SlideShare a Scribd company logo
1 of 24
廣告效果導向為基礎的行動廣告系統
Recommender as an example
Steven Chiu
RD department
Vpon Inc.
Outline
 Background, challenges and KPIs
 Basic concept
 Challenges and KPIs
 Vpon Ad service infrastructure
 AD effectiveness related work
 Recommender
 System flows
 Summary
 Q&A
Basic concept
Vpon Ad service infrastructure
Challenges and KPIs
Typical use case
Clicks
Conversions
The media
Landing pages
ADs
Ads on Vpon…
Mainly for Navigation apps, e.g. Navidog
POI (Map)
POI (Banner)
Normal
Full screen ads Video ads
Ads on Vpon…
AD Performance Evaluation
 Click Through
Rate (CTR)
 Conversion Rate
 Goals
To maximize
CTR
To maximize
conversations
Click
Conversion
Impression
Integration
Apps
Placing Ads
• Charged in CPC,
CPM
• Criteria:
• time, locations, app
categories, budget,
Performance reports
Advertisers
app
App reports
app app …
Mobile app users
Mobile app publishers
Advertisers
Ad performance reports
Vpon AD services backend
Data Archiving & Analysis
User Context
Runtime
information
User’s Ad
Requests
Ad Serving
Scalable
AD Serving
Transaction
& Billing
Real-time
Ad Selection
UserScenario
Modeling
Data
Mining
MR/Spark
HBase
HDFS
Ad-hoc
Analytics
Reporting &
Data
Warehouse
Adaptive AD
Distribution
System
Continues
Improvement
Ad
performance
P3
60+ M
Monthly Active Unique Devices
200+ M
of Daily Ad Requests
2+ T
Ad transaction records over time
25+ M
Cell Towers/Wi-Fi AP Location Data
Some numbers for Vpon AD Network
P2
Taipei, Shanghai, HK, Bejing and Tokyo
2 IDCs at Taipei, Shanghai and Some Amazon EC2 nodes
Data Analysis
Ad Requests
Ad web
service
Backend
Memory
cache
In-
memory
Grid
HBase
MapReduce/Spark
HA Proxy
Message Routing (Apache Kafka)
Ad
Request
Cue
Backend
Hadoop Distributed
File System
(HDFS)
User Profiles
Ad Requests
HTTP POST
Avro Avro Avro
Ad videos, images
HTTP Get
Data Processing and Archiving
Creative
and videos
AD
management
Report UI
(Django,
SSH)
Vpon AD services
backend functionsCDN
Recommender System
Other
undergoing
topics
Reporting system
Sales
Support
System
AD-hoc
reporting
Operation
Ganglia
Solr
AD Operation
AD
Monitoring
System
Scenario
modeling
Avro
Web
Proxy
+
Cache
Memory
cache
In-
memory
Grid
Cue
User Profiles
(Couch DB
and HBase)
Rsync, Avro Avro
Python + pig, hive,
Hadoop Streaming, spark
Python + pig, hive,
Hadoop Streaming, spark
Advertisers
Recommender as an example
Design and Implementation
Recommender
 Types
 User(imei) based recommender system
 Item(ad) based recommender system
 Steps
 Step1: Campaign/AD similarity table
 Step2: Prediction Phase
 Step3: Verification Phase
 (Continuous Improvement)
• Serve ads according to users
preference
Recommender flow
Prediction
Machine
Learning
(e.g. recommender)
Evaluation
Data
Selection
• Select user records of the Ad
Click/Conversion action by
different kinds of Apps
• Select users logs of the
Location, Date/Time, Usage
Freq., Area, Movement Speed…
• Identify relation of the conversion
types, App info, Ad info and user
info to best choose configurations
• Campaign/AD similarity calculation
• User preferences
• Advertising in accordance with
the identified targeted users
• Feedback the AD execution
results into the system for
adjusting the modeling adaptively
P5
Ad 1 Ad 2 Ad 3 Ad 4 Ad N
