1. think like Wall Street
1.1. vs traditionnal marketing
1.1.1. speed
1.1.2. precision
1.1.3. too fast for humans
1.2. challenges
1.2.1. real time
1.2.2. analytics-intensive
1.3. needs
1.3.1. good algorithms
1.3.2. analyze bigdatas
2. it is not a game
2.1. complexity
2.1.1. social
2.1.2. mobile
2.1.3. apps
2.1.4. geolocalisation
2.2. the fastest wins
2.2.1. reduce latency
2.2.1.1. everywhere !
2.2.1.2. not only in programs and servers
2.2.1.3. choose datacenters close to the users
2.3. problems
2.3.1. migration can be hard
2.3.1.1. datas
2.3.1.2. team
2.3.1.3. process
2.3.1.4. integration
2.3.2. dude, where is my money ?
2.3.2.1. it may take time
2.3.2.2. think "investissement"
3. New Idea
4. to see
4.1. predictive analytics
4.2. automated workflow platforms
4.3. personalization
4.3.1. message
4.3.2. content
4.3.3. channel
5. sources
5.1. http://venturebeat.com/2014/04/20/why-madison-avenue-is-becoming-more-like-wall-street/
5.2. https://chrisohara.wordpress.com/2011/06/23/fish-don%E2%80%99t-know-he%E2%80%99s-wet/
5.3. http://venturebeat.com/2014/04/23/marketing-tech-how-to-make-or-lose-money-with-marketing-automation/
5.4. http://springload.com/2013/07/11/how-arbitrage-works-in-digital-advertising-today/
6. Real-Time Bidding (RTB)
6.1. what
6.1.1. target user in real time
6.1.2. buy individual impressions BEFORE page is loaded
6.1.3. 0.11 sec to generate target ad
6.2. why
6.2.1. traditional online ads = inefficient
6.2.2. better value for the publisher inventory
6.3. how
6.3.1. Demands Side Plateform (DSP)
6.3.1.1. automate the purchase of ads
6.3.1.2. arbitrage
6.3.1.2.1. statistical model of users
6.3.1.2.2. historical model of inventories
6.3.1.2.3. campaign
6.3.2. Publisher
6.3.2.1. Supply Side Plateform (SSP)
6.3.2.2. provides inventory to ads exchange
6.3.3. Ad Exchanges
6.3.3.1. connect publishers and advertisers
6.3.3.2. value of the bid, like stock exchanges
6.4. problems
6.4.1. inventory quality
6.4.1.1. to have clicks you need quality content
6.4.2. datas from cookies are not magic
6.4.3. facebook and google will always have better datas than you