Transcript
SEM Performance with
@BethThouin, VP Marketing at Acquisio
Machine Learning
We empower campaign managers to thrive in the changing digital economy.
We simplify online advertising and optimization with data. Our award-winning machine learning algorithms perform millions of campaign adjustments per day on 300,000 campaigns for hundreds of clients,
enabling better advertising tracking, optimization and reporting.
AdWords - Facebook - Bing
• What is Machine Learning? • Applying Machine Learning to SEM• Data points• @acquisio
Agenda
Machine Learning and AI are disrupting work being done by
humans
It's being called "The Fourth Industrial Revolution"
First RevolutionSteam, water,
mechanical production equipment
Second RevolutionDivision of labour, electricity, mass
production
Third RevolutionElectronics, IT,
automated production
Fourth RevolutionCyber-physical
systems
Machines are faster, "smarter" and never tire
ML already delivers better marketing outcomes for advertisers of all sizes
Better campaign performance, even for low budget/volume
spends
Automated workflows that dramatically improve operating
efficiencies
Decreased OPEX, Higher customer sat, LTV and lower
churn rates
So HOW does ML work with SEM?
1 Money Management
Bid & Budget Management across platforms (Google/Bing/FB/Yahoo), campaigns, devices, geo, time of day, day of week, GDN topics, and GDN placements, RLSA, and Interactions.
Learning from each individual advertiser and benchmarking per vertical and even per geo.
2 Suggestive ‘Don’t make me think’Suggestions for account audits, keyword selection, ad copy, ad testing, audience segmentation, campaign structure, missed opportunities.
These will all ultimately have to be fully automated.
So how much does ML impact SEM?
ML saves campaign managers approx 65%* of management time. Freeing
up resources for more important work.
*With automation, CMs spend 20 minutes vs 60 minutes per campaign per week.
On average, ML increases clicks by
33%.
N = 12,000 accounts running 3 months+ on Acquisio’s Machine Learning technology. March 2017.
On accounts set to optimize for conversions, ML lifts conversions by
100%.
N = 3,000 accounts running 3 months+ on Acquisio’s Machine Learning technology. March 2017.
Does size matter?
Smaller accounts see better LTV improvements
on ML, but larger accounts are more likely
to see bigger budget attainment
improvements.N = 36,000 accounts running 3 months+ on Acquisio. March 2017.
Adwords = 34.6% improvementBing Ads = 16% improvement
LTV / ARPU
2.5X more likely
to hit budget targets on AdWords and 7X more likely on Bing
Ads.*<500$ monthly ad spend
ML-based campaign automation vs rule-based campaign automation
4X more likely
to hit budget targets on AdWords and 15X more likely on Bing
Ads.*>1500$ monthly ad spend
We were able to increase margins from 10 - 15% to 20 - 25% using ML automation. We use machines and humans in tandem to better service customers. Machines perform complex tasks around bidding, budget management and optimization. Account managers are able to focus on things like generating ad copy, landing pages, keywords and otherwise applying their industry knowledge to service customers. ~ Chris Hanson, 3GEngagement
“
What does the future hold for advertisers and agencies? Automation for every routine campaign management task and more… such as keyword selection, audience optimization, even writing ad copy and customer service.
• @BethThouin• VP Marketing @Acquisio• bthouin@acquisio.com
Thank you!
top related