User 1 0 0 1 0 0
User 2 1 1 0 1 0
User 3 1 1 1 1 1
User 4 1 1 0 0 0
User N … … … … …
Step1: Ads' Similarities
Unique
device IDs
from latest
K months
Historical and ongoing ads (App downloads as conversions)
Ad 1 Ad 2 Ad 3 Ad 4 Ad N
User 1 P(1,1) P(1,2) P(1,3) P(1,4) P(1,5)
User 2 P(2,1) P(2,2) P(2,3) P(2,4) P(2,5)
User 3 P(3,1) P(3,2) P(3,3) P(3,4) P(3,5)
User 4 P(4,1) 1P(4,2) P(4,3) P(4,4) P(4,5)
User Z … … … … …
Step2: Users' Preferences
Unique
device IDs
from latest
K months
Historical and ongoing ads (App downloads as conversions)
User 1
User 2
… … … … … …
Step3: Prediction Phase:
ADs sorted by preference
Data Analysis
Ad Requests
Ad web
service
Backend
Memory
cache
In-
memory
Grid
HBase
MapReduce/Spark
HA Proxy
Message Routing (Apache Kafka)
Ad
Request
Cue
Backend
Hadoop Distributed
File System
(HDFS)
User Profiles
Ad Requests
HTTP POST
Avro Avro Avro
Ad videos, images
HTTP Get
Data Processing and Archiving
Creative
and videos
Billing
System
CDN
Recommender System
Other
undergoing
topics
Reporting system
Sales
Support
System
AD-hoc
reporting
Operation
Ganglia
Solr
AD Operation
AD
Monitoring
System
Scenario
modeling
Avro
Billing
Proxy
+
Cache
Memory
cache
In-
memory
Grid
Cue
User Profiles
(Couch DB
and HBase)
Rsync, Avro Avro
Step3: Prediction Phase:
Serving Ads based on
Preferences
user1 ad1,ad2, ad5
user2 ad2,ad4, ad5
user3 ad4,ad5,ad6,ad8
user1
Persisted on Apache CouchDB
Replicated to in-memory grid
Step4: Evaluation Phase
Using our Optimization Model,
the CTR increased 3~4 times
Normal
1st Rnd
Optimized
1st Rnd
Normal
2nd Rnd
Optimized
2nd Rnd
Clk 987 2318 973 2330
Imp 122,514 82,229 122,397 81,882
CTR 0.81% 2.82% 0.79% 2.85%
0.00%
0.50%
1.00%
1.50%
2.00%
2.50%
3.00%
1
10
100
1,000
10,000
100,000
1,000,000
CTR
#ofImp./Clk.
Perf. Campaign
0.000%
1.000%
2.000%
3.000%
4.000%
5.000%
6.000%
7.000%
Clk v.s. Conv
0.746%
3.646%
6.386%
Clk v.s. Conv
Normal 0.746%
recm_1st lvl. 3.646%
recm_2nd lvl. 6.386%
Game App DL Clk v.s. Conv.
After our 2nd lvl optimization,
the conv. v.s. click increased 8.56 times
Step5: continuous monitoring and improvement
10,037,003
2,451,061
85.01%
81.29%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
100.00%
0
2,000,000
4,000,000
6,000,000
8,000,000
10,000,000
12,000,000
No-Optimization Optimized
Target%(Perf.)
Imp.consumed(Cost)
Imp. Consumed (Cost) Targeted % (Perf.)
0.00%
0.20%
0.40%
0.60%
0.80%
1.00%
1.20%
1.40%
1.60%
CTR
CVR
IVR
0.90%
1.53%
1.37%
0.88%
0.57%
0.50%
Optimize Normal
CTR = Click v.s. Impression
CVR = Click v.s. Conv.
IVR = Imp. v.s. Conv.
Conv. Rate increased 3 times
Cost Optimization:
Cost reduced more than 75% while
performance only decreased 3.72%
Implementation
 Hadoop MapReduce as computing platform
 Using Hadoop streaming with Python
 Map: a list of ad pairs as input for similarity caculation
 Reduce: simply aggregate the map results
 Re-modeling on a daily basis based on results
 Will go on to use Haoop HDFS + Spark + Python for performance
benefit
Summary
 Build the infra. that proves models effective or not as early
as possible
 AB testing for new models
 Automate as much as possible
 Monitoring and measurement
 Computing resource
 Properly manage Product, ad-hoc, analysis jobs
 Optimization does work
 Use Python wherever it fits

