[lecture] 왜 지금 ‘컨텍스트의 시대’를 논하는가? (App Center)

Title	"왜 지금 ‘컨텍스트의 시대’를 논하는가?" -(translation) Why now are we talking about “Age of Context?”
Place	KAIST Dogok Software Grad School 101-ho
Time	May 16, 2014 7PM~9:20PM
Speaker	박지훈, 이구환, Ciren Jang
Organizer	App Center
Focus	5 forces for contexts

I.       Message (c.f. Amazon: booklink)

1.       Related terms

1)      Definition of context: the circumstances and facts that surround a situation

2)      “If content is king, context is god.” – Gary Vaynerchuk

2.       Five forces shaping the age of context

1)      Mobile: mostly wearables

2)      Social media: real-time emotions on SNS

3)      Data: analyzable large data

4)      Sensors: eyeball tracker at Google Glass, light and humidity sensor at Plant Sensor, temperature and humidity sensor at Nest, depth sensor at Prime Sense

5)      LBS: foursquare, Check-in, Waze, Asthmas Police

3.       Application: Try to think of value-added services based on the above 5 forces

4.       Example

1)      Personalized and predictive service

①  Google Glass- e.g. displaying my stock portfolio info

②  Google Search- e.g. Jejudo ticket info can be displayed at the top of SRP

③  MindMeld- e.g. previous conversation info is displayed during video-chatting

④  IBM Watson- e.g. predict why a person has to have a certain sickness by analyzing the person’s data

⑤  Google Now- e.g. flight ticket info can be displayed on SRP

⑥  23andMe- e.g. DNA analysis service to provide probability report to get a certain disease

2)      Anticipatory services and products

①  SRI’s tempo- utilized all possible information related to calendar schedule (e.g. parking info around meeting locations and emails related to the meeting)

②  Tapingo- utilized time and place info (e.g. you used to eat Chinese food today, would you like to eat it today?)

③  Tagwhat- telling you what’s happening nearby

3)      Context means business

①  Knowing everything about customers: Uber and GE’s industrial Internet

②  Knowing your customer in deep detail

A       Audience.fm

B       Datasift- analyzes Tweeted messages via Twitter’s Firehose

C       Vintank- recommends wines based on user’s wine related tweet through their user profiling technology

5.       Case utilizing all 5 forces?

1)      Disney’s RFID bracelet- visited restaurant, used as a payment tool

①  identifies a customer spending lots of money and sends coupon notification to make customers spend more money

2)      Belgium museum’s iBeacon- sends description of a museum item to visitors

3)      Yahoo’s Marissa Mayer- announced that Yahoo search engine will provide contextual search results

4)      Mercedes-Benz- preparing a service predicting passengers’ directions based on dates, riders, and previous destinations

6.       Contextual privacy: e.g. Friends’ nearby, Edward Snowden

7.       Ask below questions for opportunities--> to come up with services saving people’s time

1)      Do you know your customer?

2)      Are you building highly personalized systems? Or are you serving those who are?

3)      Do you see a way to “Uberize” your business?

4)      Are you dreaming of a new sensor-based world and how it might be different?

5)      Are you hanging out with people who are building next generation technologies?

II.      Takeaway

1.       Anticipatory services and products cases

[lecture] 포탈회사의 연구개발 : Naver Labs 를 중심으로- 송창현 Naver Labs 연구센터장

Title	“포탈회사의 연구개발 : Naver Labs 를 중심으로” -       (translation) R&D at Portal Company : Naver Labs
Place	KAIST Dogok Software Grad School Chin’s AMP Hall 103-ho
Time	May 1, 2014 5PM~6PM
Speaker	송창현, Naver Labs 연구센터장 -       (translation) Chang Hyun Song; Research Head, Naver Labs
Organizer	KAIST Software Graduate Program
Focus	Naver Labs management

I.       Message

1.       Employee: 130 at Naver Labs

2.       Product example: voice communication, wine label and music detection

3.       Admin position: created an admin position which handles most of non-engineering works

4.       Ways to adopt new technology: licensing, HW/SW tuning, open source SW, internal TFT (e.g. expert organization)

5.       Software strategy

1)      Establishment of Naver Labs allowed Naver to move from tactical to strategic and pure research

2)      Research and development should go together

6.       Criteria for employee evaluation: growth, technology, leadership, collaboration

7.       Internal technology festival: invite employees in other departments (e.g. PM, sales) to show developed technology (e.g. Similar to Microsoft’s internal TechFest)

