Well That Didn't Last Long

Dominoes was knocked from the #1 paid slot in Cards & Casino after only a few days on top. A solitaire collection is currently #1, but Dominoes looks like it's going to hang in solid at #2. It's still selling very well (30-40/day). Dominoes is #50 overall in paid games. Comparably, the #16 in Arcade & Action is #30 on the overall list, which gives an indication of the better-selling categories. Golf Solitaire is hanging tight at #7 with 3-10 sales a day, so it doesn't take much to stay in the top 10 in that particular category.

I'll do a recap at the end of November, but it was a very good month. I'd love to see the numbers go up as more and more people buy Android phones, but I certainly would not complain if they stayed right where they were for a while.

Checkout Check

I'm just finishing up a new app that's probably the first one I've designed and built specifically for my own personal needs. It's called Checkout Check and it's an interface to Google Checkout that lets you generate simple sales reports. Here are some screens:

Here was the problem: If I want to see what my sales in the Android Market are for a given day or date range, I can log in either from the browser on my PC or on my G1. On the browser, the cookies make the login a two-step process. On the phone it's an additional step and each page that loads is quite slow. Either way, there is no way to sort sales for a given day based on the name of the app. You can export a summary to an .xsl file, but get this...you can't export orders with the name of the app!

Let's say I sold 45 apps on a given day and I want to record the sales. If I want to know how many copies of each app I sold, I have to load up the market interface in the browser and manually count them. And like I said, you can't even sort them by app name. As my sales continue to increase, this becomes increasingly tedious.

So I've developed an app that let's me select a date range and an order state and generate a simple report that details the name of each app, the number sold, the gross revenue, and net revenue, and totals for everything. It's kind of amazing to me that I'm not able to generate such a report through the official market interface.

And let me tell you, it wasn't easy using the Google Checkout API, either. There's a Notification API, which works in real time to notify you of any changes to your account, a Notification History API, which allows you to get information on past notifications, and an Order Report API, which generates reports on orders.

But here's the thing...from what I could tell, only new order notifications contain information regarding item name. So what I had to do was send a query to the server for all the new orders for a given time range, returning order number and item name, then send another query for an order report for the given time range, which contained order number, status (CHARGED, CANCELLED, etc.), and cross-index the two to generate a report with the information I wanted. Sounds fun, huh? Because who would actually want to generate a report with the name of the items they're selling, right?

Anyway, it works. Sort of. It's a bit slow, but right now it's reliably fetching sales information for a single day for charged transactions. I'm still having some issues with other order states, and I'm also getting errors when I search for time ranges over a day.

Another fun property of the Notification History API is that you can only fetch 50 orders at a time. If a query return is longer than 50 orders, what you get is an XML response with 50 orders and a next-page-token. You have to send another query using that token to get the next page, and so on for every 50 orders. I'm probably either getting a time-out error or an issue with memory for queries longer than a day, but hopefully I'll figure that out soon.

Anyway, I know I'll find it useful, and I hope other Android devs do as well. I'm sticking ads in it to hopefully generate a little revenue. If the ad revenue just isn't worth it, I'll likely just open-source it and see if other devs can help make it better. Also, right now it only supports transaction fee structures for Android merchants. Google takes 30% per transaction for Android apps, but the normal cut for Google Checkout is a tiered structure, depending on your monthly revenue. For non-Android transactions they take a small percentage and a flat fee for every transactions, and the fee gets smaller the larger your monthly transaction total. If there's enough interest and demand I'll probably augment the app to handle non-Android app transactions, but there's no need to do that right now. Besides, non-Android merchants can still use the app to get a summary of units sold and gross revenue.

Dominoes Hits #1!

Yep, I've now got the #1 paid app in Cards & Casino in the Android Market. People must love their dominoes. And revenue continues on an upward trend. I could get used to this.

I'm definitely expecting growth through the end of the year. Black Friday is coming up and I would expect another bump from discounts on the current crop of Android phones, as well as another around Christmas when gifted phones are fired up and new users start looking for apps.

Supposedly Verizon/Motorola has already sold around 250,000 Droids, and I saw a report yesterday that estimated 600,000 would be sold by the end of the year. Good stuff.

