“If you don't get noticed, you don't have anything. You just have to be noticed, but the art is in getting noticed naturally, without screaming or without tricks.”
The astounding growth of the mobile sector has attracted the attention of many entrepreneurs, particularly when combined with its low market-entry cost for developers and growing list of enviable success stories. For anyone with a mobile application to sell, Apple’s App Store and the Android Market provide easy access to the world market. However, this market accessibility and endless opportunities increase competition and challenge the entrepreneur to stay focused. This article presents some initial observations and experiences from the first year of Anystone Technologies, a mobile applications startup, as it faced the challenges of starting a new business in this attractive but highly competitive sector.
Mobile opportunities abound. Every day a new “app” is created that solves a problem and every day another new problem is created for an app to solve. But, how can mobile application developers turn their talents into a successful business?
Many entrepreneurs are attracted to the mobile applications sector because it requires very little capital investment to get started. Pay Apple $99 and you suddenly have access to a worldwide distribution channel, a platform that leverages cutting-edge hardware, and free quality assurance. With an abundance of open source applications and online services that are free of monetary cost to small companies, a studio can quickly (and cheaply) equip itself with software and services for documents, graphics, audio, issue tracking, file sharing, web page hosting, and more. Of course, an initial investment in mobile hardware is required, but startups that are interested in mobile applications tend to have these already. Further, with a strong Internet presence and inexpensive phone services that provide 800 numbers and an automated receptionist, startups can appear to be a larger company long before the revenue begins to flow in.
However, with over 500,000 apps in Apple’s App Store, getting noticed is no small feat. For a bootstrapped startup that has little marketing budget, the challenge of generating substantial revenue is great.
In this article, we reflect upon the first year of Anystone Technologies, a mobile applications startup in the Lead to Win ecosystem. Anystone Technologies was founded in 2010 by Tony Wacheski (CEO) and Sean Kormilo (CTO) to create a system of mobile applications that use speech recognition to make it fun for children to learn and love to read. While the founding team had decades of combined research and development experience in corporate telecommunications, we were new to entrepreneurship and the mobile applications sector. Here, we share our experiences with our first releases and highlight key observations and lessons learned along the way.
Our First App
While our ultimate goal was to develop a line of educational applications using speech recognition technology, we recognized that this would not happen overnight. Our initial strategy was to develop a simple application through which we could learn the process of submitting to Apple’s App Store and upon which we could begin to build our brand.
We knew that significant revenue would take time to develop, but we wanted to quickly familiarize ourselves with the existing revenue mechanisms that were provided through the App Store. At the time, “in-app” advertising was new and was showing great promise. We decided to create a simple game that would be quick to develop and would incorporate advertisements into the game play.
From this initial concept, further ideas flowed. We developed, played, and reworked until simple became complicated. However, we had not yet created a game that we wanted to play, despite enthusiastic interest in the concept from others. It was time to take a step back and re-evaluate our objectives.
Releasing our first application would provide invaluable experience and a revenue stream from advertising. However, we realized that this particular game concept did not align with our original company objectives of delivering products that make a positive difference in people’s lives. Our skills had increased in mobile application development, we were now registered with all the advertising networks, and we proved to ourselves that we could create a high-quality application, but we decided that this particular application would not be our first release.
Although our first application had been shelved, we learned that:
Great ideas may not translate into the application as expected. Mobile application development is an art. It can require many iterations of trial and error to transform a great idea into a great application. It is better to fail fast, learn, and move on quickly.
Refining game play is very time consuming. An additive game requires the right balance of challenge, reward, and fun, which can only be verified by hours and hours of play. We now know how important it is to build a team of testers who are interested in our project.
Counsel from our advisor to “stay focused on building your brand” was easily ignored. The excitement and enthusiasm in the early stages of design is very enticing.
The promise of mobile advertising was kept: it is now set to grow faster than web-based advertising on PCs and laptops.
Our First Released App
Our second attempt at a first release was much more successful. This time, it was five days from conception to submission to the App Store. The application was unique, it was cute, it helped people, and it was intended to start building the brand for our reading application. The application, Tuto’s Nite Light, helps parents transform their iPhones into night light sleep promoter for young children, who would watch the owl character (Tuto) gently fall asleep as the screen slowly dimmed in step with a timer set by a parent.
Initially, we offered Tuto’s Nite Light for 99 cents in the App Store. Surely, parents would gladly pay 99 cents for even one smooth bedtime! However, the predominant sales model in the App Store includes a free or “lite” edition of the application with limited capabilities. Consumers are reluctant to pay for an application without trying it first. As Chris Anderson indicates in his book, Free, making an application free removes a powerful mental barrier from potential purchasers.
