Last week we received an e-mail from Apple’s App review team notifying us that after 14 days of review, our latest update to iCombat Lite would not be accepted because of “inappropriate keywords.” The offending keyword was “wii tank,” and we had chosen this because many of our users have told us that our game reminded them of the tank mini game that is part of Wii Play.

While we knew not to use current iPhone app names as keywords, it had never in a million years occurred to us that “wii” might be problematic.  In Apple’s words, they “cannot post applications that contain irrelevant keywords in their search criteria” and suggested that “it would be appropriate to remove ‘wii tank.’”  Interesting since they: 1) had already approved our iCombat paid version with the same keyword, 2) have approved an app with Wii in the title, and 3) they had already rejected iCombat Lite two weeks prior for some other reason without mentioning any problems with the keywords.

What is so frustrating about this latest round of trivial rejections is that the app review “feedback” seems to always come on day 14 (at the earliest), and happens serially.  To give you an idea, we first submitted iCombat Lite update 1.1 on September 8th, 41 days ago!  After being rejected for an issue Apple reported with the code on September 22nd, and spending several days working on replicating the bug (which we never even managed to), we resubmitted the exact same keywords and code on September 28th.

On October 12th (14 days later) we received notice that the entire update would need to be resubmitted because of the “wii tank” keyword.  Had anything changed from the approved iCombat Paid version or the previously rejected lite version?  Nothing at all…and so we deleted the words, resubmitted and for the third time started another 14 day approval cycle.  All in all, if we are lucky we expect the iCombat Lite update to be approved on October 26th, just 48 days to get to market (42 if we subtract the days we took to work on the first rejection).

Reasons to Avoid Developing for the iPhone

Ignoring how illogical this last keyword rejection has been, the real damage of the current app approval process is that it has created a slow and arbitrary development environment that does nothing but discourage indie developers.  The biggest issues with this setup are:

1. Slower development cycles – As if figuring out what users wanted wasn’t hard enough, now add a 14 day approval delay which quickly turns into 1 month with any rejection and you have a buffer that really starts to isolate developers from their users and this constrains the feedback – iteration loop
2. Product/ Market fit is replaced with Product/ Apple fit – To use Andreesen’s advice (worth a read), entrepreneurs should “do whatever is required to achieve product/ market fit.”  Here the only thing that matters is finding what users want and giving it to them.  Yet with Apple as the gatekeeper success is not determined by the market but first by whether Apple will LET you play in its garden. This perverts the goals of the developer and ultimately reduces the chances that an efficient product/ market fit can occur.  You could even argue that on the off chance that you find an exceptional product/ market fit you are at even a higher risk of being cannibalized or pulled or copied by Apple itself.
3. Impossible ROI calculations – If you are trying to run a business based off of App development, how can you possibly calculate the return on your investment when you have no control over your launch to market? Unless you are ngmoco with funding from Kleiner Perkins then how can you build a business on top of such uncertainty (market and execution risks should be more than enough to contend with without having to worry about the d-bag app reviewer risk)
4. App approval amnesia and the lack of a fast track system – What seems to be happening all too often is that previously approved apps, after waiting weeks in the queue, get rejected for features that had already been approved in past releases.  This approval amnesia combined with being lumped in with new app approvals creates a developer disincentive to work on refining applications.  Does it make sense when iCombat Lite, having been live for 3 months with 100k installs and no complaints, suffers a 40+ day delay because it is being forced to the back of the line over and over again to wait amongst what is new crapware?  The sooner these apps can get updates out the sooner they can deliver high quality experiences to Apple users.

All of these factors serve to undermine developer confidence, reduce the quality of apps in the store, and ultimately choke app development activity.  Developers are already looking to other platforms and are limiting investment as the environment has simply become too unpredictable to work with.  Sure Apple has its reasons, namely pushing its 85k or 100k or 250k apps commercials to prove it has the most evolved app ecosystem versus its peers.  But if Apple doesn’t fix these problems soon, those numbers will begin to mean less and less, and at some point the number of apps in the App store will be about as meaningful as the number of videos uploaded to YouTube.

