You’d think after 13 years I’d know better. Some lessons, though, are easily forgotten.

On Thursday last week we shipped version 1.1.2 of powerOne Financial Calculator to the App Store. It was accepted into the store on Friday. Saturday morning the first bug report rolled in that there was something wrong with template creation. It was pulling other templates in. Testing on this problem led me to a bigger problem — any templates created before version 1.1.2 were hidden from view.

On Saturday evening, after a half hour bug fix and a day’s testing we shipped version 1.1.3 to Apple, which was accepted into the App Store today. Luckily we escaped major catastrophe as neither bug was fatal — all previous templates were recovered and new templates are now created correctly again.

We have been at such a frantic pace here that we got lazy about testing. Sure, we were testing around the areas that were changed but we weren’t spending time testing everything before release. We didn’t think we had touched these areas — user-created templates — with version 1.1.2. We were wrong.

I learned this lesson the first time in March 1999. We shipped version 2.0 of FCPlus Professional… then version 2.0.1 then version 2.0.2 then version 2.0.3 all on day one. Nasty lesson learned.

So a reminder for this decade: write test cases, expand test cases, and actually step through them all before shipping.

A lesson learned that only caused a day’s panic and three days worth of consternation… but nothing else. Luckily.

We receive a number of requests each week regarding powerOne software calculators on different platforms or different devices that we don’t support yet. We love these, by the way, so keep them coming as it helps us prioritize when it comes to new work.

We currently have product for iPhone, iPod touch, BlackBerry (non-touchscreen devices), Windows Mobile (touchscreen devices), and Palm OS. The Palm OS version runs on webOS devices if you purchase Classic emulator. And the iPhone version will run on iPad, although we have already started working on a native version.

There are some market dynamics at work, and this is why we have not yet supported Android, webOS natively, BlackBerry Storm or Symbian devices. I’ll outline that here.

First we are a small company, able to focus on one platform at a time. Developing for a new platform is expensive and the return on investment is long due to the fact that powerOne software calculators are fairly complex to develop. There are lots of development reasons to focus on one platform but there are some market dynamics also.

BlackBerry

We wrote a first BlackBerry app before RIM announced their touchscreen devices, which are a completely separate development process on the front end. RIM has plenty of users — adding about 20M per year — and hypothetically these folks are right in our wheel house as they are a “professional” device.

For us, there are two major problems here. First, it is not clear that the BlackBerry devices are really available to third-party apps like ours. A lot of those sales are to consumers who like to text and send email but wouldn’t purchase a financial calculator. Another large chunk of customers work for corporations that lock down those devices, not allowing their employees to install software on them.

The second major problem is one of discovery. Since there is no unified location for people to find BlackBerry apps, it is very unclear how powerOne would be discovered beyond word of mouth. We spent months marketing this product without a lot of success. We also partnered with RIM thinking they’d help but their Alliance Program is geared toward what we’d do for RIM rather than what RIM would do for us.

I’m still convinced BlackBerry is a good place for us, but clearly not the one where we are going to make a lot of money in the short-term, and, as we needed to prioritize, iPhone with its 75M devices and centralized discovery site took precedence.

Android

Android is just getting hot but few developers are making money on that platform yet. What do I mean by making money? I mean making enough money to pay for three people and cover the bills, which runs at least $20,000 per month (for very low salaries).

There was a great story recently about a developer who made $13,000 in a month (not per month, in a month) and this was heralded by the press as a sign that developers can make money now on Android. But it was clearly an outlier case that occurred because Google featured the app. An outlier case for iPhone is millions of dollars per month, a huge difference.

I have other concerns — like platform fragmentation — but these won’t stop us. Could we do well on Android? I think so. But we started on iPhone and need to see that through before we jump to a new platform and take on that multi-month development cycle and learning curve.

Symbian

We have more requests for Symbian versions than any other, simply because of the length of time it has been around. I have two concerns here. First, will it survive? The main hardware platform has been Nokia and Nokia is putting a lot of time and effort into its Maemo and now MeeGo platforms. Will the company keep using Symbian OS?

The second concern is localization. Since powerOne is highly specialized and Symbian is primarily used in Europe, we’d need to understand the computations used in Europe to make sure we create the right ones. Plus, we’d need to localize the app into European languages, which we’ve never done before.

webOS

So many questions here: will Palm survive? Can they sell enough units to pay back the investment required? Is the Classic emulator with powerOne for Palm OS platform good enough?

Until a month ago we couldn’t even think about supporting it as we have a large chunk of code written in C or Java that we needed to use. Palm had no way of supporting that and had denied to my face that they ever would.

I have to admit there is another factor here. In 2004-5 Palm made three moves that hurt Infinity Softworks badly. After all the work we did to support their platform, these three moves — ending our bundling relationship, jacking up the costs to sell software through their site, firing their education team — killed our business plans and sent Infinity Softworks on a downward spiral that almost ended the company in 2007. It’s a lot like being dumped. It’s really hard to stay friends with old lovers.

Windows Phone 7

Uhhh…. vaporware at the moment.

iPhone, iPod, iPad

We made our bed a year and a half ago when we picked iPhone as the platform to rebuild Infinity Softworks. We’ve learned a ton over the past year about developing and marketing a product. Now is the time for us to make it work. And that’s why we stay focused on iPhone.

I’ve been thinking a lot lately about pricing and business models. When we started in the business, we developed high-functioning software calculators for Blackberry, Palm, Windows Mobile and Windows and were able to derive prices in the $60-$160 price range, depending on the product and niche. After a series of customer surveys, reading other developer blogs and books, and playing with our own prices in Apple’s App Store, it has become clear that these customers are no longer willing to pay these kinds of company-sustaining dollars. Instead, the price expected is diving toward 0 and there is now a disconnect between “price” and “value”, at least for digital goods, that had been there in the past.

(Please keep in mind: I’m not complaining. It’s reality and I’m trying to figure out how to adjust.)

In the old days, there were two classes of software: Nice To Have and Must Have. The Nice To Have software was just that. It was great if I had it but if it’s not there, I won’t fret about it. Since the customer really didn’t need it and there was no urgency to have it, customers didn’t value it and the price was low or free. Must Have products were the opposite: it was needed and there was urgency to have it, and thus prices were high. While we were developing a niche product, it was a high-need item within those niches and we could charge good money for that software.

Now it seems like the market has shifted. Now, because of the breadth of the Internet and the App Store and the ability to find people who Must Have your product or service, it seems that the Nice To Have category of products has disappeared. Instead, Must Have has split into two segments: Must Have for me personally and Must Have where I am a contributor.

If I’m a contributor — meaning that there is something I am creating that must be retained or stored or shared — than I am willing to pay for it. But if it’s for personal use than the expectation is free.

When did this change? I don’t know but it was occurring long before the price pressure of the App Store. Google has been rolling out “free” services (meaning free to the end-user) for years and the freemium model (where some stuff is free and other stuff is paid for) has become a predominant model for web applications.

We, too, have been going through a tremendous transformation, re-evaluating what should be paid for and what should be free in the FastFigures eco-system, re-evaluating whether we are a mobile application with web components or a web application with mobile extensions, and trying to figure out how we build a stable and growing company again.