More Related Content

What's hot

CityAds Affiliate Marketing
CityAds Affiliate MarketingCityAds Affiliate Marketing
CityAds Affiliate MarketingNikolay Khokhlov
 
Paid search management technology
Paid search management technologyPaid search management technology
Paid search management technologybbullockRKG
 
Lurker intro ppc_seo_2011
Lurker intro ppc_seo_2011Lurker intro ppc_seo_2011
Lurker intro ppc_seo_2011RebelRouse
 
Full Picture For Ad Serving On Multiple Screens
Full Picture For Ad Serving On Multiple ScreensFull Picture For Ad Serving On Multiple Screens
Full Picture For Ad Serving On Multiple ScreensAxel Hoehnke
 
Online Ad Serving
Online Ad ServingOnline Ad Serving
Online Ad ServingNeha Gupta
 
Watson Customer Engagement
Watson Customer EngagementWatson Customer Engagement
Watson Customer EngagementHeber Lopes
 
Intro to Programmatic Advertising with Matt Prohaska from Prohaska Consulting
Intro to Programmatic Advertising with Matt Prohaska from Prohaska ConsultingIntro to Programmatic Advertising with Matt Prohaska from Prohaska Consulting
Intro to Programmatic Advertising with Matt Prohaska from Prohaska ConsultingStukent Inc.
 
Internet Advertising for Dummies
Internet Advertising for DummiesInternet Advertising for Dummies
Internet Advertising for DummiesSajid Abdul Rahiman
 
Programmatic ad buying in 3 slides
Programmatic ad buying in 3 slidesProgrammatic ad buying in 3 slides
Programmatic ad buying in 3 slidesOne Marketing Ltd
 
Iab weborama october
Iab weborama octoberIab weborama october
Iab weborama octoberiabrussiaprez
 
Display Advertising Basics
Display Advertising BasicsDisplay Advertising Basics
Display Advertising BasicsBidGear Inc.
 
Attribution marketing, from theory to reality - Kwanko
Attribution marketing, from theory to reality - KwankoAttribution marketing, from theory to reality - Kwanko
Attribution marketing, from theory to reality - KwankoKwanko
 
SiteScout DSP Update (September 2014)
SiteScout DSP Update (September 2014)SiteScout DSP Update (September 2014)
SiteScout DSP Update (September 2014)sitescout
 

What's hot (20)

CityAds Affiliate Marketing
CityAds Affiliate MarketingCityAds Affiliate Marketing
CityAds Affiliate Marketing
 
Paid search management technology
Paid search management technologyPaid search management technology
Paid search management technology
 
Lurker intro ppc_seo_2011
Lurker intro ppc_seo_2011Lurker intro ppc_seo_2011
Lurker intro ppc_seo_2011
 
Full Picture For Ad Serving On Multiple Screens
Full Picture For Ad Serving On Multiple ScreensFull Picture For Ad Serving On Multiple Screens
Full Picture For Ad Serving On Multiple Screens
 
Online Ad Serving
Online Ad ServingOnline Ad Serving
Online Ad Serving
 
Watson Customer Engagement
Watson Customer EngagementWatson Customer Engagement
Watson Customer Engagement
 
Cheat sheetmonetization1
Cheat sheetmonetization1Cheat sheetmonetization1
Cheat sheetmonetization1
 
Intro to Programmatic Advertising with Matt Prohaska from Prohaska Consulting
Intro to Programmatic Advertising with Matt Prohaska from Prohaska ConsultingIntro to Programmatic Advertising with Matt Prohaska from Prohaska Consulting
Intro to Programmatic Advertising with Matt Prohaska from Prohaska Consulting
 
Internet Advertising for Dummies
Internet Advertising for DummiesInternet Advertising for Dummies
Internet Advertising for Dummies
 
Programmatic ad buying in 3 slides
Programmatic ad buying in 3 slidesProgrammatic ad buying in 3 slides
Programmatic ad buying in 3 slides
 
iMobiTrax Overview
iMobiTrax OverviewiMobiTrax Overview
iMobiTrax Overview
 
Iab weborama october
Iab weborama octoberIab weborama october
Iab weborama october
 
Aumark Marketing Automation
Aumark Marketing AutomationAumark Marketing Automation
Aumark Marketing Automation
 
Display Advertising Basics
Display Advertising BasicsDisplay Advertising Basics
Display Advertising Basics
 