Related articles by Zemanta

• iPhone Apps: From Idea to Implementation (slideshare.net)
• Big Game Publishers Muscle In On iPhone Upstarts (wired.com)
• ngmoco Hires MySpace Executive Jason Oberfest to Helm Social Applications (wireless.ign.com)
• Shazam hits 50 million user mark and raises funding from Kleiner Perkins (thenextweb.com)

The biggest question with the change is the last sentence of the above, which could mean that the only way to change keywords for an app is to submit to Apple an updated version of the app for approval. This may not sound like a big deal but if indeed app updates are required it could open developers back up to the pschizophrenic app approval process and only further delay the approval process (currently at 12-14 days). Traditionally everything but the binary (app summary, price, icon image, etc) can be changed without resubmitting an application so that should be the case here.

The addition of keywords will be another filter separating actively developed versus abandoned apps (as was the 3.0 compatibility update) but it will simultaneously bring back the topic of keyword optimization. This keyword marketing strategy has been around for quite some time so we will see if Apple does anything to curb outright app name poaching. For example, a developer could choose to put the top 10 app names as keywords instead of words relating to the type of app being marketed.

Hopefully the additions of keywords will help discoverability but Apple needs to still do a better job of keeping developers in the loop. For anyone who takes app development seriously as a business, Apple’s poor communication only serves to further alienate developers from the platform. No one wants to be reminded they have a gatekeeper to their business, and Apple does a great job of making developers feel helpless almost any chance it gets.

Update: It looks like Apple has resolved some of the confusion around the keyword character limit as well as the keyword submission process.  As with all other app description features developers can change their keywords from within iTunes Connect without submitting a new build – this however only applies to the first time.  Once keywords are set, developers will have to resubmit their app to change them.  This clearly will not help with app approval delays.

Related articles by Zemanta

• Google Voice iPhone app rejected, current GV apps lose connection with iTunes (engadget.com)
• Google Voice app GV Mobile ported to jailbroken iPhones, web app version in the works (engadget.com)
• Watching an iPhone Feud on ‘The Daily Show’ (blogs.wsj.com)
• Spotify iPhone App (paidcontent.co.uk)

Conversion rates well above 5%

Most developers report conversion rates between 0.5% and 2%, and while Ethan Nicholas of iShoot fame at one point claimed (see comments) a conversion rates of 8% (200k sales off of 2.4MM lite downloads at the time) we didn’t think that was a reasonable expectation given the increased competition in the App store since (and in the gaming category more specifically).  And since there is no way of knowing where users are coming from we thought at minimum we should break out the conversion rates controlling for the previous month’s sales average.

I was surprised when I found our conversion rates in the control case close to 7% percent for U.S. sales.  Even when iCombat was clearing over 2.5k lite downloads a day our conversion rate was peaking at above 9%. While the early high conversion rate could be attributable in part to iCombat’s update that also occurred during the period, the averages trended steadily around this rate both before and after the update went live.  Aside from the new release it is worth noting that most other conditions were controlled for as there was no press, no features by Apple, no reviews, advertisements or anything else that might have spiked sales (a real marketing coup we know).

As for the big surge in monthly sales (5x), this is a bit skewed due to the low average volume of sales iCombat had experienced in the month prior to the launch of the lite version.

Some lessons learned:

1. Should have released lite version from the beginning – There was no point to waiting and sacrificing the initial new release buzz.  Since it is harder to get featured once your app is launched, say for app updates, it is important to strike early and hard with your app release.

2. Lite does NOT cannibalize sales – If your app is a gimmick then it might not make sense but in all other cases it only helps to increase sales (see our previous post on this topic)

3. Get the bugs out for your lite release – users churn lite apps and are fine giving you 1 star if they don’t like the experience.  This is especially bad because the App store prompts users to rate an app when they try to delete it

4. Lite sales trail off too but paid sales remain higher – if you don’t have the x-factor that is needed to spread the word your lite downloads will fall as they have for iCombat, but in our case paid sales have continued to sell at a minimum rate several times higher than the pre-lite period

5. Frequent releases do juice downloads – Pocket God and other frequently updated apps have benefited from a weekly sales bump as they show up in the new releases section of the app store (users also like this episode style model)

Related articles by Zemanta

• Google Voice apps yanked from iTunes App Store (geardiary.com)
• You: AT&T stores to sell MobileMe starting August 2nd (thenextweb.com)
• Apple Flip Flops on “Mature” iPhone App Policy (readwriteweb.com)
• How an Apple Tablet Could Pit iTunes Against Amazon.com (wired.com)

I came across an interesting post on “How to use Twitter, Facebook and YouTube to market your mobile games” and it reinforced what I have noticed since readying to launch iCombat – there are too many channels to manage when it comes to reaching the end user! This may sound like a high class problem to all of the game and content developers that have faced the gatekeepers of distribution over the years but the sudden blossoming of dozens of channels to reach users has created its own set of problems. Now the burden of brand management and marketing has fallen into the lap of the indie developers, where they must become “marketers” of their product if they hope to get noticed.

