If you are following along you will know that the latest iOS App Store changes caused us some (hopefully temporary) misery. Instead of doing what I was supposed to do yesterday, I spent the morning getting new revs of powerOne back into the App Store so we can change the keywords. Apple’s end-of-week changes meant that we were no longer found for searches like “mortgage calculator” and “rpn calculator” although we were front-and-center for searches like “mortgage” and “rpn.” In the end, the changes will be a good thing for customers and us but the near-term pain is a 30% drop in sales.

I’ve been at this a long time and app store changes wreaking havoc is nothing new. In the Palm OS/Windows Mobile days, we worked with two primary e-sellers, PalmGear and Handango. As sales slowed for all handheld devices, these two decided they would jack up the rates they charge developers. Instead of 25% it soon became 35% then 65%, which was more than our physical product resellers were charging. Handango also served as the back-end for Palm’s website. I was so annoyed that I lobbied Palm to switch to PalmGear. What’d I get? Worse terms when they switched. Not only did they keep the rate at 65% but they also forced everyone to remove their own website links from their apps.

When we rely on a third-party to make our sales, we are just asking for trouble.

In a nutshell, here is the vicious loop of what has happened: with the iOS App Store being the only source for sales to iOS customers, every sale goes through it. Because every sale goes through it, all customers search there for the product they want. Because distribution is restricted and marketing in the App Store is limited, competition starts revolving around price alone. Prices get compressed and with lower prices the opportunity to spend money on marketing the app is limited, forcing us to focus on the App Store for our marketing, which reinforces to customers that it is the only place to find apps.

This isn’t exclusive to iOS. Android has two leading stores, Google Play and Amazon’s store. (Yes, there are a number of vendor sponsored ones as well but the reality is the same: customers search for apps in these stores.) Microsoft has one. RIM has one.

Furthermore, this not a rant against any of the app stores and providers. There are good things to this, too. Installation, reinstallation, and purchasing are all so much simpler than they were back in the day. With smaller volumes than we have now, we used to need a full-time support person who spent 80% of his time handling install and reinstall problems. We even needed telephone support with 800 numbers because many of the issues weren’t resolvable via email.

Nor is this a problem for all kinds of apps. Games, for instance, seem to thrive in this environment.

But what it leads me to conclude is that app stores are no place to make money. Not real money, anyway. App stores are great for distribution. App stores are extremely efficient for the process of transferring money from customer to vendor. But what they are not is a great place to differentiate, a great place to market. We are too much at the whims of the app stores.

We are working on something new (I have hinted a hundred times and will share very soon). We are exploring a freemium (get some for free, pay for more) subscription model. The pricing and recurring revenues would actually support us marketing and selling the product outside of the app stores. No surprise, there will be a heavy mobile component to this app and we look forward to offering pay options through Apple’s, Amazon’s, Google’s, Microsoft’s and other stores.

But in no way will I restrict myself to those channels. The opportunity is too great.

Four years ago we put our first app in Apple’s App Store, a product called FastFigures. This was our first experiment and, for a short while, it did quite well. This were back in the days when Apple looked at your name, description, ratings and downloads to rank your app. Since we had an app that did lots of stuff — everything from finance and business to conversions and scientific functions — and we discussed that in our description, we ranked very high in a lot of searches. We tried a number of different price points, finding that $5.99 was the right combination of units and revenues to generate the most income. We also tried $9.99, $4.99 and $1.99, where $1.99 was the best at getting us high in our category. We were ranked #2 in Finance at that price.

But then Apple made a change. They stopped using the description and started using keywords and title only. Keywords were restricted to 100 characters (including spaces and commas to separate them). While the title seemed to weigh higher, the keywords and title were mixed together to create a pool of words with which Apple used to rank us for search. This change caused us to drop from #2 in our category out of the top 100, pretty much killing our sales in a matter of weeks.

Because of what happened here, we decided to revert back to our original brand and released a new version as powerOne. We were smarter here. We did a bunch of research, figuring out keywords to use, making adjustments with each release. For the last couple of years we could count on a pretty consistent volume of sales day in and day out, allowing us to plan for the future with a consistent base of sales.

Sometime over the last few days, all of that was thrown out the window again as Apple changed their algorithms. We pay the closest attention to our powerOne Finance Pro product, which traditionally generates 50%+ of our sales. We went from top 25-28 on iPhone Finance category with our primary paid product, powerOne Finance Pro, to top 50-75. (Interestingly, we didn’t change drastically on the iPad.) Our sales dropped across our entire product line nearly 30%.

The reports coming in from others (see here, here, and here for a few of them) believe that, among other changes like increasingly emphasizing downloads and ratings, the biggest change is that Apple is now treating titles as titles and keywords as keywords, not mixing the two.

My data proves this out. Every one of our keywords were specifically chosen and honed over the past two years but now we don’t appear for a number of them. Key phrases like “mortgage calculator,” “loan calculator,” “business calculator,” and “real estate calculator” do not show our products at all. Each of the first words are in our keyword list but we omitted “calculator” since it is in our title. As mentioned the old rules didn’t require it to be in both places. One example that proves this change is our keyword “rpn.” We rank 10 and 8, iPhone and iPad respectively, for the term alone. But if you change it to “rpn calculator” we don’t rank at all. Another example is “mortgage”. No rating for “mortgage calculator” but “mortgage” alone ranks us 14 and 5.

We do, however, still show for “finance calculator” and “financial calculator,” which is very interesting. We have traditionally had to include related words like this. However, we do not have “financial” anywhere in our title or keywords, which means Apple is now grouping related or similar words. This is a great improvement.

In the end these changes (and I’m sure others to come) will be a huge improvement in finding the app for which you are looking. Given that, every one of these changes means massive changes for developers.

Now I get to fret about sales for the next few weeks as I wait to see the ramifications of these changes. I also have to divert attention to a new set of powerOne releases so I can update the keywords. (You can only change app name and keywords on a new product release.) Luckily, it is a simple fix, though, as adding “calculator” to our keywords should solve our problems.

Given all that, the worst part is it feels like I can never count on revenues for long in this world. But that will be a discussion for another day.