Emarketing
EmarketingEmarketing
Emarketing
 
Attribution marketing, from theory to reality - Kwanko
Attribution marketing, from theory to reality - KwankoAttribution marketing, from theory to reality - Kwanko
Attribution marketing, from theory to reality - Kwanko
 
Overview RTB ecosystem
Overview RTB ecosystemOverview RTB ecosystem
Overview RTB ecosystem
 
Qpx CLab unipro
Qpx CLab uniproQpx CLab unipro
Qpx CLab unipro
 
SiteScout DSP Update (September 2014)
SiteScout DSP Update (September 2014)SiteScout DSP Update (September 2014)
SiteScout DSP Update (September 2014)
 
Deck
DeckDeck
Deck
 

Viewers also liked

數位廣告的血淚進化 20150714
數位廣告的血淚進化 20150714數位廣告的血淚進化 20150714
數位廣告的血淚進化 20150714Ruby Kuan 關芸如
 
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理悠識學院
 
如何社群行銷?就是不銷而銷!
如何社群行銷?就是不銷而銷!如何社群行銷?就是不銷而銷!
如何社群行銷?就是不銷而銷!綠生活 GreenLife
 
十分鐘關鍵字廣告上手
十分鐘關鍵字廣告上手十分鐘關鍵字廣告上手
十分鐘關鍵字廣告上手bahn hong
 
最新網路行銷廣告策略與分配管理
最新網路行銷廣告策略與分配管理最新網路行銷廣告策略與分配管理
最新網路行銷廣告策略與分配管理Norika
 
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長悠識學院
 
打倒程式化購買術語
打倒程式化購買術語打倒程式化購買術語
打倒程式化購買術語NT150 Com
 
施典志(Tenz):社群工具的本質與應用方法
施典志(Tenz):社群工具的本質與應用方法施典志(Tenz):社群工具的本質與應用方法
施典志(Tenz):社群工具的本質與應用方法開拓文教基金會
 
新手創業家的哀愁與美麗
新手創業家的哀愁與美麗新手創業家的哀愁與美麗
新手創業家的哀愁與美麗Norika
 
融合與衝突 是促進還是毀滅世界的武器
融合與衝突 是促進還是毀滅世界的武器融合與衝突 是促進還是毀滅世界的武器
融合與衝突 是促進還是毀滅世界的武器Norika
 
創業這條不歸路
創業這條不歸路創業這條不歸路
創業這條不歸路Norika
 
Facebook廣告操作實務
Facebook廣告操作實務Facebook廣告操作實務
Facebook廣告操作實務Norika
 
創業的美麗與哀愁
創業的美麗與哀愁創業的美麗與哀愁
創業的美麗與哀愁Norika
 
網路廣告基礎入門
網路廣告基礎入門網路廣告基礎入門
網路廣告基礎入門Norika
 
Facebook 粉絲團經營教學
Facebook 粉絲團經營教學Facebook 粉絲團經營教學
Facebook 粉絲團經營教學Vince Liao
 
How to Successfully Run a Remote Team
How to Successfully Run a Remote TeamHow to Successfully Run a Remote Team
How to Successfully Run a Remote TeamWeekdone.com
 
20 Fantastic Flat Icons and Their Meaning In Logo Design
20 Fantastic Flat Icons and Their Meaning In Logo Design20 Fantastic Flat Icons and Their Meaning In Logo Design
20 Fantastic Flat Icons and Their Meaning In Logo DesignDesignMantic
 
2016 Digital predictions for marketing, tech, pop culture and everything in b...
2016 Digital predictions for marketing, tech, pop culture and everything in b...2016 Digital predictions for marketing, tech, pop culture and everything in b...
2016 Digital predictions for marketing, tech, pop culture and everything in b...Soap Creative
 

Viewers also liked (20)

數位廣告的血淚進化 20150714
數位廣告的血淚進化 20150714數位廣告的血淚進化 20150714
數位廣告的血淚進化 20150714
 
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理
[SDX2016] 網站分析工作的領悟 / 鍾喬后 Isobar 安索帕 資料分析經理
 
如何社群行銷?就是不銷而銷!
如何社群行銷?就是不銷而銷!如何社群行銷?就是不銷而銷!
如何社群行銷?就是不銷而銷!
 