The rise of the fractured, multichannel market has created dozens of sites that developers now need to visit to build goodwill and help gain exposure, something that is fundamentally different from what they probably want to be doing. In our case this translates into 18-20 hours of work a day since launch to manage the marketing effort. While we have a dedicated team for this, most developers don’t have the luxury of even dedicating all of their own time to managing their launch.

I have compiled a list of some of the tools and sites I think necessary to run a solid app marketing effort:

Top Down channels – kissing the ring

1. Bloggers – aka the “digital influencers” – sites with major traffic like Gizmodo, Techcrunch, GigaOM
2. Targeted review sites – Touch Arcade, 148Apps, Pocket Gamer, etc. – we have counted 60+ serious ones in all

Bottom Up channels – connecting with your users

1. Facebook – both personal status and group page
2. Twitter
3. MySpace
4. App/ Development/ Gaming Forums - Touch Arcade, maybe iPhonedevSDK, etc.
5. Product website & blog – here you need to have a demo, maybe a news section and your blog
6. YouTube – a great way to communicate with avid gamers
7. iTunes App store summary

Analytics Tools – monitoring usage, downloads, buzz

1. User downloads – iTunes Connect – Heartbeat App or AppViz (super easy to use but less flexible than Heartbeat)
2. Usage behavior – Pinch Media – gives you uniques, geo data, version & device stats, as well as unique data by action in your App (very cool)
3. App ranking – AppRanking by Michael Dorn or Mobclix
4. Web traffic - Google Analytics or Wordpress Stats
5. Affiliate data – Linkshare – here if you are an Apple affiliate you can earn back 5% of your 30% cut that goes to Apple – and monitor clickthrough conversion data from your site
6. Brand management – ScoutLabs or Google Alerts – Scout doesn’t really seem to work for small apps or early on in a campaign as it is, for these smaller apps Google Alerts is a great basic filter

While this list is not comprehensive this is basically everything we are using for the marketing of iCombat. I didn’t realize just how many channels and tools this involved until trying to post an urgent piece of news. When iCombat was reviewed in Gizmodo’s weekly roundup of iPhone apps we quickly wanted to update every channel we were using.

Spreading the word for that first update went something like this: first update Twitter, then the Facebook status, then the Facebook “Fans of iCombat” group news section, then the iCombat website news page, then the Touch Arcade forum thread discussing iCombat, then update the app description in iTunes for the good review.

Once we had finished with the news updates, hours later, we started to comb the anaytics tools: first Pinch media, then user reviews in iTunes, then Mobclix to see if rankings had hit, then Wordpress stats to see website hits, then Linkshare to see the conversion rates, then Scout Labs to check for buzz, etc.

What we experienced was the effect of the new multichannel model where the developer has had to become a social media explorer, pushing into every nook and cranny of the social web. From gamer forums to niche blogs run by 12 year old app reviewers (they do exist), developers risk ignoring these communities at their own expense. With users fatigued by so much new product noise, it is easy for an unmonitored launch to go unnoticed by these users.

Not all products are good enough to sell themselves, and even if they are amazing, the system isn’t efficient yet at rewarding quality. So while I concede part of the problem is that most developers don’t have the time, especially amateur ones, to dedicate to marketing, I do think we all still need to shift our attention to just how important app marketing is. Traditionally, developers seem to look at rising to the top in the App store as primarily a question of the product. I would argue that in most cases now for the indie developer it is equal parts product and packaging (dare I say it might even be even more about the packaging?). This new balance should be weighed when estimating the time, cost and energy it will take to execute a new app idea – i.e. when laying out a development time line, product features and depth of marketing focus should be weighed as equally relevant factors (as each is a drain on your resources, be it time, money or energy).