8.       Sensual content filtering: manual filtering e.g. 100 employees

9.       People management: anytime interruption is allowed unless the employee’s timeslot is marked as no interruption

II.      Takeaway

1.       Create an admin position to accelerate development

2.       Ways to adopt new technology 

[personal UX/UI review] packaging stickers

product: School Food's packaging stickers

issue: difficult to open

recommendation: try using easily detachable stickers 

reasoning: unboxing experience is not just important for electronics companies but also in any products which require customers touch

[personal UX/UI review] McDonald's card reader design

Product: McDonald's card reader

Issue: difficult to tell which side of card needs to go where when swiping 

Recommendation: make each side of the reader read cards

[personal UX/UI review] iOS bug

Product: iOS for iPhone

Issue: box within a box

Recommendation: keep box interaction consistent 

[lecture] Fireside Chat with Mitch Lowe and 유영철

Title	Fireside Chat with Mitch Lowe, App Marketing
Place	D.Camp 5F Seminar Rm C
Time	April 10, 2014 4:30PM~9:30PM
Speaker	Mitch Lowe, 유영철
Organizer	Wisdome
Focus	multi-platform service, app marketing

I.       Mitch Lowe, CEO of Quarterly

1.       Message

1)      Netflix Info

①  Business: launched in 1998 as Kiball? (Netflix’s previous name) (c.f. Amazon 1994 launch)

②  Internal rule- if business doesn’t succeed, reiterate business model renewal every 90 days

③  Stat: grew up to 40 million customers, 7 million DVD via mail before he took off the company

2)      Redbox Info

①  Business: started small with 6 kiosks->40,000 kiosks in North America in 2012

②  Pattern: 40% of DVDs are returned via different kiosks than the ones they rented out from

3)      Advices for start-ups

①  Execution is the key for success

A       Build the right team: find someone with whom you can be with long hours

B       Try zombie test to find the right partner: meaning when you get bitten by zombies, will he tell you about that fact

C       Keeping the momentum going: improve and do something about your business to keep investors and customers continuously interested

4)      Regrets

①  Didn’t fire bad people early enough

②  Got lazy in raising money (c.f. US Congress raise money every 2 weeks)- one of the key reasons why raising money is to recruit best people

③  Didn’t hire enough smart people and didn’t overpay them

2.       Q&A

1)      What factors did you consider when launching a service in different platforms; which platform to introduce first?

①  Studied cancellation rate for each platform and asked what makes users quit and how often do they quit using them

2)      For CineMatch, can you make a comment on how do you find matching movies for individuals? -> Collaborative filtering <source: link>

①  Searches the CineMatch database for people who have rated the same movie - for example, "The Return of the Jedi"

②  Determines which of those people have also rated a second movie, such as "The Matrix"

③  Calculates the statistical likelihood that people who liked "Return of the Jedi" will also like "The Matrix"

④  Continues this process to establish a pattern of correlations between subscribers' ratings of many different films

3.       Takeaway

1)      Recommendation algorithm: CineMatch

2)      For start-ups, the importance of interviewing applicants with as many employees as possible and quickly firing people if they are not a good fit with the company

II.      유영철, CEO of GameBerry 

1.       Message

1)      App marketing strategy: utilize platforms (e.g. “카카오 게임하기” ranking, “카카오 친구초청하기”), tracking rules (e.g. CPK: 카카오 연동)

2)      General strategy

①  Game app- usually peak downloads go one month and can last up to three months

②  Non-game app (e.g. utility apps)- try discount pricing

3)      Success case: Polaris (e.g. US- Facebook; JN- Twitter; KR- blogs)

①  Tried price discount

②  Paid attention to competitive apps and put marketing money when the number of those apps is going down

2.       Takeaway

1)      For mobile app marketing, trends in Korea are moving to non-incentive ad network such as AdMob or Inmobi

2)      For non-game app marketing, put budget on mobile ad platforms such as Facebook

3)      For Google Play marketing, put money incrementally such as 1 million KRW-> 5-> 10 instead of the other way around

4)      Shoot to be featured in each app marketplace to increase your revenue

5)      Take advantage of competitors’ app ranking trends (e.g. perhaps with App Annie): meaning put money in when they are going down

6)      Facebook ads provides 10~20% conversion rate on average (c.f. 1~3% on AdMob)

7)      When advertising, always put analytics and tracking tools together

8)      For CPI marketing, run a test such as 1,000 downloads first?

[personal UX/UI review] designing a button icon

-. case: a button design on YouVersion website

-. issue: the button looks clickable but it isn't in reality

-. recommendation: do not differentiate the background color for the button if it's not going to be clickable :)