Google’s Machine Learning: The Case for Smart Bidding

May 7, 2019
Posted in Paid Search
May 7, 2019 Jen Lopez

Google’s Machine Learning: The Case for Smart Bidding

Ever sit and wonder about today’s tech advancements, and how they were once dubbed as just “make-believe” in our favorite movies? It’s amazing how times have changed, and with the rise of artificial intelligence (AI), so many things have become SO much easier.

If I need the temperature changed on my air conditioner and my hands are full, I can ask Alexa. If I’d like to listen to some relaxing music before bed, Siri will handle it for me. If I am looking to find out when my local grocery store closes, I can have Cortana do a search for me and report back. And, as much as people like me enjoy the advancements, there are always the naysayers, or those slower to adopt the change.

In fact, though I have a full smart home and the ability to start my car from an app on my phone, I was at first very hesitant to allow AI into my workplace. As a PPC Manager for some of the world’s largest and most successful brands, testing AI made me nervous, as it felt like I would be giving up control and allowing my Google Ads campaigns to rely on something I had little insight to.

Enter: Smart Bidding for Google Ads.

This is the science behind Google’s real-time bidding (RTB) technology that analyzes 70 million unique signals at the exact moment a web page loads, in order to set the ideal bid for your prospect on Google Ads.

Sounds cool, huh?

Well, it is, and after enough testing and tweaking, I am finally a true believe in Google’s Smart Bidding Technology.

To best understand how the auction-time, multi-signal bidding works, it’s important to understand all the exclusive signals that Google has access to in Search and Display that you don’t always see in your campaign reports. Here are just a few of the things that Google is considering while adjusting your query-level bids in real-time:

  • OS
  • App
  • Browser
  • Ad Creative
  • Language
  • Actual Query
  • Search Partner
  • Location
  • Time
  • Re-marketing List
  • Device
  • User Behavior (sites previously visited, cross-device usage)
  • Demographics (age, gender)
  • Ad format
  • Ad performance
  • Onsite Behavior (how recently a user left your site, how many pages they viewed, values of the products a user viewed on your site)

Prior to using Google’s Smart Bidding, like most PPC Managers, I was using a combination of manual bidding and automated rules. However, as Google Ads became more complex, and the signals above were harder for my team to track, manage and react to, I heard a quote that resonated with me. It was, “Infinite permutations call for automation,” The quote was directly out of a Google presentation on AI, and sitting in that room, alongside other marketers, I knew there was something to this.

With so many placements, messages, ad formats, micro-moments, and all of the other pieces of the puzzle I was trying to focus on to get a prospect to convert, testing Smart Bidding (a HUGE piece of said puzzle), allowed me to focus on more of what the user sees from the brands I manage. A user doesn’t see what I am bidding to get their attention; they see how I am speaking to them, what I am asking from them, what I am willing to give in return, and so much more. That is where I want to focus my attention, not on setting bids which is less visible to my prospect.

Interested in learning more about Smart Bidding and how all these signals come into play?

Stay tuned for the next installment of this series, where we will walk through some of my favorite bid strategies available in Google Ads. We’ll learn about strategies that focus on visibility, conversions, revenue and clicks.