In the App store the definition of what is “good” is all over the place, open to interpretation across the huge breadth of gaming, niche, utility, gimmicky and entertainment apps as well as age and demographic groups – so the opportunity to define what works is wide open for anyone willing to get down into the dirt and push. While it is fantastic that there is no system of patronage or well laid tracks in the app store in terms of rankings and exposure, it is a bit like the Wild West in that everyone is in a land grab racing for market position (whether it be in App reviews, app analytics, brand management, or the apps themselves). What’s clear is that the ones who will triumph aren’t those sitting waiting for the meritocracy to kick in, but rather those who act aggressively and start thinking creatively about how they can get noticed.

Related articles by Zemanta

• Google Voice iPhone app rejected, current GV apps lose connection with iTunes (engadget.com)
• Confessions of an iPhone game maker (macworld.com)
• Is mobile finally here? AdMob / Comscore, Forrester, CCS Insight and Lightspeed research suggests it might be (nickburcher.com)
• How to get ahead in the Apple App Store by mogeneration for mobile monday Sydney 04/05/2009 (slideshare.net)

- Raven Zachary, Founder Small Society

The 2009 Creativity and Technology event featured an excellent panel discussion on “What’s Now and What’s Next in Mobile” (for video see here).  While the title should have been what’s now and next for the iPhone, the truth is that this year mobile as been almost all about the iPhone.  And for good reason, the statistics are astonishing – with over 40+MM devices sold, 1.5Bn apps downloaded and a feverish growth rate that does not seem to be slowing down.

The most interesting part of the panel occurs when Raven Zachary, founder of the iPhone strategy firm Small Society begins to talk about the general trends in the app market (go to minute 12 in the video here).  His conclusions are bold and bright, pointing to the accelerating trend of brands and companies coming into the app store.  Zachary wisely suggests that brands though should be less concerned about speed to market and more concerned about a clean, quality 1.0 user experience as it ultimately ends up being far more impactful.  The balance of delivering utility or entertainment with branding, takes time to get right and too often the rush to be first is at the cost of a quality experience (see my rant about the prematurely launched Dunkin’ Run app here).

Perhaps the most insightful comment Zachary makes is that the “the app store is not a marketing vehicle but a distribution vehicle.”  Companies should approach the App store “as you would approach any product or marketing effort” and “drive people to the app in the same way you would drive people to anything you create in the company”.  The idea that applications can be developed overnight and with little additional marketing, product launch or strategic thought is outdated because the size of the app store makes organic discovery difficult.  And as the novelty effect of the iPhone wears off (where blowing and pouring are enough to drive downloads) the utility and entertainment components of apps will come to the fore.

Related articles by Zemanta

• Staying Afloat In a Sea of iPhone Apps (apple.slashdot.org)
• iPhone app buyers spend 50 percent more on weekends (venturebeat.com)
• Mobile phone applications grow up (news.bbc.co.uk)
• Wildly successful iPhone App Store hits its first year (computerworld.com)

The Pitch to Users

Dunkin’ Donuts had stumbled upon a unique opportunity, to further their customer engagement while actually delivering value to users through an iPhone and web application.  And the marketing team at Hill Holliday crafted a powerful narrative around the groundbreaking innovation.  Hill Holliday’s Chief Media Officer was quoted as saying the app “perfectly complements the ‘we’re in it together’ ethos of the Dunkin’ Donuts brand” and their site goes as far as to quote an internal employee who says that “Dunkin’ Donuts doesn’t just serve coffee, it also fuels the “hope and ambition and energy and drive” of America”. Wow.

The Negative Response

The problem with such bold positioning and a large launch hype-fest was that the app does not actually work from a user’s perspective. So the narrative comes across as a naive oversell and a typically “corporate” attempt at marketing.  What has followed has been UGLY – I have seen very few iPhone applications with such poor ratings and reviews.  Users are angry, at the time of writing this 225 of 376 (60%) ratings are ranked 1 out of 5 with 298 out of 376 being 3 or below (79%).  User rejection has been nearly unanimous. So why did the application fail?  Because its designers ignored the very thing that it was offering to users, “speed, accuracy and fun.”

“What Did I Do Wrong?”