十分鐘關鍵字廣告上手
十分鐘關鍵字廣告上手十分鐘關鍵字廣告上手
十分鐘關鍵字廣告上手
 
最新網路行銷廣告策略與分配管理
最新網路行銷廣告策略與分配管理最新網路行銷廣告策略與分配管理
最新網路行銷廣告策略與分配管理
 
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長
[SDX2016] 2016年 SEO 的關鍵在 UX / 連啓佑 將能數位行銷 執行長
 
打倒程式化購買術語
打倒程式化購買術語打倒程式化購買術語
打倒程式化購買術語
 
施典志(Tenz):社群工具的本質與應用方法
施典志(Tenz):社群工具的本質與應用方法施典志(Tenz):社群工具的本質與應用方法
施典志(Tenz):社群工具的本質與應用方法
 
新手創業家的哀愁與美麗
新手創業家的哀愁與美麗新手創業家的哀愁與美麗
新手創業家的哀愁與美麗
 
融合與衝突 是促進還是毀滅世界的武器
融合與衝突 是促進還是毀滅世界的武器融合與衝突 是促進還是毀滅世界的武器
融合與衝突 是促進還是毀滅世界的武器
 
創業這條不歸路
創業這條不歸路創業這條不歸路
創業這條不歸路
 
Facebook廣告操作實務
Facebook廣告操作實務Facebook廣告操作實務
Facebook廣告操作實務
 
創業的美麗與哀愁
創業的美麗與哀愁創業的美麗與哀愁
創業的美麗與哀愁
 
網路廣告基礎入門
網路廣告基礎入門網路廣告基礎入門
網路廣告基礎入門
 
Facebook 粉絲團經營教學
Facebook 粉絲團經營教學Facebook 粉絲團經營教學
Facebook 粉絲團經營教學
 
How to Successfully Run a Remote Team
How to Successfully Run a Remote TeamHow to Successfully Run a Remote Team
How to Successfully Run a Remote Team
 
Melt (Beta)
Melt (Beta)Melt (Beta)
Melt (Beta)
 
The Build Trap
The Build TrapThe Build Trap
The Build Trap
 
20 Fantastic Flat Icons and Their Meaning In Logo Design
20 Fantastic Flat Icons and Their Meaning In Logo Design20 Fantastic Flat Icons and Their Meaning In Logo Design
20 Fantastic Flat Icons and Their Meaning In Logo Design
 
2016 Digital predictions for marketing, tech, pop culture and everything in b...
2016 Digital predictions for marketing, tech, pop culture and everything in b...2016 Digital predictions for marketing, tech, pop culture and everything in b...
2016 Digital predictions for marketing, tech, pop culture and everything in b...
 

Similar to 廣告效果導向為基礎的行動廣告系統

Qcon London 2017 - Architecture overhaul - Ad serving @ Spotify scale
Qcon London 2017 -  Architecture overhaul - Ad serving @ Spotify scaleQcon London 2017 -  Architecture overhaul - Ad serving @ Spotify scale
Qcon London 2017 - Architecture overhaul - Ad serving @ Spotify scaleKinshuk Mishra
 
Epam BI - Near Realtime Marketing Support System
Epam BI - Near Realtime Marketing Support SystemEpam BI - Near Realtime Marketing Support System
Epam BI - Near Realtime Marketing Support SystemDmitry Tolpeko
 
Rd Online Deck 3.0
Rd Online Deck 3.0Rd Online Deck 3.0
Rd Online Deck 3.0Zestadz
 
Data Science at Flurry
Data Science at FlurryData Science at Flurry
Data Science at Flurrysoupsranjan
 
KB Seminars: Working with Technology - Advertising; 10/13
KB Seminars: Working with Technology - Advertising; 10/13KB Seminars: Working with Technology - Advertising; 10/13
KB Seminars: Working with Technology - Advertising; 10/13MDIF
 
20141209 meetup hassan
20141209 meetup hassan20141209 meetup hassan
20141209 meetup hassanNanda Kishore
 
Smadex Company Profile
Smadex Company ProfileSmadex Company Profile
Smadex Company Profilesmadex
 
Deepak-Computational Advertising-The LinkedIn Way
Deepak-Computational Advertising-The LinkedIn WayDeepak-Computational Advertising-The LinkedIn Way
Deepak-Computational Advertising-The LinkedIn Wayyingfeng
 
