Report: The future of paid-search marketing is machine learning and AI

Machine learning-optimized campaigns saw 71 percent better conversion rates and lower CPCs than those not using it.

Chat with SearchBot

Paid Search 165661541 Ss 1920

Acquisio released a report (PDF) this month that argues machine learning algorithms can dramatically improve paid-search marketing performance vs. accounts that don’t use them. Acquisio calls its suite of machine learning tools “Acquisio Turing,” after British computer scientist Alan Turing.

The company looked at thousands of campaigns over a two-year period, comparing results from machine learning-supported accounts with those not utilizing machine learning. Acquisio concluded, “Across all business types and budget sizes globally, the accounts using Acquisio’s machine learning technology are outcompeting those that aren’t, giving advertisers an advantage.” This is especially true at low budget levels.

More than 30,000 accounts were analyzed and compared in the report.

Acquisio says that just over a third of its customers manage small or local accounts (e.g., media publishers that provide SEM to small business advertisers) and “just over 50 percent our client base is made up of marketing organizations servicing local markets, with a large portion of them focusing on the automotive market.”

About 60 percent of the company’s customers serve North America. The remainder of customers and accounts are distributed across regions, with a larger concentration in Australia.

Acquisio reported that machine-learning accounts saw improved metrics across the board, including higher conversions, lower CPCs and higher click-throughs:

  • Conversions increased by 71 percent on average
  • CPCs decreased by 7 percent on average
  • Clicks increased by 15 percent on average
  • 64 percent of accounts saw a lower CPA

These accounts were also better at achieving “budget attainment” (reaching their target spend levels) vs. accounts not utilizing machine learning optimization.

The company added that machine learning-supported accounts also experienced lower churn on the platform than those not using machine learning, and this was especially true for accounts operating at less than $500 per month. “Businesses with the lowest spend actually had the highest success in terms of lifetime value,” the report states.

I would expect to see more studies and results like these in the near future. It thus seems inevitable that automation will replace humans in most areas of paid-search marketing.


Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.


About the author

Greg Sterling
Contributor
Greg Sterling is a Contributing Editor to Search Engine Land, a member of the programming team for SMX events and the VP, Market Insights at Uberall.

Get the must-read newsletter for search marketers.