Automating Google Ads anno 2019 - where are we?

Letting Google handle bid calculations & campaign creation used to be a daunting idea, but since Google’s algorithms have become smarter over time, adopting these automation tactics offer a huge competitive advantage.

2019 therefore promises to be a challenging year for digital advertising experts as they need to focus even more on strategic decisions while relying on Google’s automation technologies such as smart bidding & smart campaigns

We will zoom into these two topics in the upcoming blog posts, kicking off with smart bidding. 


What is smart bidding?

Smart bidding strategies are solid strategies combined with clever use of automation: machine learning algorithms train on data at a wide scale to get predictions that are more accurate across your account. This way, the impact of different bids might affect conversions or conversion value. More importantly, these algorithms take into account a vast range of parameters that might influence performance - more than any person could ever process. 
The most common signals Google takes into consideration are:

  • Demographics
  • Device
  • Geographic location
  • Browsing history
  • Account performance
  • Audience membership
  • User/site history

For each auction, Google’s smart bidding algorithms will evaluate any or all of these signals to determine the potential value to your business of appearing with an ad. Based on this evaluation and the criteria you defined, Google will customize a bid for each and every auction. 

Types of smart bidding

Enhanced Cost Per Click (CPC) has been around for a good few years and has been implemented by a lot of marketers. However, the true capability of using Smart Bidding is something marketers are only just beginning to grasp onto and utilize over the last two years. During that period, bidding strategies have become even more robust and even smarter, making them the ideal partner in the online marketer’s toolbox.
Let’s take a deep dive into the different smart bidding strategies:
  1. Target Search Page Location: target your ad either at the top of search results or on the first page
  2. Target Outranking Share: automatically outbid your competitors with the goal to appear above their ads
  3. Maximize Clicks: choose this model if you aim to get the most clicks possible out of a certain budget
  4. Maximize Conversions: automatically set your bids to get as many conversions as possible within your campaign’s budget. By choosing this strategy, you don’t aim to have a particular cost per conversion.
  5. Target Cost Per Acquisition (tCPA): Automatically sets your bids to get as many conversions as possible within your CPA goal, without a specific target on volume of conversions.  You no longer need a minimum amount of conversions to start with this strategy, a notable change within the last 2 years
  6. Target ROAS (tROAS): Set your bids to generate as much revenue as possible within your target.  You need at least 15 conversions passing conversion values through Google Ads within 30 days.
  7. Maximize conversion value: this strategy is a fit if you want to maximize the total conversion value of your campaign within your specified budget. This is the default bidding strategy used for smart shopping campaigns. 

Following the deprecation of average position, Google will retire the Target Search Page Location and Target Outranking Share automated bidding strategies at the end of June 2019.

Avoid these common smart bidding mistakes

Keep in mind that not having a strategy at all can backfire in the end. You need to know the possibilities as well as the pitfalls and prerequisites of each bidding strategy before applying a smart bidding strategy at all.
Let us guide you through the possibilities. 

Tip 1: choose a non-last click attribution model

Before using smart bidding strategies, one of the first steps to undertake, is choosing a solid attribution model. This way the right credit is assigned to the different campaign types (upper/lower funnel).  Having a clever attribution model in place will therefore contribute to the effectiveness of your bidding strategy.  
In general, the advice is to choose a non-last click attribution model, with a high preference for a data-driven attribution. 

Tip 2: Cool kids experiment, so do cool marketers

Before applying smart bidding immediately to your campaign(s), we advise to set up an experiment. With advancements in smart bidding, choosing automation produces better results if you have tested enough.

Tip 3: Avoid initiating a new learning period

 Don’t change any bids, budgets or ad copy during the learning period. Let the Google algorithm take its time to learn & optimize. Google needs around 7 days to learn how to use your budget and then it is out of the ‘learning’ phase.


Tip 4: Start with a realistic tCPA & gradually decrease to your ideal tCPA

Take the last 30 days into account & ignore the last 7 days (depending on your conversion window). You can consider this tCPA as your average tCPA to set when starting; afterwards you can gradually decrease your tCPA to your desired tCPA.

Tip 5: Choose your smart bidding strategy based on your goals, not on luck 

This rule not only applies to what strategy you should choose, but also to the reporting metrics you use to evaluate performance. Let’s take the example of a retailer that has always been reporting on sales, but also on impression share when evaluating campaigns. Using particular smart bidding strategies focus will generate better sales numbers, but might negatively affect your impression share because only qualified users will get to see an ad.

Tip 6: campaigns should not be budget constrained 

This applies in particular to tCPA & tROAS: these bidding strategies will look at the totality of the auctions it could target, without any budget awareness. This implies that applying these bidding strategies to campaigns that are limited by budget, might lead to a ROAS that is below your objective.  

Tip 7: Smart bidding for seasonal times 

Despite using auction-time bidding, the algorithm can’t predict short spikes in traffic and conversions (Black Friday, Cyber Monday, etc.). If you expect your conversion rate to change from 2% to 4%, the targets of your smart bidding strategies will need to change accordingly. 

Smart bidding can also drive store visits

Smart bidding strategies are a great opportunity for businesses with an omnichannel approach. That’s why bidding strategies such as target ROAS or maximize conversions now also include store visits, using all of the same signals used for online conversions.  In effect, this means optimizing for total omnichannel revenue or omnichannel conversions at the advertiser's desired target omnichannel goals. Once the algorithm knows the value of both online and offline conversions, it will take into account both the value and probability of the conversion to optimize bids. 


Google has been moving forward with more and more automation for Google Ads, with more pushes quickly coming our way in 2019. One of the biggest, and oftentimes biggest black box to marketers, is automated smart bidding. However, there’s really no better time than now to set your old ways aside and give one of these new methods a try.
Google’s machine learning is giving us as marketers the tools to work faster, smarter, hit our goals & focus on strategy and eventually accomplish more in less time.   

Are you convinced of the possibilities automation has to offer, but you still like some advice from our experts? Lets talk! It’ll be our pleasure to make your project a success. 

Don't miss out

The Reference has its office in the heart of Manhattan.
“I want to wake up in that city that never sleeps, and find I'm king of the hill, top of the list, head of the heap” – Frank Sinatra