When it comes to the Windows Phone 7 App Hub, we have seen the big money grabs, big number milestones, and read about 3rd party support from from popular app makers. At the same time, when looking at the Windows Phone 7 apps themselves, the speed they make it into the App Hub, the services they use, and the prices they sell for, there are some interesting points yet to be made. The data below represent the some of findings after watching the Windows Phone 7 App Hub for a little over a month.

App Release Rate

On average, Wednesdays looks to be the day your app will likely make it into the App Hub. After looking at daily totals, it is the day the highest number of apps are released each week. Perhaps this mid-week peak is indicative of some internal work flow or app release strategy. Then again app submissions from developers may be driving this trend. Either way, this release rate looks to be independent of the the European and North American product launches. One would think that prior to the release of Windows Phone 7, in its respective regions, there would be a rush to have App Hub “stocked” with new apps. To an extent, looking at the three day period before October 21st and November 8th, this appears to hold true as 235 and 225 apps were released during the respective periods.

Total Apps

By my count, the Thanksgiving holiday set back the progress of getting 3,000 apps in the App Hub by the end of last week. No apps were released between November 26th and November 28th. As a result I believe the App Hub did not reach the 3,000 app milestone, for the US Marketplace, until today. Screenshots from the website Marketplace Browser as well Bing Visual Search for Windows Phone 7 Apps as of late in the evening on November 30th support this claim as both photos show totals hovering just below 3,000 apps. Excepting the first two weeks, the growth in the total number of apps in the App Hub has been roughly linear over the past month. Initially growth started slowly but since then it has definitely picked up. For the time being, the growth appears to be constant.

Looking at growth, when comparing one week to another, there is a definite “hump” that characterizes the weeks on the graph. When looking at averages for each day of the week and plotting them as a stereotypical week, it becomes apparent why this is the case. An average week of app releases has a definite hump-like characteristic to it. As you can see again, Wednesday represents the peak day each week for apps to be released in into the App Hub. I should note that this chart and the ones above it try not to include apps that are no longer for sale in App Hub.

App Prices are Moving

While I have tracked apps in the App Hub, I noticed a few price changes. Generally, the price changes that I saw were those that lowered the price of apps. There were a few apps that did raise their price as well. Thanks to the way I catalogued apps, I noticed these price changes within a smaller sampling of the overall catalog of apps. While this only represents a small portion of the overall marketplace it is still an interesting trend to note. Considering the short period most of these apps have been in the App Hub this could simply be developers looking for a sweet spot to sell their apps. In particular this looks to be the case with Guitar Hero 5 as app reviews noted $9.99 to be too expensive. Currently, Guitar Hero sells for $6.99.

Something for the Road

While it is great to read other's findings and look at charts of their data, playing with that data yourself is even better. That is why I want to take these results to the cloud by putting them in a shared folder on our SkyDrive and sharing what I have been working on in the form of two separate spreadsheets. Below is a copy of the full spreadsheet:

Before you continue let me provide some context for these spreadsheets. Over the past month I tracked twenty-five attributes of just under 3,000 apps in the App Hub. Among other items, the data includes cost, app category, and services used by each app. Because it is such a daunting task, just over 800 apps have all twenty-five attributes filled out. At a minimum all apps include their name, cost, the date I first saw the app, and its rating at that time.

The first spreadsheet contains all the apps I have catalogued from the App Hub. In addition it includes the formulas and charts used for this post as well as Part 1. There were two times that I counted all the apps in the App Hub on the same day more apps were being released, 11/7 and 11/16. As a result it is possible I missed a few apps on those days. This is something I am double checking to make sure all apps are accounted for. This is the spreadsheet that contains a minimum of four points and a maximum of all twenty-five. It is possible there are spelling errors in app titles that have not been fully cataloged. The website Marketplace Browser has been helpful in helping me correct my overzealous typing. So far I used the site to help me with six apps that I could not find on my own because of spelling errors or app name changes.

The second spreadsheet is a limited list of just over 800 curated apps that have all twenty-five values filled. This list should be spelling error free, as I found the app and was able to look at its attributes. This curated list includes the top 30 free and top 30 paid apps. There are also some apps that are not in English. However, I am not sure this curated list can be seen as representative of the App Hub as a whole. Among other apps, it includes most of the Tip Calculators, Flashlight apps, and the first 300 apps in the App Hub. The reason I over-analyzed the first two app types is that it helped me identify apps with the same or similar names from different developers (e.g. apps called ‘Tip Calculator’ or ‘Flashlight’). I included the first 300 apps because my initial process was to catalog all twenty-five attributes before moving on to the next app. This was going slowly so I switched to a faster process that is the basis of the full spreadsheet. With this second spreadsheet you can answer questions such as how many $0.99 apps include a free trial (134 of 813 apps).

Apps with no ‘Description Language’ are in English while apps with no indication of whether they are ‘Ad-Supported’ are that way because I cannot tell from the services listed, app description, or preview pictures provided whether the app serves ads. The app ratings and number of reviews per app come from anywhere between the date the app was first seen to the date I last updated the spreadsheet. As a result, there may be some apps in this spreadsheet that have fewer reviews than their current total in the App Hub. Apps highlighted in red are apps that are no longer available. These apps are usually in hidden rows but using a filter can make them appear again. I will be updating the first spreadsheet more frequently than the second. Using the Pivot Table sheet provided in the first spreadsheet will probably give you similar results to the second spreadsheet.

I mentioned taking this to the cloud and I mean it. I am looking for some help. If you want to help fill in app attributes or know how to parse the XML entries that pertain to each app so that this data entry process can be automated send me a message (Twitter, Zune) so we can get to work. If not, feel free to download either spreadsheet and play around with the data. If you do find these spreadsheets useful or use the data provided within either spreadsheet, be sure to say thanks below and link back to this blog post.