Automation is Google’s innovation

If you haven’t realised, machine learning is already everywhere. From Dominoes food preparation timer to Spotify’s unique recommendations, today there is no escape. Google is using machine learning in many of its products. One of their latest innovations is smart bidding.

What is machine learning?

There is a general confusion that Artificial Intelligence (AI) is the same as machine learning. The reality is that machine learning is a technique used to develop AI. Therefore, one step below AI.

Machine learning, in brief terms, is when computers can learn and act as humans do. Through understanding algorithms and how users behave and then perform certain functions. They give marketers more time to focus on more strategy-based work.

We are already using machine learning with Google products: Google Translate (with camera), Google Photos & Google Trips. But now we can incorporate it into our marketing optimisation.

Why should you care about smart bidding?

Smart bidding will mean marketers will spend less time focusing on bidding, manually adjusting thousands of keyword bids. And instead, let Google control this process. This will not only allow testing to occur continuously, learning the best bid is for each user. But will allow marketers to dedicate more of their time to creating engaging copy and formulate effective strategies.

In summary, smart bidding is valuable because…

  • True auction-time bidding
  • Adaptive learning at the query level (ML algorithms rely on robust conversion data)
  • Richer contextual signals and cross signals analysis


4 Smart Bidding strategies

  1. eCPC (enhanced Cost Per Click)
    This strategy takes the implemented keyword-level bids and bid adjustments into consideration while providing auction-time bidding capabilities. It automatically adjusts manual bids based on each click’s likelihood to convert.
  2. If you are not ready for Google to take full control of your bidding strategy than we recommend this strategy. Allow Google to adjust your bids, set your maximum CPC, and attract traffic to your site.

  3. Maximise Conversions
    This strategy automatically sets bids for every auction to get the most conversions for the campaign while spending the provided budget. By adjusting bids for every auction, it will be more accurate, which will lead to an increase in conversions and a decrease in CPA.
  4. We would recommend this if you do not have a target CPA, are constantly maxing your budget yet want to drive as many conversions as possible.

  5. Target CPA (Cost Per Acquisition)
    This bidding strategy lets the algorithms predict the conversion rate outcome for a click in each auction based on the relative signals present. The goal of this strategy is to maximise conversions while meeting the target CPA.
  6. If you have a set CPA in mind, then this bidding strategy will help drive conversions by optimising your bids to reach that target.

  7. Target ROAS (Return On Ad Spend)
    This is the only bidding strategy to predict conversion values for every single search auction. It is available as either a standard strategy for a single campaign or a portfolio strategy across multiple. A portfolio with multiple campaigns will likely perform better than the individual campaigns as the budgets can be cross-allocated.

If you are a larger company with a specific ROAS target, then implementing Target ROAS bidding strategy will work effectively.


Automation is here to help us marketers improve our campaigns. Through dedicating more time to high-level tasks and less time to manually adjusting. Depending on what needs and goals you might have, smart bidding will help optimise your campaigns to increase performance.

We hope you have been enjoying our Google Series so far. If you missed out on our first blog, we summarised all the updates released during the Google Marketing Live 2018. If you are wanting to optimise your ads for the always-on device, then stick around for part 3.