Is Google’s “Smart” Bidding Smart For You?

Robert Brady

Robert Brady

Google Ads offers four types of "smart" bidding that leverage machine learning in your account, but are they right for you? Let's talk about it.

Big tech companies are talking a lot about machine learning and AI right now. And if you believe the hype it seems like everything from curing cancer to world peace will be solved once we get enough computing power and the right algorithm. While most of that is still in the distant future, for advertisers using Google Ads, the future is now.

What Is Smart Bidding on Google Ads?

Smart bidding is the term Google uses for four specific automated bidding strategies; Target CPA, Target ROAS, Maximize Conversions, and Enhanced CPC. These bidding strategies utilize machine learning to optimize several factors in accordance with the goal of the specific bidding strategy. Let’s take a moment to explain what each of these bidding strategies actually does:

  • Target CPA – This bidding strategy takes a desired CPA from the advertiser, say $50, and will try to achieve as many conversions as possible while meeting this average CPA. The machine learning is specifically looking at historical information from your campaign as well as auction-time signals from the user to determine when and how to show ads.
    • Remember: Device bid modifiers still function with this bidding strategy, but will change to modify the Target CPA. So if you wonder why your Target CPA isn’t being met it’s wise to first check your device bid modifiers.
  • Target ROAS – This bidding strategy requires advertisers to have values for their conversions (either static or dynamic will work) and will operate much like Target ROAS above. I see this more frequently used with ecommerce advertising where the conversion is a transaction.
    • Remember: A campaign must have 20 conversions in the past 45 days to be eligible, but Google recommends 50 conversions in the past 30 days for best results. And when the platform gives a recommendation for best results you’d be crazy to try their machine learning with less data.
  • Maximize Conversions – This bidding strategy automatically adjusts bids to get the most possible conversions while using your entire budget. This strategy only works for campaigns with their own budget.
    • Remember: Google has two goals with this strategy; spend all your budget and get the most conversions possible. In that order. If you enable this for a campaign that hasn’t historically spent all it’s budget you will see cost go up. This is especially true in campaigns with limited impression volume because if the clicks & impressions aren’t there then Google machines are smart enough to increase your bids until the number of clicks takes your full budget.
  • Enhanced CPC – This is the simplest bidding strategy as it basically layers on top of manual CPC bidding. It will adjust CPC up or down based on the likelihood of a conversion (here’s where the machine learning has to earn it’s keep) but will strive to keep the average CPC at or below the max CPC.
    • Remember: Enhanced CPC takes effect AFTER applying all bid modifiers. If you’ve got a lot of modifiers in place already, turning on eCPC can really make determining your actual max CPC a mathematical riddle.

Now you have a basic understanding of what each bidding type does. So what exactly is that machine learning looking at? Let’s talk about it.

Contextual Signals Used By Smart Bidding

Google Ads offers several areas where you can manually set bid modifiers. These include device, location, time of day, day of week, and others. However, Smart bidding goes even further to include additional contextual signals. We don’t have the time to go through all of them, but here is Google’s list:

Google Ads Smart Bidding Contextual Signals

As you can see, this unlocks a lot of potential factors for the automated bidding strategy to use, with even more coming in the near future. Keep in mind that bid adjustments are using the historical performance (and current performance) of the campaign to determine which, if any, of these factors should be influencing bids and in which direction.

And this all brings us to the ultimate question.

Is Google Smart Bidding Smart For You?

The answer for this question will vary for each advertiser (that’s a clever way of saying “it depends…”) based on a lot of factors. However, there are a couple of factors that I think should carry the most weight in your decision:

  • How reliable is your conversion data? You need to ensure that the conversions you track in Google Ads (and enable for usage with automated bidding) are the conversions that align with your business goals. It won’t matter if Google Ads produced 1,000 leads if those leads don’t turn into signed contracts for real revenue. I highly recommend an audit of your conversion tracking and settings before enabling Smart bidding in any form.
  • How much conversion data do you have? As I mentioned for Target ROAS above, Google often has recommended levels of conversions for bidding strategies (if you dig hard enough into their support documentation). If your campaigns don’t have enough data then automated bidding can do really weird things, most of them bad. Especially keep in mind any seasonality you might have in your business. You might have plenty of data during peak season, but in the slow season does it drop below the recommended level?
  • How much time can you dedicate to your Google Ads account? If you’re majorly crunched for time then Smart bidding can appear as manna from heaven. However, just because Google says they’re handling things doesn’t mean you can ignore it. You still need to be checking on your campaigns regularly, though you will likely need less time to dig deep unless things are going haywire.

Overall I think that Smart bidding is a trend of the future. Google is constantly improving their machine learning and the more data they crunch, the better the results will be for advertisers. Just make sure that you’re implementing them in your account based on your specific situation. Good luck!

Comments

Other PPC journal entries

PPC (Pay Per Click), or the process of helping businesses drive traffic to their site from places like Google, LinkedIn, Facebook, etc., has proven to be one of the fastest ways to get the results you need to grow your business. These articles are focussed on the best practices of online advertising.

Robert Brady
  • PPC
  • 01.15.2019

Testing Responsive Search Ads in Google Ads

Google Ads is always testing new formats for ads that appear on search results pages. In the

Read more
Robert Brady
  • PPC
  • 11.12.2018

Facebook Drops 5000 Targeting Options

Facebook has been facing immense pressure concerning their advertising platform. The Cambridge

Read more
Robert Brady
  • PPC
  • 08.24.2018

Big Changes Coming To Google Display Network Campaigns

Over the years Google has built a massive network of websites that display advertising. These ads

Read more
Robert Brady
  • PPC
  • 06.22.2018

Old AdWords Interface Will Be Retired on July 10th

Ever since the announcement of the new AdWords interface at Google Summit in 2016, there has been an

Read more
Subscribe to recieve monthly advice and
best practices on design, marketing, and more.