App Stores, Then and Now, a Distribution Evolution

The AppStore to some, seems like a new phenomenon, but Electronic Sales Distributors (ESD's) have been around a long time. Selling software has evolved tremendously over the years and has helped make sales easier to complete. In the case of the AppStore, distribution has pretty much been taken out of the equation in how it can negatively affect sales. We've been writing mobile software since 1999, and writing it as Creative Algorithms since 2003. This post will explain how distribution and sales have evolved over the past decade, through our personal experiences.

On a side note, this is my 10th #idevblogaday post, and, until the new design goes into effect, my last one for a bit. I've enjoyed the challenge of writing every two weeks, but am also looking forward to a brief break. Of course, I'll be writing on my blog as I find things I want to share, so I'm not ending my blog, just my regular stint on iDevBlogADay.

App stores, electronic software distribution, and such, are new phenomena, cropping up as the internet took hold. In the past, only large software houses or publishers, could sell software easily. They could afford to build the inventory of discs and CDROMS, in boxed packaging, and distribute them to the shelves of brick and mortar stores, such as BestBuy. Back in 1999, we released our first mobile app, Date Wheel, as freeware, for the Palm Pilot. We would have loved to sell the app, but there was no good way to do so. Instead, we put our mailing address in the About page, and asked people to send a donation if they liked the software. Needless to say, the checks were not pouring in—I don't recall receiving even one. But it was exciting to see your products in reality and I beamed Date Wheel to many program manager colleagues in automotive.

Eventually, we listed Date Wheel as freeware on PalmGearHQ, one of the first Electronic Sales Distributors (ESDs). We watched as our downloads hit the thousands. It was also quite exciting. When Palm released a major update to their OS that broke all past apps (something they did not commonly do), and the timing was right, we started Creative Algorithms and updated Date Wheel, releasing it's 2.0 version as “shareware”. With shareware, you offer a trial period and people purchase an unlock code. We had to design our own encryption method, which took more time than the whole app. We started selling the app on our website, using PayPal, a new way to collect small amounts from customers. Date Wheel retailed at $15. We also listed Date Wheel at the numerous free software archives around the web, such as Palm Boulevard, TuCows, EuroCool, Palm Pilot Archives, and Version Tracker, to name a few.

In order to increase sales, we decided to upload our new shareware onto some of the non-free distribution sites, such as PalmGearHQ, Handango, PDAGreen, Pdassi, and PDATopSoft. Each of these ESD's charged a commission for each sale. In addition, all ESD's required that you sell at the same price on all the ESD's and on your website, so you couldn't discount 20% (the cost of the commission) if people bought direct. People would download the free trial, then come back to these sites to make a purchase (or just make an impulse purchase). On our About screen, we listed all the sites where we listed our software, along with our own website. Customers could buy software on different sites than where downloaded, so some of the ESD's tried to stop this “leakage” with new rules.

We prepared to release the Date Wheel update on PalmGearHQ, who took 20% of the revenue for listing on their site, but discovered the very distasteful new rule in the process. All developers had to direct customers ONLY to the site where the app was downloaded. Customers could still buy from wherever, but they could only get info on where to buy from one location. The ESD's thought this was reasonable, but to developers, now we had to release separate builds for each ESD and our website. Beaming apps thru the IR port was also very common, so we were concerned that when the app “sold itself” that the “Buy it at PalmGear” About screen would also transfer, which was not palatable. So, we devised an algorithm that would strip this note after beaming, and revert back to our website as the place to buy. Many developers combated this new rule with additional gaming of the system as well.

To add insult to injury, some developers started to game the free listing sites, so every day they would re-upload their software, via scripts, and rise to the front pages of the listings. These older websites had been put on auto-pilot, with ads, so the owners could care less if the listings were fair, as long as the ad revenue continued to come in.

As selling Palm software started to wain, the ESD's consolidated and imposed more and more rules on the developers. However, by this time, if you wanted to make sales, you HAD to stick to the ESD's, since many of the free sites were hacked. It was frustrating and some developers even formed the ESD Union, which never took hold, or had any effect on the system. Motricity came along and bought PalmGear and PDAGreen (and Pocket Gear, which sold software for the Microsoft WindowsMobile platform). Handango picked up a few smaller houses. Two main ESD's emerged: Palm Gear and Handango. Because they had a duopoly, the developers suffered more.