On the flip side, Golf Solitaire has slid down to #7, but I'm not going to complain.

T-Mobile Carrier Billing for Apps

Apparently T-Mobile had announced that starting Nov. 17th users would be able to purchase Android apps by easily charging them to their mobile bill. They also announced a custom T-Mobile channel in the market. It's Nov. 18th, and browsing the market and searching for T-Mobile, I don't see either of these functions. I have T-Mobile service, but clicking "Buy" for a paid app still takes me to Google Checkout, with no other payment options.

I was excited when I heard they were implementing this, hoping that more payment options would facilitate more sales, especially impulse purchases, and possibly fewer returns. It's not up by the announced date...I do hope they get it up and running soon.

And another quick update: Dominoes has just moved up to #2, but Golf Solitaire has dropped a slot to #6. Having an app in the #1 slot would be oh so sweet. Just one more to go.

Update

Well, a few days ago Golf Solitaire was #3 and Dominoes was #5. Now their positions are reversed, with Dominoes at #3 and GS at #5. Dominoes is still selling strong, making up about 30% of my revenue right now.

It's been a while since I've shared some hard numbers, so here they are in a pretty graph. This is net revenue by day for the past 6 weeks:


I had been steadily making about $50/day before the release of Dominoes, which gave me a very nice, and so far robust, increase.

As you can see, revenues had started to dip back down slightly before the release of the Droid on Nov. 6. According to reports, they sold 100,000 Droids in the first weekend after it went on sales, and about 250,000 in the first week. It looks like the Android market is poised for steady, sustained growth, mostly on the back of the Droid. Though the bulk of my users are still on the G1 and MyTouch, I'm seeing a steady rise of Cliq, Hero, Eris, and Droid users in my analytics.

My graphics probably look a little grainy on the Droid, though I have yet to confirm this...still wrestling with whether or not to get a Droid. The only confirmed technical issue was with Kilopod, my Centipede clone. I'd built it for trackball support, and hadn't built in controls for a d-pad, which is what comes on the Droid. A user alerted me to this and I fixed it within an hour.

And that's net revenue, after Google has taken their 30%. Revenue from ads makes up about 55% of the daily take, leaving about 45% to net sales. Ads have tended to make up about 50% of revenue since I started this.

In the works are a Google Checkout client for Android called Checkout Check (I'm not happy with the way Google Checkout exports data), a multiplayer version of Dominoes using Google App Engine as the server (Philip is working on this), a local version of Relativia which will be rebranded as Puzzle Lords, and a new game called Mobile Murder Mystery.

2 in the Top 5

Dominoes has edged it way into the #5 slot in Cards & Casino in the Android Market. Golf Solitaire is wedged in at #3. So I now have two games in the Top 5 in their category, which is very nice. And Dominoes shows no sign of slowing. It might very well pass GS this week and take the #1 (currently held by Ahoy Matey).

Meanwhile, InfusionCalc had dropped out of the Top 10 in Health, but I've had a nice trickle of sales lately, enough to nudge it back into the #10 slot.

I've noticed now that a lot of my niche apps (like Concrete Calculator) are selling in low, steady numbers again, probably due to the infusion of Droid users.

Update

Sales and ad revenue are still strong, so I'm happy. Golf Solitaire is still #3 in its category, and Dominoes has moved up one slot to #7, and is still selling briskly.

I came close to buckling down and buying a Droid and just going month-to-month on Verizon, but then I heard that there's a GSM version of the Droid called the Milestone that's released in Germany and will soon be released in Canada. See, T-Mobile, the carrier I'm on right now, uses GSM. Verizon uses CDMA. That means the phones don't run on each other's networks. However, even if the Milestone isn't officially released in the US, you should be able to use it on T-Mobile's network. So I'm going to wait a little bit and see if I can make that happen. That would really be the best option.

Gamers Apparently Don't Want to Roam

So after crapping out of the ADC2, I released a free, ad-supported version of Relativia on the Android Market. It's had about 500 downloads, with a subpar rating. Turns out that people don't really want to have to visit real-world locations to play a game...or at least not my game. The good news is that they do seem to generally like the puzzle combat system. Here are some comments from users:

Good idea but cant play because im not willing to travel around in real life to do so. Please fix!