After limited success at the 99 cent price point and some experience from our second released application (described below), we later developed a free, or “lite” edition of Tuto’s Nite Light that contained a subset of capabilities from a new and updated paid version. The free edition would point the consumer to the paid edition, but we decided not to include advertising or in-app purchase features in the free edition because it is intended for small children. This new sales model resulted in a steady downloads and increased sales.
We also released Tuto’s Nite Light on the Android platform, an open source software stack for mobile devices, developed through the Android Open Source Project (AOSP), which is led by Google. Android runs on many hardware platforms, which means that developers have to cover many design and testing permutations. In contrast, the Apple ecosystem, where Apple controls the entire vertical slice, limits the hardware variants tremendously. However, we found that Android does provide some good mechanisms to handle the many platform configurations.
With Tuto’s Nite Light, we stuck to our plan of keeping it simple and gained the experience that we sought. We also learned that:
There are differences between developing for Apple and Android. For example, Apple tests and accepts or rejects applications submitted to the App Store based on their own criteria. (For example, our first free edition of Tuto’s Nite Light was rejected because it included too many explicitly disabled features showing what was available in the paid edition.) There are no alternative markets. You must follow Apple’s rules, but you receive free validation and some confidence your application will work for all users. The Android Market does not screen submissions. You can sell an application that crashes on launch. The Amazon Appstore is an alternative Android Market that does validate applications before releasing them. Another difference is that the App Store does not allow you to contact your customers and provides limited visibility of deployment details. The Google Market allows you to send customer email messages (through a proxy) and provide customers refunds. The Android Market provides some detailed statistics about users, including platform version, device, country, and language.
Price changes are another interesting tool that can influence download rates and sales. Setting Tuto’s Nite Light free for a day resulted in hundreds of downloads. There are sites and applications dedicated to discovering price drops, which means that a price reduction can be more than just a means to convert potential customers; it becomes a publicity tool. Similarly, word of mouth and social networking are key marketing tools for mobile application startups. Giving away applications to seed these conversations is one of the few free marketing tools available. We received the following review from a customer who downloaded Tuto’s Nite Light when it was free for a day: “simply beautiful! - Beautifully simple very usable timer I love their other app so I had to check out this one. It was such a useful gift I decided to upgrade the other (free) app ‘Anytune’ to the paid version to give a little back!”
Know when to say when. We have several more ideas for additional functionality and marketing campaigns for Tuto’s Nite Light. However, while we had thousands of free downloads, the additional revenue would not have justified significantly more effort even if we converted 100% of these users into paying customers. We needed to focus on our other projects that were already underway.
Our Second Released App
For our next application, we decided to focus on increasing our expertise with audio, which we knew we would need for our longer-term plans. Our second release was Anytune, an audio application designed to help musicians learn to play or transcribe songs by allowing them to slow down the tempo, adjust the pitch, and repeat loops. Guitarists, for example, instantly loved the idea. We released Anytune as a free application with advertising and gave the user the option of removing advertising through in-app purchase.
We received a much better response with this application: steady downloads and low but steady sales. Releasing new versions proved to be our most effective way to increase sales. We have released new versions quickly and have added new capabilities in each one. The richness of Anytune’s functionality is one of its competitive advantages, along with the elegance of its design.
Over time, we noticed that customers appear to be more comfortable purchasing a full edition outright than purchasing upgrading to a full edition through in-app purchase, as we described in a recent blog post, “Musings On In App Purchase”. We released Anytune Pro to allow customers to buy the full functionality directly as an alternative to the in-app purchase.
In a later version, we included support to share through Twitter, Facebook, email, and SMS. Allowing users to tweet their favourite pitch/tempo setting to their followers, for example, provides us with free advertising by encouraging users to tell others about Anytune. To reduce the effort of adding these features, we used ShareKit, an open source software package licensed under the MIT License. Anytune also uses the SoundTouch open source libraries under the GNU Lesser General Public License (LGPL, v2.1), with special dispensation since we could not dynamically link in iOS, Apple’s mobile operating system.
In our most recent release, Anytune became “universal”, meaning that it is optimized for the iPad as well as the iPhone and iPod. This update required us to rework and improve the user interface, but it now gives Anytune a better position in the App Store search results. There are several established applications in direct competition with Anytune but few are optimized for the iPad. We also raised our price from $4.99 to $7.99. The improved design and higher pricing has increased our revenue from this application.
From releases of our second application, we learned that:
Considering all form factors in the initial design will influence the development and avoid rework. When we optimized for the iPad, we reworked and improved the iPhone interface. Having an application that is optimized for the iPad has had a significant effect on sales.
Purchasing behaviour should be monitored closely. We have observed more sales through purchases of the separate, fully paid edition of Anytune Pro than through in-app purchases in the free edition, even though they are functionally equivalent and we expect most users download the free edition before buying.
Revenue from advertising is increasing but is mostly negligible. For this revenue stream to make a significant impact, we need to substantially increase our download rate. The advertising in the free edition is primarily to “encourage” users to buy the Pro upgrade, which removes advertising.
