Technical Revolution of Digital Marketing Bryan M. Minor, Ph.D. Chief Scientist, Acquisio Performance Media Platform
Technical Revolution of Digital Marketing
Jul 02, 2015
While most of us spend our days toiling away in the deepest details of paid search campaigns, scientists of paid search are running experiments, testing regression models and analyzing larger trends to unlock the fundamental secrets of paid search auctions, attribution, economics and search behaviors. Distinguished marketing scientist, Bryan Minor, presents his latest discoveries, insights and knowledge.
Slide notes:
Slide 2
1. Constantly using the best current information to adjust current Bid.
2. Produces superior precision to native AdWords daily budget control.
Slide 3
Compared to a canon with fire and wait model.
Slide 5
This is CPC per day for the Budget Group on x axis, and Clicks/day for BG on y axis
Slide 6
True for clicks or conversions, all campaigns in Budget Group with same goal, either maximize clicks or conversions.
Slide 7
Ellipse points are B, Hexagon are C, with time stamp numbers sequencing up on each.
Slide 8
Easily seen that with BBM (big dot centroid) is more clicks at a cheaper CPC.
Slide 9-12
Yellow date is the first dates with BBM.
Slide 13
Yellow are all cases where obey CPC constraint, Red they are disobeying. Middle graph good statistical significance, right graph poor statistical significance.
Slide 14
Budget Survival is the percentage of the day the Budget Group is running. Red are below 80%, yellow higher. Right graph has good statistical significance, Left graph has very poor.
Slide 15
Budget Spend = (Spend/(Planned spend)) for Budget Group. Yellow are good cases, red poor. Right graph has good statistical significance, Left has poor.
Slide 16
86% of money spent well obeying the constraints, while 14% does not.
Slide notes:
Slide 2
1. Constantly using the best current information to adjust current Bid.
2. Produces superior precision to native AdWords daily budget control.
Slide 3
Compared to a canon with fire and wait model.
Slide 5
This is CPC per day for the Budget Group on x axis, and Clicks/day for BG on y axis
Slide 6
True for clicks or conversions, all campaigns in Budget Group with same goal, either maximize clicks or conversions.
Slide 7
Ellipse points are B, Hexagon are C, with time stamp numbers sequencing up on each.
Slide 8
Easily seen that with BBM (big dot centroid) is more clicks at a cheaper CPC.
Slide 9-12
Yellow date is the first dates with BBM.
Slide 13
Yellow are all cases where obey CPC constraint, Red they are disobeying. Middle graph good statistical significance, right graph poor statistical significance.
Slide 14
Budget Survival is the percentage of the day the Budget Group is running. Red are below 80%, yellow higher. Right graph has good statistical significance, Left graph has very poor.
Slide 15
Budget Spend = (Spend/(Planned spend)) for Budget Group. Yellow are good cases, red poor. Right graph has good statistical significance, Left has poor.
Slide 16
86% of money spent well obeying the constraints, while 14% does not.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Technical Revolution of Digital Marketing
Bryan M. Minor, Ph.D.Chief Scientist, Acquisio
Performance Media Platform
Continuous SEM Optimization
Performance Media Platform
Features:
1. Examination and adjustment of Bids in regular intervals many times per day
2. Examination of Budget spend precision many times per day with hyper accurate control
3. Updating of modeling parameters in algorithms on a longer characteristic time scales
4. Auto detection and dealing with anomalies
Results in accelerated learning and optimization
Algorithm Model
• Cruise missile model
• Dynamic Non-linear optimization
• Small steps more often
Performance Media Platform
Problem statement:
Performance Media Platform
• For a fixed Budget for budget period (month)
– With a group of Campaigns (Budget Group)
• Make Daily Budget last whole Day
• Maximum Average CPC per day limit
• Fairly compete Campaigns based on value of Clicks (conversions)
• Maximize Clicks (conversions)
Performance Media Platform
Theoretical ABC graph
C B
A
A
A
minCPC
0 2 4 6 8 100
5
10
15
20
25
30
35
CPC
Cli
cks
day
ABC graph explanation:
Performance Media Platform
• B graph – Daily Budget spent
• C graph – Daily Budget not spent
• A – location of maximum number of Clicks for a fixed Daily Budget obeying constraints
• minCPC – Lowest value of CPC produces Clicks
Constantly searching for Optimal solution
Performance Media Platform
Experimental ABC data #1
ABC Data Analysis
Performance Media Platform
• BBM start: 24 May 2014
Performance Media Platform
Results: X-graphs #1
Start Clicks Start CPC End Clicks End CPC
1,066 $0.51 1,848 $0.27
• BBM start: 26 May 2014
Performance Media Platform
Results: X-graphs #2
Start Clicks Start CPC End Clicks End CPC
490 $2.07 892 $1.21
• BBM start: 23 Apr 2014
Performance Media Platform
Results: X-graphs #3
Start Clicks Start CPC End Clicks End CPC
29 $1.24 56 $0.63
• BBM start: 23 Apr 2014
Performance Media Platform
Results: X-graphs #4
Start Clicks Start CPC End Clicks End CPC
23 $2.08 76 $0.64
CPC Constraint Obedience
Performance Media Platform
Budget Survival
Performance Media Platform
Budget Spend
Performance Media Platform
Spend CPC Constraint Obedience
Performance Media Platform
Conclusions
Performance Media Platform
• ABC theory validated – A exists!
• Continuous updating of Bids produces superior results
• Precise Budget spend control over Budget Period (2%)
• Statistical Significance matters (143 issue)
• Constraint matters
• Ability to deal well with dynamic changes Budgets
Constraints
Creative changes
Google Settings changes
Google algorithm changes
Related Documents