ROI without monetization - Stefan Bielau
ROI without monetization - Stefan Bielau ROI without monetization - Stefan Bielau
ROI without monetization - Stefan Bielau Adjust
 
Reach targeted audience segments with top-quality 3rd party data.
Reach targeted audience segments with top-quality 3rd party data.Reach targeted audience segments with top-quality 3rd party data.
Reach targeted audience segments with top-quality 3rd party data.reklamajans
 
ANTS Programmatic Agency - Credential
ANTS Programmatic Agency - CredentialANTS Programmatic Agency - Credential
ANTS Programmatic Agency - CredentialANTS
 
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017Marcel Prothmann
 
Epom Ad Server For Networks
Epom Ad Server For NetworksEpom Ad Server For Networks
Epom Ad Server For NetworksEpom
 
Data Science Salon: Enabling self-service predictive analytics at Bidtellect
Data Science Salon: Enabling self-service predictive analytics at BidtellectData Science Salon: Enabling self-service predictive analytics at Bidtellect
Data Science Salon: Enabling self-service predictive analytics at BidtellectFormulatedby
 
LINE's Vision of an Ideal Advertisement Platform
LINE's Vision of an Ideal Advertisement PlatformLINE's Vision of an Ideal Advertisement Platform
LINE's Vision of an Ideal Advertisement PlatformLINE Corporation
 
Project Lotus Intro Deck Aug
Project Lotus Intro Deck AugProject Lotus Intro Deck Aug
Project Lotus Intro Deck Augmartinsunwenhua
 
JBP-Cobalt all services
JBP-Cobalt all servicesJBP-Cobalt all services
JBP-Cobalt all servicesDuy Nguyen
 

Similar to 廣告效果導向為基礎的行動廣告系統 (20)

Qcon London 2017 - Architecture overhaul - Ad serving @ Spotify scale
Qcon London 2017 -  Architecture overhaul - Ad serving @ Spotify scaleQcon London 2017 -  Architecture overhaul - Ad serving @ Spotify scale
Qcon London 2017 - Architecture overhaul - Ad serving @ Spotify scale
 
Epam BI - Near Realtime Marketing Support System
Epam BI - Near Realtime Marketing Support SystemEpam BI - Near Realtime Marketing Support System
Epam BI - Near Realtime Marketing Support System
 
Rd Online Deck 3.0
Rd Online Deck 3.0Rd Online Deck 3.0
Rd Online Deck 3.0
 
Enliven cem clickstream solution
Enliven cem clickstream solutionEnliven cem clickstream solution
Enliven cem clickstream solution
 
Ad exchange product description
Ad exchange product descriptionAd exchange product description
Ad exchange product description
 
Data Science at Flurry
Data Science at FlurryData Science at Flurry
Data Science at Flurry
 
KB Seminars: Working with Technology - Advertising; 10/13
KB Seminars: Working with Technology - Advertising; 10/13KB Seminars: Working with Technology - Advertising; 10/13
KB Seminars: Working with Technology - Advertising; 10/13
 
20141209 meetup hassan
20141209 meetup hassan20141209 meetup hassan
20141209 meetup hassan
 
Smadex Company Profile
Smadex Company ProfileSmadex Company Profile
Smadex Company Profile
 
BOLO2010 Portugal
BOLO2010 PortugalBOLO2010 Portugal
BOLO2010 Portugal
 
Deepak-Computational Advertising-The LinkedIn Way
Deepak-Computational Advertising-The LinkedIn WayDeepak-Computational Advertising-The LinkedIn Way
Deepak-Computational Advertising-The LinkedIn Way
 
ROI without monetization - Stefan Bielau
ROI without monetization - Stefan Bielau ROI without monetization - Stefan Bielau
ROI without monetization - Stefan Bielau
 
Reach targeted audience segments with top-quality 3rd party data.
Reach targeted audience segments with top-quality 3rd party data.Reach targeted audience segments with top-quality 3rd party data.
Reach targeted audience segments with top-quality 3rd party data.
 
