User Analysis
We have identified three demographics who would likely benefit using a mobile device to look up products while shopping: technological innovators, power users, and casual users. In this section, we describe these demographics. We discuss their problems and explain how our application may benefit them. Finally, to substantiate our claims, we located three research subjects and interviewed them about how a mobile product identification software will assist them in their lives. Our results are intriguing and suggest that our proposed mobile phone application can have a significant impact.
Technological Innovators
Overview: A compelling demographic are technological innovators and researchers who are likely to be early adopters of technology. People in this demographic typically use sites like Amazon.com and Newegg.com to find products and read their reviews. These users have grown accustomed to quick access to reviews and product information while shopping. In other words, they desire efficiency. Consequently, we believe this demographic will find a mobile lookup application useful since they want to replicate the online experience while in stores.
User Study: We located one research subject who was currently looking to purchase a laptop. We asked him how he makes purchasing decisions. He responded that he typically uses a variety of metrics such as price, appearance, functionality, and reviews. In order to find reputable reviews while at stores, he uses web-services such as Amazon to type in queries for products he is on the verge of purchasing. Despite its slow speed, he prefers to use a text-based query input method. He argued that picture-based product lookup applications from Amazon or Google do not work well because he found them difficult to use. He often wanted to specify multiple objects. In other words, he found the user interface inadequate for this needs because he was unable to easily specify the products he wanted to lookup. More explicitly, he pointed out that product ratings are relative and that the most important measurement is a relative cue. He wanted to be able to compare two products, not just see absolute ratings.
Proficient Users
Overview: Another demographic are the power users of technology. They adopt working technology into their lives, but do not necessarily use it its full potential. While such users are aware of online services like Amazon.com, they are comfortable shopping without them. Consequently, in order for to foster rapid adoption of technology into their lives, products must have high learnability. Once they learn how to use the software, they will continue to use it.
User Study: We brought one research subject to a furniture store and asked her to buy a sofa. Since this subject identified herself has a proficient but not savvy user of technology, she immediately located a salesperson and asked questions about the couch. The thought of using her iPhone to lookup information did not occur to her until we asked her about it. Upon explicit prompting, she stated she would normally lookup products on her personal computer at home before going to the store. Upon this reminder, she also stated that she wanted to go home to lookup sofas before making this purchase. She did not want to use an iPhone application because she often found them both a) rude to use in a store, and, more importantly, b) worried that they would be biased. In this vein, she felt that if she looked it up on Amazon, they would make their own products look better and preferred to shop using an independent source.
Causal Users
Overview: The final demographic we consider are users who only use a casual amount of technology in their life. These users are typically characterized by senior citizens. While they rarely use technology, when they do they need it to work reliably and minimize mistakes. Indeed, these users need safety and simplicity. They want to just see information about one product.
User Study: After finding a more senior research subject (i.e., postdoc), we brought them to a coat store and asked them to determine which coat to purchase. Again, this user went straight to the salesperson. When asked about computerized lookup methods, they stated they normally do not use them because they do not want to cumbersomely type in product information. When we explained the idea of our project, they agreed it would be very useful, as long as it worked well. They thought they would use it because it only required a few clicks to work.
Summary
Our preliminary experiments indicate that while a mobile product look service would be useful and impact, user interfaces are slow and awkward. We believe that by developing an intuitive user interface for mobile image annotation and recognition that we can build a useful product lookup service. In
Task Analysis
Inspired by our experiences after interviewing subjects, we identified three primary tasks for a mobile phone application: quickly locating reviews for a product, efficiently comparing products, and clearly showing results to the user. In this section, we discuss these tasks for each user demographic.
How do users lookup products?
Current systems either have no method for specifying the object (instead, they assume it is centered), or they have (obtuse) interfaces for drawing a rectangular bounding box. Due to the sensitivity of state-of-the-art computer vision algorithms, poorly placed localizations lead to significantly incorrect identification results. Both cause losses in efficiency which compel users to use typed queries, or to not use their mobile device at all. Indeed, we need more efficient methods for specifying objects of interest in a photo.
How do users compare products?
Current systems do not support looking up multiple items in a single image, causing the user to have to repeatedly do the same task for multiple products. What is a visible, learnable and efficient interface for allowing users to specify multiple products in the same image?
How do users get unbiased results?
Current systems (e.g. Amazon Remembers, Google Goggles) will open a product page or search engine results page. Neither method would scale to multiple lookups. Even for individual lookups, it isnot clear that these are the best methods for conveying what a user is likely to be interested in for both prices and reviews. What is the best way to convey results to the user?