To describe the problem all you have to do is look at the first screen input screen of the app.  Imagine you are a user who has read about this revolutionary app that will help make your life easier, and then you open it and you wonder what to do next.  You think to yourself, “Do I have a password? Maybe I can just enter one and it will work. Who knows?”  So you try and it says invalid password, and you look for clues as to what to do next but there are none.  Most users just deleted the app at this point (probably most of the 225 1-star ratings happened at this point through Apple’s ratings prompt that occurs upon app removal).  In 10 seconds Dunking Donuts had managed to anger and alienate some of its most loyal customers.

What is worse is that the web experience for those who do go the extra step is even more frustrating, where the problems can be summed up with this one review: “You have to register from your computer which is stupid.  Secondly, why does it need my address, phone number, and mother’s maiden name?  We are talking about a dough nut order here, not banking information.” Well said.

The shame here is that the application, once you get past the sign-up hurdles is actually nicely thought out.  This is not worth anything though if the most important step, the first one, is made so cumbersome that 60% of your users decide to delete your application and give up.

The Solution: Talk Less, Deliver More

The app store is all about user expectations.  And in a place where reviews and ratings come all too easily, you need to be careful that you actually deliver on your what you offer.  And while the Dunkin’ Run idea is a great one, it was oversold too soon and that only contributed to the user backlash.

The point is that you can deliver value and create a truly useful application for your users without risking your brand, and the way you do that is to launch something useful and complimentary to your business but first make sure it works!  Good app design has little to do with a huge project or marketing budget and more to do with listening to your users.  Whether it be through Beta testing before launch or iterative improvements after a soft launch, it is a fundamental step that can’t be ignored if you dare to claim you have created the next big thing.

To check out the latest ratings/ reviews for the Dunkin’ Run app press here to be directed to iTunes.

Update: The app has been updated to 1.1 and now includes a prompt at the bottom of the screen (shown above) that allows the user to sign-up.  Still the reviews appear to be very poor (as of 7/18 34 of 76 ratings are still a 1 with an average of 2.5) and the comments reflect that the app is still more work than it is worth.

Related articles by Zemanta

• iDunkin: Top 7 Apps to Make Life Easier (abcnews.go.com)
• Dunkin’ Run Automates Your Coffee Breaks on the Web and iPhone (mashable.com)
• Dunkin’ Run launches, assumes you want to carry 18 coffees back to the office (inquisitr.com)
• Latte with Fries? McDonald’s Takes Aim at Starbucks (time.com)

The Pocket Agent app implements its features well, and provides a much more efficient process for filing claims or staying in contact with State Farm than ever before.  As for whether you need to install it beforehand, the beauty of the iPhone App store is that it can be downloaded in just a minute or two in the event that something does happen.  Of course such a rare frequency of use application like this requires recall on the part of the user that such an app exists, but State Farm has done a good job so far of spreading the word to its customers.

While this is one of the more niche uses of the iPhone’s capabilities, it is representative of a larger trend hitting mobile where companies seek to enhance customer engagement by leveraging the powerful feature set of mobile devices such as the iPhone.  By offering solutions that utilize the touch screen, accelerometer, GPS and camera, companies can prove to their costumers that they are innovating and staying ahead of their competition.

See the slightly random but effective State Farm video summary below:

Related articles by Zemanta

• Apple kills Routesy app, my iPhone gets less useful (digital.venturebeat.com)
• Find My iPhone helps when AT&T’s insurance can’t (mobilecrunch.com)
• Who Exactly Is In Charge Of The App Store? Anyone? (techcrunch.com)

I think mock-up tools like Balsamiq are even more useful than making great specs and writing out usage scenario examples.  Not only does it help you better develop your idea but it also gives you the ability to share your ideas with other people instantly.  For example, if you are thinking about outsourcing development or are talking to other team members, Balsamiq will let you share your quickly thrown together mock-ups and convey clearly what you want to do.  And by visually demonstrating the flow of how you might want your app to work, not only will other people have a better idea of what you are looking for, but so will you.

Oftentimes we think that conceptually working through an idea is painful, much the same way writing or preparing a speech can be.  But sometimes having the right tools can make it that much easier to throw things together.  Give it a try, you might surprise yourself!

Related articles by Zemanta

• iPhone app buyers spend 50 percent more on weekends (venturebeat.com)
• More iPhone Apps for the Home Cook (thekitchn.com)
• $99 iPhone Rocks E-Book World (mediabistro.com)