Just make the game a world map. Battles are interesting though.

This is a good idea but is dumb that u have to go to real locations.

There are more like this. The general consensus seems to be that people like the game except for the part where you have to go to a coffee shop to enter a dungeon.

So, I need to think about retooling the game to just a local version, which would probably do well. In the meantime I've removed the current version from the market.

The Droid Bump

So Google and Verizon released the Droid on Friday, with a huge marketing push. And they must be selling the things pretty well, because a number of devs in the discussion groups are reporting an increase in downloads, in some cases up to 100% more than their daily average.

I'm not seeing that big an increase, but I am getting a bump. It's more in the 20-30% range, but it is significant.

I've seen +$100 net revenue every day since Friday, and Sunday was a record high. Here's hoping it keeps up!

The rankings in the market are weird. My Dominoes game (which has only been out about 3 weeks) is about to pass my Golf Solitaire game (which has been out about 6 months) in sales. They cost the same price. The only difference is the ratings...GS is about 4.5 stars, while Dominoes is nearer 4 stars. The retention rate for Dominoes is actually a couple of percentage points higher. I don't think anyone knows exactly how rankings are computed, but it has been speculated that it's likely a combination of downloads, ratings, and retention, hopefully with some way to filter out scamming. Meanwhile, Golf Solitaire is holding steady at #3 in its category, while Dominoes hasn't moved from #8 in about a week. It's going to be very strange if Dominoes surpasses GS in sales, has a higher retention rate, and is still ranked lower.

Anyway, it's very nice to have two apps in the top ten in their category, both selling well. And InfusionCalc has also seen a bump in sales and climbed back to #10 in its category after sliding out of the top 10. I was trying to find out if there was a review or mention of it somewhere, but I couldn't find anything new. It could just be that a lot of nurses and paramedics just bought Droids.

I was in a Verizon store yesterday and a guy in scrubs was buying a Droid, so who knows? Speaking of which, I probably won't be getting a Droid, at least not right now. Their price plan would have me paying $15 a more per month than what I'm paying with T-Mobile, which might be worth it to insure that I have 3G coverage here at home. The T-Mobile rep assured me that 3G would be in place in Lafayette by the end of November. Right now the network is a little slow, but it's fine for my purposes...email and very light browsing. It's like being back on dialup. Unfortunately, I was not impressed with either of the reps at either of the two Verizon outlets I visited this weekend. What I really wanted to see on the Droid was how well it browsed the internet, rendered video, and I wanted to handle the hardware keyboard. The guy at the mall had let the battery drain out of the only test model they had and didn't have a working power cord. Okay. At the other store, the rep let me handle a phone fresh out of the box, but didn't have one with an account configured, so again I couldn't test how it performed on the network.

So I'll just wait. The G1 seems outdated, even though I've had it less than a year. But it's still a solid phone, and not worth ditching for the current generation. The Droid is the nicest phone out right now, but for me it's not worth either paying +%500 for the phone and going month-to-month, or being locked into a contract for $15 a month more than what I'm paying now. Android phones are coming out fast, and I'd like to be off-contract so I could upgrade and potentially move between carriers whenever I want. The current generation of Android phones are all sporting around 500MHz processors, but I know there are 1GHz phones in the pipeline, and hopefully one of those will be released by the end of the year and will be compatible with my T-Mobile service, which will hopefully be upgraded to 3G in my area.

Until then I'm staying put, but I'm certainly happy that lots of other people are either switching carriers or upgrading to new Android phones.

Well Crap

I just got an email from Google notifying me that Relativia didn't make the cut to the second round of the ADC2. On the bright side, the email did inform me that my game was rated among the top 25% overall. That's something, at least.

I now need to figure out what I'm going to do with the app. Obviously I had hoped to place in the money and ride the wave of publicity to fame and fortune. The monetization plan for the app (besides winning prize money) was to form agreements with companies such that they would pay in order to have their business name incorporated into the game as a physical location to visit. I may still just release the game for free and pursue that option. Alternatively I could try to work ads into the game or try to sell it. I need to think about that.

No word on the actual finalists, yet. Someone in the Android Developers Group speculated that the winners have known for a while now, but were bound by an NDA. My guess is that this is probably the case. It would allow them to modify their entries using the 2.0 SDK for compatibility to coincide with the release of Droid tomorrow. I hope to be blown away by the top 20 in my category. That would make me feel less shitty about not moving on.