ANTS Programmatic Agency - Credential
ANTS Programmatic Agency - CredentialANTS Programmatic Agency - Credential
ANTS Programmatic Agency - Credential
 
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017
OMLIVE 2017 - ADWORDS TOOLS & SCRIPTS for PPC-NERDS 2017
 
Epom Ad Server For Networks
Epom Ad Server For NetworksEpom Ad Server For Networks
Epom Ad Server For Networks
 
Data Science Salon: Enabling self-service predictive analytics at Bidtellect
Data Science Salon: Enabling self-service predictive analytics at BidtellectData Science Salon: Enabling self-service predictive analytics at Bidtellect
Data Science Salon: Enabling self-service predictive analytics at Bidtellect
 
LINE's Vision of an Ideal Advertisement Platform
LINE's Vision of an Ideal Advertisement PlatformLINE's Vision of an Ideal Advertisement Platform
LINE's Vision of an Ideal Advertisement Platform
 
Project Lotus Intro Deck Aug
Project Lotus Intro Deck AugProject Lotus Intro Deck Aug
Project Lotus Intro Deck Aug
 
JBP-Cobalt all services
JBP-Cobalt all servicesJBP-Cobalt all services
JBP-Cobalt all services
 

廣告效果導向為基礎的行動廣告系統

  • 1. 廣告效果導向為基礎的行動廣告系統 Recommender as an example Steven Chiu RD department Vpon Inc.
  • 2. Outline  Background, challenges and KPIs  Basic concept  Challenges and KPIs  Vpon Ad service infrastructure  AD effectiveness related work  Recommender  System flows  Summary  Q&A
  • 3. Basic concept Vpon Ad service infrastructure Challenges and KPIs
  • 4. Typical use case Clicks Conversions The media Landing pages ADs
  • 5. Ads on Vpon… Mainly for Navigation apps, e.g. Navidog POI (Map) POI (Banner) Normal
  • 6. Full screen ads Video ads Ads on Vpon…
  • 7. AD Performance Evaluation  Click Through Rate (CTR)  Conversion Rate  Goals To maximize CTR To maximize conversations Click Conversion Impression
  • 8. Integration Apps Placing Ads • Charged in CPC, CPM • Criteria: • time, locations, app categories, budget, Performance reports Advertisers app App reports app app … Mobile app users Mobile app publishers Advertisers Ad performance reports
  • 9. Vpon AD services backend Data Archiving & Analysis User Context Runtime information User’s Ad Requests Ad Serving Scalable AD Serving Transaction & Billing Real-time Ad Selection UserScenario Modeling Data Mining MR/Spark HBase HDFS Ad-hoc Analytics Reporting & Data Warehouse Adaptive AD Distribution System Continues Improvement Ad performance P3
  • 10. 60+ M Monthly Active Unique Devices 200+ M of Daily Ad Requests 2+ T Ad transaction records over time 25+ M Cell Towers/Wi-Fi AP Location Data Some numbers for Vpon AD Network P2 Taipei, Shanghai, HK, Bejing and Tokyo 2 IDCs at Taipei, Shanghai and Some Amazon EC2 nodes
  • 11. Data Analysis Ad Requests Ad web service Backend Memory cache In- memory Grid HBase MapReduce/Spark HA Proxy Message Routing (Apache Kafka) Ad Request Cue Backend Hadoop Distributed File System (HDFS) User Profiles Ad Requests HTTP POST Avro Avro Avro Ad videos, images HTTP Get Data Processing and Archiving Creative and videos AD management Report UI (Django, SSH) Vpon AD services backend functionsCDN Recommender System Other undergoing topics Reporting system Sales Support System AD-hoc reporting Operation Ganglia Solr AD Operation AD Monitoring System Scenario modeling Avro Web Proxy + Cache Memory cache In- memory Grid Cue User Profiles (Couch DB and HBase) Rsync, Avro Avro Python + pig, hive, Hadoop Streaming, spark Python + pig, hive, Hadoop Streaming, spark Advertisers
  • 12. Recommender as an example Design and Implementation
  • 13.
  • 14. Recommender  Types  User(imei) based recommender system  Item(ad) based recommender system  Steps  Step1: Campaign/AD similarity table  Step2: Prediction Phase  Step3: Verification Phase  (Continuous Improvement)
