gpsAssassin Analysis: Week 1

As we prepared to launch gpsAssassin, I can't tell you how grateful I was for all of the generous Apple Developers that were kind enough to publish their stats and observations. To continue in the spirit of generosity, I'm going to post our stats and observations from time to time.

Time line info:

05/08/09 - Submitted to App Store
** Note: We saw an apple tester create an account within 8 days and play for approximately 5 minutes. We know it was Apple because we entered their latitude and longitude on Google Maps - revealing their location to be Apple Headquarters :) After the initial review, we saw no further activity until we were approved.



06/04/09 - App Approved
** We made a rookie mistake and had the 'release' date set to May 31st when we initially submitted. Based on external feedback, we thought the app would be reviewed within a week and felt the 31st gave us plenty of time to prepare for launch. Unfortunately, it was approved on June 4th and since our release date was in the past, it became immediately available for download. LUCKILY, we set the initial price at $4.99 and that discouraged the masses from downloading the app before it was 100% ready.

06/05/09 - First Sales from our Beta Testers
** Our beta testers helped us clock our first sales and test the store-delivered application. BETA TESTERS ARE CRUCIAL to your success. Not only did they help us develop the game, they also helped us find hundreds of bugs. Its been a week and the beta tester are still reporting bugs - which is great since most unknown users won't give you a chance to fix and will only give you a bad rating.

06/07/09 - According to Flurry, New Users surpass 150+ by Noon
** At first, you want to believe that your application is so amazing that everyone and their brother is talking about it - thus driving new sales. But after a deep breath, you realize that going from 10 sales the day prior to 30+ sales per hour indicates something else. We were watching the data and it seemed new users were popping up in Europe, Africa, Asia, Canada, South America, and Australia. It was then that we knew we were hacked.

By that evening, we had surpassed 600 new users - far exceeding the number of sales we were showing in the iTunes data portal. Our efforts to scan the web paid off and we found the app was hacked and released. The irony: We planned on reducing the application to 'free' within the next 2 weeks - after we had enough time to clean up the app for the masses. So in this scenario, the pirating community actually helped us spread the app like wildfire - something future Developers may want to consider.

06/09/09 - Here's what the data tells us

Day-over-Day Growth (as compared to previous day)






** While not precise, I can infer that hacked software is most desirable the day it is released and then quickly tapers off. I'm guessing it will continue its rapid decrease until the pirated versions only represent a handful of new users per day. If we gain mainstream popularity or media coverage, the pirated copies may increase accordingly.

Market Research for our 'Free Version'
** Since day one, our plan has been to release gpsAssassin as 'free' and monetize through in-game transactions & related in-game marketing/advertising. The $4.99 price was temporary and allowed us to collect revenue for a launch party. Surprisingly, some people still bought the app and we were able to watch their game play for the first two days. Their gameplay was similar to the other beta players in terms of frequency and length. In contrast, the usage profile for the players of the pirated version was much different.






Players who purchase the game want to play it a lot - thus the average user playing 9.24 times per day. Paying money increases loyalty/intensity of gameplay. In contrast, the player-session averages dropped dramatically once it was hacked/free. HOWEVER, its interesting to note that it levels off around 4x per day. After months of testing, our beta testers maintained a daily session average of 4.64. This leads me to believe that despite the significant drop in active users, by day 3 the free/hacked players were already playing the game at a normal pace.

In summary, it seems that the price of your app will directly affect the number of long term players you capture (common sense). Players that pay are much more likely to become loyal and will show an intense initial game play. Free players aren't vested in the app, so you'll see many of them abandon the app before they become loyal. However, once a player becomes engaged - it seems that free vs. paid doesn't matter. Common sense would indicate that all engaged players simply judge the app on its merits once they surpass the initial 'try out' period.

Other Observations
** To me, the most fascinating part of the 'hack' version is that now we have seed-groups in foreign countries. Our application is social and relies heavily on 'seeding' new areas to gain foothold. Marketing in countries like the UK, China, or Japan would have been cost prohibited. Now, with the hacked version in the wild, we have seed groups on every continent. This trend may continue once we make the app free.

Another observation relates to the psychology of success. Inevitably, people will ask you how your application is doing. Most, if not all, want to know how many times it has been downloaded. The subject of money is often taboo, so you'll rarely (if ever) have someone ask you how much money you've made. Since downloads are often the benchmark of success, a hacked version of your application can bode well for you.