Good customer reviews encourage new downloads, but are somewhat beyond your control. We were delighted that our first ten reviews for Anytune were all five-stars. However, we later learned that reviews can be a double-edged sword when we received a one-star review and had no recourse to help the customer or gather information about the problem that prompted the poor review. For details, see our blog entry, entitled “There is a Support Button”.
Anystone Store Kit
Alongside the development of our first applications, we attempted to refine and improve upon our marketing strategy. Marketing an App Store application means trying to get noticed in a very saturated market. Startups depend on reviews of their applications on popular review sites or mentions on popular blogs and news sites. Anytune has been reviewed by several sites and was recently named one of the “Best new iPhone music apps: July 2011” by Product Reviews. We did notice an increase in downloads but no appreciable jump in sales corresponding to the time of this latest review.
Providing influencers (reviewers and bloggers) with early access to full functionality is essential to gain access to these essential marketing tools. Review sites welcome requests for reviews and always ask for a “promo code”. Promo codes can be used only once to download the paid version of an application for free. Although Apple allows promo codes for downloading applications, it does not provide this service for in-app purchases. Since our original release of Anytune only had a free download and in-app purchase feature, we could not provide promo codes to allow influencers to access the full functionality.
The lack of promo codes for in-app purchases and some challenges with Store Kit, Apple’s framework for handling in-app purchases, caused us to consider alternative modules. Apple’s Store Kit supports in-app purchases, but we found it to be overly complicated and difficult to provide a magical user experience due to lack of feedback mechanisms. After identifying a gap in the desired functionality, we decided to create our own Store Kit.
We decided to create an open source project to deliver the Anystone Store Kit, which is now available on GitHub under an MIT license. We hoped an open source project would provide some visibility for our company. From what we learned developing our first application, we wanted to enlist the crowd to validate requirements and test our software.
In addition to enabling promo codes for in-app purchases, the Anystone Store Kit included general usability improvements, improved error handling, and added-value features. Also, the library could be used independently or in combination with server functionality. We used Google’s App Engine for the server side, with the possibility of providing this functionality as a service to other developers for a fee.
In our experience with Anytune, we have observed more sales through the purchase of the fully paid edition than upgrades from the lite edition to the Pro features through the in-app purchase mechanism, even though they are functionally equivalent. This does not remove the need for the Anystone StoreKit, which supports the full range of in-app purchase options used by the “freemium” model, including lite-edition upgrades (as used in Anytune), add-ons, gambling, pay-to-progress and subscription.
The Anystone StoreKit project was funded in part by Coral CEA, a not-for-profit open innovation network (accelerator) that is based in Ottawa. Coral CEA agreed to fund part of the project because of its potential value to its members. Details of the research, references, and table stake requirements can be found in our blog entry, “Introducing the Anystone Store Kit”.
Shortly after releasing the Anystone Store Kit, other iOS developers with similar issues noticed our blog. The Anystone Store Kit had its first user and contributor. We were realizing the benefits of an open-source project. Early adopters were validating requirements, finding bugs, contributing to the code base, incorporating the code into their own applications, and asking for more. It was clear that others shared our pain, and our solution was needed.
The Anystone Store Kit is still under development, but the base capabilities have already been deployed in Anytune and Warp Plus. Through this project, we have learned that:
A useful open source project will quickly provide early requirement validation, testing, and contributions. We received some excellent feedback, bug catches, and a security module from our contributors - all from a single blog and the use of GitHub.
Example code sometimes becomes product. One of our first users incorporated the example storefront code into their product, including a prominent attribution in their application. This provided us some of the visibility we were seeking when we started the project.
While we learned through our initial releases how to operate in the mobile market, our original objective was still being worked on in the background. However, in the end, our research revealed that our resources were insufficient to successfully compete against the large players in our domain and our vision of creating a system of mobile applications that use speech recognition to make it fun for children to learn and love to read was shelved. Nevertheless, simply having a longer-term strategy was important because it brought focus to our decision-making processes and forced us to constantly re-evaluate the steps we were taking to build our brand.
We have plans to continue development and marketing efforts for our existing applications, including:
Improving our applications in response to user requests, such as increasing the performance of Anytune’s time-stretching algorithm and adding further sharing capabilities.
Creating promotional videos to effectively communicate the value of our applications.
Connecting directly with the user communities of our products.
Also, we are now working towards an ambitious venture that uniquely leverages the latest mobile technologies to engage people around the world.
While we have had some commercial success with our first applications, the real value of the first year of Anystone has been the experiences gained through this exciting learning adventure that immersed us in the very dynamic and growing world of mobile applications and entrepreneurship. The more we learned, the more we discovered we did not know. The challenges are great, but the entrepreneur community in Ottawa is vibrant and willing to help. There is nothing we would rather be doing.