Developers used to get email addresses of all customers who purchased products, so could amass a large base of subscribers to very effective newsletters that informed users of updates and new software releases. This was a boon for developers because they could upsell and cross sell software directly to their current customers and distribute directly through their website, at a lower cost. (Today this marketing technique is most effective, even on the AppStore, where devs can have people sign up/opt in for such news.) ESD's slowly ended this access, especially when they started to power Palm's software store. Palm seemed to be the most hurtful of its developers and wouldn't even allow the developers to link to help on their own websites for help. In fact, any mention of a developer's website was strictly forbidden.

ESD commissions started to go up. 20% became 25%, then 30%, then 40%. The Palm storefront took 50% from its developers, AFTER fees. ESD's could discount your apps at any time, up to 20%, without your permission. Handango started giving under the table, sales price 'discounts' to channels such as the telecoms, so the net revenue to developers could get as low as only 20% of the purchase price (80% commission!). All this was done with little transparency, so it was difficult to argue proper payment. Speaking of payments, the payments were also drawn out to nearly 60 days from purchase, and some sites just stopped paying developers. We still have outstanding payments from some of these sites.

In addition to commissions, the ESD's made money on the front page by selling advertising to the developers. Those who purchased these spots really increased their awareness, but the sales on the site rarely paid for the advert, since the commissions were so high and the apps had to direct back to the ESD. However, general awareness did increase, so net revenue could increase, but the multiplier wasn't very high.

A new ESD started up, Mobihand, who was much more friendly to developers. In fact, our website shopping cart still uses Mobihand (we moved off PayPal years ago). Mobihand gave developers back many of the benefits we once enjoyed and charged a much lower commission. But they were late in the game and the Apple AppStore still choked off one potential mobile app revenue stream for them. Not as agile as Mobihand, Motricity eventually sold off PocketGear and PalmGear to one buyer, then the new PocketGear combo actually purchased a dying Handango a few years later.

ESD's were historically only one part of the distribution problem. The installation and fulfillment of unlock keys was also more difficult. Wireless or on-device purchasing was not always possible. In order to browse, buy, install, and use software, you had to go through these steps:

1) Visit an ESD or developer's website from your computer.
2) Download a zip file to your PC or Mac (Mac support was sporadic).
3) Unzip the files, move the .prc (or equivalent) into the Palm desktop (or equivalent) software.
4) Plug in your mobile device to your computer.
5) Hot Sync the files to the mobile device.
6) Try the software for 30 days (and then it would stop working).
7) Remember where you downloaded it.
8) Return and purchase a "Registration key."
9) Wait for the email from the developer with the key (and hope it wasn't spam filtered).
10) Launch the software on your device.
11) Manually type in the (sometimes long) registration key code.
12) Now you were up and running.

These steps left many points at which customers would abandon the process. Some developers added self-installers to their zip files. Some developers included directions on how to purchase inside their software. When wifi came out for mobiles, and smartphones appeared, there were tricks where you could download the software wirelessly to your device, but it was not standard and didn't always work. Sales to download ratios were horrible. Plus the awareness of software was spotty--most mobile users didn't even know they could add third-party software apps.

Android and Blackberry methods of distribution, including Amazon, appear to be a hybrid of the old ESD's. In many cases, the app store is not installed on the device when people buy it, so the app awareness is still lower. New users to mobile are still left without a clue how to get new apps or how to install the app stores. Only as a byproduct, the success of the iOS AppStore has brought apps to the attention of these other platforms, so people do ask how to get apps. In addition, on some stores, such as Amazon, sales discounts can be given at any time, without the developer's input.

In the case of the AppStore, much of the distribution issues have been resolved. The only obvious negative to the AppStore is pricing, but why pricing raced to the bottom is a whole other discussion and only partially due to the improved distribution model. On the positive, awareness is high, due to Apple promoting the apps and AppStore. The effects of prime promotional spots have not changed. However, devs do not pay for these (and hopefully it never comes to that because the small devs will be shut out as the demand will drive the price sky high). Distribution and installation is streamlined—just buy with one click and the app is installed (also no trial periods). Terms for developers are 30% (though a lower rate would be nicer), and there are no hidden fees. Developers control the price for their products, so there are no hidden discounts to special interests or sales at the whim of Apple. Developers are allowed to mention/link to their own websites. Devs do not get emails for each sale (which protect customers' privacy), but developers can ask their customers to sign up on their own. All in all, the rise of the AppStore ESD model has greatly improved the experience for both the developer and the customer.