We'll see.

Google Implements Analytics for Android and iPhone

Yep. The SDK and documentation is here. This is pretty cool, but I'm currently using Flurry, and I'm extremely happy with their data and visualization. I may go ahead and use Google Analytics on my next app to see how things compare.

In other news, two sites that I used to rely on heavily to track the status of my apps have been down for a week. Cyrket is a site that provided an up-to-the-minute snapshot of the Android Market. It now directs to a blank page. Speculation in the dev forums indicates that Google may have changed something on their end that screws up the way Cyrket was scraping the information. Not sure if that's true or not, but I hope it's only temporary. In forum posts where people are asking about Cyrket, usually the first response is always, "just use AndroLib". AndroLib is okay. There are a couple of problems with it. One, it's much slower to update, as in days. Two, it doesn't display rank information just as the Android Market does. You can sort apps by date or by rating (which is kind of nice...the Android Market doesn't have this function), but you can't see a view of apps based on their current ranking (at least that I can tell). Want to find out what the #1 ranked app in a given category is? You'll have to actually go to the Android Market on your phone. Lame.

The other site that's currently down is AndroidStats. Actually, it's not completely down, it just hasn't updated with new data for a week. From what I've heard, AndroidStats gets its data via Cyrket, so since it went down, they both went down.

This wouldn't be a big deal if Google had a decent web interface to the market that displayed this information. One could possibly hope that the change they made might be related to implementing their own decent web interface to the market. We'll see. They've been pretty good about consistently updating Android itself, but not so much its web presence.

Another issue: I got an email today from a customer who had purchased Dominoes. She loved it, and wanted to give it as a gift to a friend. She asked that I simply charge her credit card again and send him a copy. Only problem is, I can't do that. And the Android Market doesn't currently support gifting apps. So I told her she could either buy him a gift card to let him buy the app on the market, or we could do an end around the Android Market via PayPal and email. Neither option is great. I have read that Google is considering adding this option. I really hope they do...it could only increase sales.

So, daily movements: Golf Solitaire has dropped back down to #4, but Dominoes keeps inching up...it's currently #8.

T-Mobile to Start Offering On-bill App Purchases This Month

This is welcome news:

T-Mobile will let its subscribers pay for Android applications on their monthly mobile bills starting Nov. 17, also introducing its own section of the Android Marketplace that day.

Very nice. One long-standing complaint with the Android Market has been the lack of payment options, only allowing users to buy apps through Google Checkout. Hopefully the on-bill purchases will be relatively streamlined, and it should only help with sales.

Speaking of which, Golf Solitaire has now moved up to #3 in paid apps in Cards & Casino, and Dominoes has moved up to #10 (and at the rate it's going, I would almost expect it to overtake GS).

October Update

Ah, October 2009 was a very good month. The combination of AdSense and Dominoes allowed me to make three times as much this month as my best previous month. I've now made more in the last two months than in the first six months I started publishing Android apps.

And the really good news is that things are actually picking up steam. The Motorola/Verizon Droid will be released this Friday, and it should do well, which should mean more downloads. Less well-hyped is the release of another Sprint Android phone today, the Samsung Moment.

In the meantime, Golf Solitaire has pushed up to #4 in paid apps in Cards & Casino, and Dominoes has vaulted up to #11. At this rate I wouldn't be surprised to see it in the top 5 by the end of the week, and it sure would be nice to have two top 5 apps in place for the release of Droid. Today I also released a new version of Dominoes with support for larger screens and for Spanish localization. I took Spanish as an undergrad, but it's not good enough to do translations. I relied on Google translate. Hopefully I won't get too many angry emails. At least if they're in Spanish I won't be able to really understand them.

Still...no word on the Android Developer Challenge 2. Today is a Monday, which seems like a good day for an announcement. Then again, so did the last three Mondays. We're a day shy of a month since they announced the close of round 1 and said it would only be "several days" before the finalists were announced and round 2 would start.

If they aren't going to announce the finalists this week, it would at least be nice to get some kind of update regarding the delay and an estimate about when we might hear something.