcast/posts' is marked as crashed and should be repaired

Lovely. I couldn’t fix that myself, so I went looking for someone who could. The Simplaris website doesn’t mention Blogcast anywhere; however their name got onto it, they’ve apparently washed their hands of it. The real creator’s website comes up 404 Not Found. Now I’m stuck using this abandonware because integrating this blog is important to my nebulous FB strategy and I don’t know of any other program that does it. While I debated whether to reinstall it for a third time or abandon it myself, it suddenly cleared up. Whether somebody repaired FB’s MySQL database or the program miraculously healed itself, it’s fine now -- until next time…which I fully expect to be moments after I finish writing today’s post.

Sunday’s Sunshop upgrade from version 4.2.0 to 4.2.2 went smoothly. But hours later, a customer complained of the dread PayPal mystery loop bug. Crap! I had thought that this patch addressed that, but I was mistaken; it actually fixed a different PayPal bug that I reported. Sometimes I feel like I work for Turnkey’s QA.

I wish that I were paid accordingly. I wasted three more hours on Monday trying once again to reproduce this bug. Once again, I failed. I abused the checkout procedure in every way that I could imagine with a variety of different products, and my test transactions went through properly every time. I’ve gotten only sketchy information from the customers who report this bug. They seem to be low-skilled computer users so they’re probably using Internet Explorer, and quite possibly an old version, but they can’t even tell me that much. Maybe they’re infected with spyware or viruses (as most unskilled users are). For reasons that you won’t care about, I think that there’s more likely something wrong with their computers than with them. My strongest suspicion is a cookie problem – once they encounter the loop, these customers can’t escape it even in subsequent sessions -- but none of them have been able or willing to delete their cookie and try again. It’s almost like they don’t even want to volunteer in the Kraken Enterprises QA department!

Ordinarily I would take this bug to Turnkey or pay my developer to figure it out, but the first thing a developer needs is steps to reproduce. You can’t fix a bug that you can’t see. And so the PayPal loop endures. I’m sure that I’m losing some sales to it, but I don’t think it’s very many. It’s mostly embarrassing and frustrating.

Sunshop 4.2.2 wasn’t a total waste, though. It introduces a one-page checkout screen that might improve conversions very slightly (possibly even offering a workaround for the PayPal loop…not that I can test it). It also integrates a plugin called ShopBuddy that exports my product catalog to a Facebook tab. Now users can browse my store without leaving FB. I don’t understand why anybody would want to do that, but I don’t understand the appeal of FB in the first place. So I’m doing a 6-month free trial of a program that will supposedly cost $8.99/month starting in August – that’s more than I pay for web hosting!

OK, so my blog and my store are integrated with my FB page. It even has a newsletter signup form. I think it’s time to use the $25 FB ad credit that I’ve had since Dec. 9. Should I advertise a particular product or my store in general? Generic advertising has always flopped before, but FB is entirely about the mass market; I don’t think that I can promote anything other than my page. Is that worth doing? Maybe I can buy an answer if I match that credit with $25 of my own. I don’t expect it to drive sales directly, but getting new followers could make it worthwhile.

March will need a little goosing after a February for the record books. Only seven of my 50 months in business have surpassed this February, and six of those seven months were Novembers and Decembers. Just look at these numbers.

February, the Miracle Month:

Total income: +148.6%
Total COGS: +180.3%
Payroll: +114.8%
Net Income (Profit): +133.7%

Year to Date:

Total income: +62.3%
Total COGS: +93.6%
Payroll: +78.5%
Net Income (Profit): +149.8%

I ought to just retire right now. Those numbers will never look better than that. The Cost of Goods Sold is slightly worrisome; that should track total income more closely. But that’s looking for the dark cloud inside a silver lining. Get this: My advertising spend is only up 2% over LY! Consumer confidence hit its historic low point in February 2009, so beating LY ought to be easy for any retailer. The economy still sucks, but less than LY.

March’s target looks achievable until week 4, when I had another one of those outlying corporate sales last year. I’ll be happy if I can get over that $1,200 bump and finish flat with LY.