Bringing Technologies To Mobile Applications

Editor’s note: GD (Ram) Ramkumar is a serial entrepreneur and computer scientist. He was founder and CTO of SnapTell (acquired by Amazon in 2009) and is now the Founder and CEO of Concept.io, a new mobile startup. He holds a Ph.D. in Computer Science from Stanford.

I started as a mobile entrepreneur in the pre-iPhone era in 2006 as the founder of SnapTell, the first successful mobile app in the image recognition space. SnapTell was acquired by Amazon’s subsidiary A9 in 2009. In 2011, I left Amazon to join Charles River Ventures to start a new company, Concept.io, which launches later this year. I reflected on lessons learned before embarking on the new venture and wanted to share them with the community. This article shares lessons I learned and discusses mobile trends that have emerged since.

The Key Lesson: Choose a problem and frame it well

Our first product at SnapTell was a service that allowed consumers to send in a photo of a shelf tag in a store for comparison shopping. The service read text from the label using OCR, determined product and price, and returned product and price information to the user. The solution was slow and cumbersome. It only worked with high resolution images on high-end auto-focus camera phones. Only a few such camera phones existed in the market in those days, sold by Nokia, Sony Ericsson, and HTC.

We learned quickly that this was not the right problem framing. We found that it is better and more intuitive to recognize a product cover instead of a shelf tag. Technology called ‘image matching’ was shown to work in research labs to compare a camera phone image to a set of thousands of images. We set out to create a scaled version that works for millions of images. This time the problem was well framed. Over the next year, our team worked hard to create a solution that emerged soon after the iPhone App Store launched in 2008. The resulting product was the first iPhone app that allowed users to search for media items visually using pictures. After SnapTell’s acquisition, the technology was deployed in Amazon’s mobile applications.

We set out to create a scaled version that works for millions of images. This time the problem was well framed. Over the next year, our team worked hard to create a solution that emerged soon after the iPhone App Store launched in 2008. The resulting product was the first iPhone app that allowed users to search for media items visually using pictures. After SnapTell’s acquisition, the technology was deployed in Amazon’s mobile applications.

I learned some lessons out of the experience.

For a mobile app or product to turn into a sustainable business, it must support a daily use case that turns into a habit. Comparison shopping didn’t provide a frequent use-case for most consumers.
Successful apps must create a “wow” experience and bond users emotionally. Having a great UX designer on board early is essential to a successful startup product.
It is important to identify a problem and focus on creating a solution for the consumer, rather than build technology and look for applications.

Emerging Mobile App Opportunities

The mobile ecosystem today is dominated by the app marketplaces of iOS and Android. Many new opportunities exist on these platforms, particularly to build “always-on services.” Such services leverage strong API support in iOS and Android for sensors and wake-up mechanisms to serve the user contextually. Background location streams offer powerful semantic data. Background voice-over-IP and audio APIs support new use cases. Consumers that use such a service do not need to remember to launch the app. The service knows when it is useful and chimes in automatically.

The user goes about their daily life as they normally do. Mobile service apps of the future wake up and serve the user in context. For example, such services may (1) Let you know the dry cleaner is open and clothes are ready and available as you drive home; or (2) Tell you about traffic where you are headed for your commute; or (3) At a convenient time, let you know that a friend you wanted to speak to is available right now for a call. Early examples of always on services are the newly announced Google Now product, the iOS Reminders app and Highlight.. It’s still early days with many services to come in the future!

Advances in Mobile OS

New features and APIs on the mobile platforms often make room for innovative services. iOS 6 announced the ability for Siri to launch apps (with the promise of more to come), tighter integration of Facebook, more efficient access to Bluetooth, and support for Passbook, a ticket and payments organizer. As an example, apps will emerge that integrate with Passbook to facilitate smoother travel.

Conclusion

Despite our collective focus on technology and design, social science and human behavior remain critically important. Understanding consumer behavior and applying technology to serve users in context will be a theme for many mobile services to come. Creating a “wow experience” and a positive emotional response in the user is very important. Siri has been popular in part because it goes to great lengths to answer fun questions such as “what is the meaning of life?” Technology exists to serve humans, not vice versa, and this will play out in interesting ways as the mobile ecosystem continues to evolve.