Google Develops Data-Driven Simulation Tool To Measure Automated Bidding
Google Develops Data-Driven Simulation Tool To Measure Automated Bidding
Google on Tuesday announced it will launch a simulation tool for eligible advertisers. The tool will provide insight into how automated bidding would affect a live data-driven attribution campaign. The measurement window is seven days.
The goal is to help marketers see how data-driven attribution will affect an account’s performance before deciding to switch to data-driven models.
The goal is to make advertisers feel confident in their decision.
Charles Huyi, group product manager of attribution at Google, wrote in a post that the company plans to bring data-driven attribution to more advertisers and more ad types.
Historically, data-driven attribution has supported Search, Shopping, Display and YouTube ads.
Huyi explains in the post that Google will expand support to app conversions and begin supporting Discovery formats — including those in Performance Max — next year.
As data seems to be disappearing, Huyi said Google is committed to helping advertisers more accurately measure campaign goals, and to provide tools needed to succeed as advancements in machine learning and automation attempt to deliver positive results.
Advertisers that switch to data-driven attribution from another attribution model typically see a 6% average increase in conversions, according to Google data.
In a data-driven attribution model, machine-learning algorithms assign fractional credit to customer touch points that may have previously been undervalued. Smart Bidding can then react to these opportunities, resulting in performance gain, Huyi wrote.
Mercedes-Benz Germany, which upgraded all of its conversions to data-driven attribution and scaled out the use of Smart Bidding across all campaigns, ran tests with Google in Smart Bidding.
The company originally ran conversions using a last-click attribution model. An experiment across multiple campaigns tested several bidding strategies such as Maximize conversions or Maximize clicks to examine how they would perform with data-driven attribution.
Huyi wrote that the six-week experiment using data-driven attribution with Maximize conversions saw a 37% increase in conversions.
(37)