The Opportunity

66% of smartphone users say they will turn to their smartphones to lookup something they saw while watching a TV commercial. I am one of those people. By leveraging this known behavior Google can open the doors to a delightful shopping experience that is as dynamic as it is useful. Thus tapping into the full marketing power of movies, and shows.

Closing the gab between product placement and checkout

The purpose of this project will be to outline and design a product with the following attributes:

1. A system that allows a user to act on their impulses when they are inspired to purchase a product. 

2. This system will use context aware search and discovery for frictionless entry into purchasing funnel.

3. Google Now will be used as the content delivery mechanism via an Android enabled smartphone.

4. Spot will not be used to transact at the end of the checkout funnel but instead will only provide guidance as to where to buy.

Dependencies

1. An Android smartphone

2. A system and method to add metadata to television content by producers, corporations and networks. 

3. A system and method of enhancing user search results by determining a television program currently being displayed in proximity to an electronic device. Ref. US Patent no.: 8,839,303 B2

Process & Design Thinking

Assumptions 

Considering the space and research associated I made the following assumptions: 

#1 The User

Millennial and Generation X'ers will be the target users. Not only do these users watch over 20 hours per week of television but they also spent the greatest amount of time with a second screen; particularly the smartphone. Additionally the Millennial spends more time focused on the smartphone while in front of the television while the inverse is true for Generation X'ers. 

#2 The Use Case

A user is often inspired to buy a garment due to a trigger (sale, scarcity, perceived value). Reduction of friction is fundamental to preventing lost purchases. This means that the user must also know what it is that they are attempting to purchase and how to purchase it.

#3 The Signal

The system must recognize when a user is interested in a product, explicitly or implicitly, by actions such as subscription, search, browsing, or content consumption. 

#4 The Data

Content owners must provide metadata so that the end user is exposed to useful and accurate content leading up to a purchase.  Content owners (Brands, Media Production Houses) will see this as another distribution channel by which to monetize. 

Exploration 

Based on the assumptions and tech constraints I explored the viability of the following solution:

Snap & Buy (Discarded)

A user can be anywhere and see a garment they want and take a photo of the item.

The following should take place:

• The garment could be spotted in a shop window, on another person or even in a magazine. The user will open the app and take a photo.

• Once a photo is taken computer vision will be used to identify the garment(s) and categorizes them into types (skirt, blouse, sweater, etc.). 

• Check these categories against available data based on geolocation. Data points like surrounding stores and regional purchasing patterns would be accounted for using Google Shopping Insights. 

• provide the user with ability to purchase the item(s).

Tech Dependency 

Google Shopping insights 

Composite template for cloth modeling and sketching

Problems with this solution

1. Heavy on processing  

2. It's creepy to take photos of people in public

3. Barriers to adoption are high. It is not intuitive to the user that they should launch an app take a photo and expect this sort of result. 

4. High degree of error. Requires near perfect conditions to work and will not likely return exact matches when it does work. Instead it will return suggested matches leading to a reduction in experience. 

5. Hard to detect accessories like jewelry. 

Final Solution

Dynamically detecting content based on proximity to TV

A natural environment solution had too many variables to consider so building a snap and buy system was sure to fail on some front. I explored impulse purchases in the living room instead because I found that a captive audience exhibits more predictable behavior, data can be more structured, and content creators (brands and TV production companies) could actively participate in providing a layer of metadata that would allow a user to purchase anything from programming.

The following should take place:

• A user watches their shows, movies, or commercials.

• A smartphone will detect the content that the user is watching. (US Patent no.: 8,839,303 B2)

• When a user conducts a google search of "What is Sandra Bullock wearing at the Oscars?" from their browser the system will be context aware and return a notification.

• User will be directed to Google Now. where they can see the garment and accessories along with name, price, and description. 

• Returning users will learn that they can go directly to Google Now to browse or use voice search from their home screen to ask about a garment. 

• Eventually the system will learn a user's habits and preferences and will send garment recommendations from their favorite programming or brands. 

• From the product details page a user can click on the shop button and will be forwarded to the place in which they can purchase the item. The sale will not be made in the app. 

Context Aware Search

The television is just the supporting device. The real star here is search, although TV provides a way to bring context and help provide structured data. If you're near your TV and you initiate the voice command from the home screen of your smartphone and say "I love her look!" while watching The Big Bang Theory you'll get the results you're looking for. 

Workflow

Visual Implementation 

When we set out to make events last forever we found it was the events and activities that got the most social media chatter were exposed to users for longer periods. With this increased visibility user got a better feel for events and organizers beyond the scope of the actual event itself. 

More Featured Projects