  • 15. • Serve ads according to users preference Recommender flow Prediction Machine Learning (e.g. recommender) Evaluation Data Selection • Select user records of the Ad Click/Conversion action by different kinds of Apps • Select users logs of the Location, Date/Time, Usage Freq., Area, Movement Speed… • Identify relation of the conversion types, App info, Ad info and user info to best choose configurations • Campaign/AD similarity calculation • User preferences • Advertising in accordance with the identified targeted users • Feedback the AD execution results into the system for adjusting the modeling adaptively P5
  • 16. Ad 1 Ad 2 Ad 3 Ad 4 Ad N User 1 0 0 1 0 0 User 2 1 1 0 1 0 User 3 1 1 1 1 1 User 4 1 1 0 0 0 User N … … … … … Step1: Ads' Similarities Unique device IDs from latest K months Historical and ongoing ads (App downloads as conversions)
  • 17. Ad 1 Ad 2 Ad 3 Ad 4 Ad N User 1 P(1,1) P(1,2) P(1,3) P(1,4) P(1,5) User 2 P(2,1) P(2,2) P(2,3) P(2,4) P(2,5) User 3 P(3,1) P(3,2) P(3,3) P(3,4) P(3,5) User 4 P(4,1) 1P(4,2) P(4,3) P(4,4) P(4,5) User Z … … … … … Step2: Users' Preferences Unique device IDs from latest K months Historical and ongoing ads (App downloads as conversions)
  • 18. User 1 User 2 … … … … … … Step3: Prediction Phase: ADs sorted by preference
  • 19. Data Analysis Ad Requests Ad web service Backend Memory cache In- memory Grid HBase MapReduce/Spark HA Proxy Message Routing (Apache Kafka) Ad Request Cue Backend Hadoop Distributed File System (HDFS) User Profiles Ad Requests HTTP POST Avro Avro Avro Ad videos, images HTTP Get Data Processing and Archiving Creative and videos Billing System CDN Recommender System Other undergoing topics Reporting system Sales Support System AD-hoc reporting Operation Ganglia Solr AD Operation AD Monitoring System Scenario modeling Avro Billing Proxy + Cache Memory cache In- memory Grid Cue User Profiles (Couch DB and HBase) Rsync, Avro Avro Step3: Prediction Phase: Serving Ads based on Preferences user1 ad1,ad2, ad5 user2 ad2,ad4, ad5 user3 ad4,ad5,ad6,ad8 user1 Persisted on Apache CouchDB Replicated to in-memory grid
  • 20. Step4: Evaluation Phase Using our Optimization Model, the CTR increased 3~4 times Normal 1st Rnd Optimized 1st Rnd Normal 2nd Rnd Optimized 2nd Rnd Clk 987 2318 973 2330 Imp 122,514 82,229 122,397 81,882 CTR 0.81% 2.82% 0.79% 2.85% 0.00% 0.50% 1.00% 1.50% 2.00% 2.50% 3.00% 1 10 100 1,000 10,000 100,000 1,000,000 CTR #ofImp./Clk. Perf. Campaign 0.000% 1.000% 2.000% 3.000% 4.000% 5.000% 6.000% 7.000% Clk v.s. Conv 0.746% 3.646% 6.386% Clk v.s. Conv Normal 0.746% recm_1st lvl. 3.646% recm_2nd lvl. 6.386% Game App DL Clk v.s. Conv. After our 2nd lvl optimization, the conv. v.s. click increased 8.56 times
  • 21. Step5: continuous monitoring and improvement 10,037,003 2,451,061 85.01% 81.29% 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00% 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 No-Optimization Optimized Target%(Perf.) Imp.consumed(Cost) Imp. Consumed (Cost) Targeted % (Perf.) 0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 1.40% 1.60% CTR CVR IVR 0.90% 1.53% 1.37% 0.88% 0.57% 0.50% Optimize Normal CTR = Click v.s. Impression CVR = Click v.s. Conv. IVR = Imp. v.s. Conv. Conv. Rate increased 3 times Cost Optimization: Cost reduced more than 75% while performance only decreased 3.72%
  • 22. Implementation  Hadoop MapReduce as computing platform  Using Hadoop streaming with Python  Map: a list of ad pairs as input for similarity caculation  Reduce: simply aggregate the map results  Re-modeling on a daily basis based on results  Will go on to use Haoop HDFS + Spark + Python for performance benefit
  • 23.
  • 24. Summary  Build the infra. that proves models effective or not as early as possible  AB testing for new models  Automate as much as possible  Monitoring and measurement  Computing resource  Properly manage Product, ad-hoc, analysis jobs  Optimization does work  Use Python wherever it fits