Testing Automated Bidding with Google’s Maximize Conversions

It’s easy to sit around and theorize about bidding strategies. It’s even easier to switch a bidding strategy only to revert it back after a week of less-than-stellar results.

If you’re going to test a bidding strategy, the first step is to do your research. Make sure your chosen strategy is the right one to test, and set your expectations accordingly. In this post, we’ll dive into one Google Ads strategy: Maximize Conversions automated bidding. I’ll share the results of our three-month test, what we learned, and when it makes the most sense to use it.

What Is the Maximize Conversions Bidding Strategy?

Automated bidding was introduced by Google back in 2010 with the release of Enhanced Cost-Per-Click, and Google has been encouraging advertisers to adopt its ever-growing array of automated features ever since. In Google’s words, automated bidding “takes the heavy lifting and guesswork out of setting bids to meet your performance goals” by utilizing advanced machine learning.

Maximize Conversions as an automated bid strategy came about in May of 2017. It claims to factor in multiple signals such as remarketing lists, time of day, and browser, coupled with machine learning technology to adjust bids for each auction based on the likelihood of a conversion by that user.

Is Maximize Conversions the Right Strategy for You?

In theory, Maximize Conversions sounds ideal if conversion count is the end goal. However, there are instances where this could certainly be the wrong bidding strategy for your campaign. If any of the following KPIs are your primary metric, Maximize Conversions might have disastrous results:

  • CPA – Consider a Target CPA strategy, as Maximize Conversions could drive that number in the opposite direction.
  • ROAS – If revenue is variable and you’re able to pass accurate revenue data to Google Ads, consider Target ROAS over Maximize Conversions.
  • Volume – Maximize Conversions can limit your reach to only those likely to convert.
  • Max CPC – Maximize Conversions offers no control over how much you bid for a click.
  • Budget – Straight from Google’s own support page: “Maximize conversions will try to fully spend your average daily budget, so if you’re currently spending much less than your budget, Maximize conversions could increase spend significantly.”

If the total number of conversions is the primary metric you’re chasing, and you have some wiggle room in the areas listed above, Maximize Conversions may be just the ticket.

Putting Maximize Conversions to the Test

We tested Google Ads’ Maximize Conversions bidding strategy on one of our B2B clients that sits in a somewhat niche industry. This client’s focus is on improving total conversions year-over-year and keeping a target CPA of $100, though they are willing to go up to $150 if the total number of conversions warrants it.

On November 17th of 2019, we traded in our trusty Target CPA of $100.00 to a Maximize Conversions bidding approach. In terms of volume, here’s how the campaign performed the three months before swapping the bidding strategy, and the three months following the swap. Our bidding strategy change is marked with a pink line:

The line graph here shows a jump in both clicks and impressions once the bidding strategy was swapped to maximize conversionsThe line graph here shows a jump in both clicks and impressions once the bidding strategy was swapped to maximize conversions

In terms of conversions and cost, here’s that same date range with the same pink line:

Again, the line graph shows a jump, this time in terms of conversions and cost once the bidding strategy was swapped to maximize conversionsAgain, the line graph shows a jump, this time in terms of conversions and cost once the bidding strategy was swapped to maximize conversions

Our date marker is hardly necessary in either of these graphs. We can see that in just the first week, our metrics jumped exponentially, as laid out in the table below:

This chart shows within the first week of switching our strategy, impressions went from 1,145 to 4,050; clicks went from 98 to 175; conversions went from 2 to 13; and cost went from $245 to $596.This chart shows within the first week of switching our strategy, impressions went from 1,145 to 4,050; clicks went from 98 to 175; conversions went from 2 to 13; and cost went from $245 to $596.

But machine learning takes time, and automated bidding strategies need a few days to, well, learn. So let’s remove the initial week of trial-and-error on Google’s part (the week of November 18th), and compare our three-month periods to each other sans volume spikes.

After removing the initial week, comparing three months of performance to the previous three months, we saw an 189% increase in impressions, 71% increase in clicks, 135% increase in conversions, 108% increase in cost, and a 12% decrease in CPA.After removing the initial week, comparing three months of performance to the previous three months, we saw an 189% increase in impressions, 71% increase in clicks, 135% increase in conversions, 108% increase in cost, and a 12% decrease in CPA.

The Results

In the above comparison table, we can see that not only did conversions increase from 26 to 61, but our cost-per-conversion surprisingly dropped from $95.12 to $84.13. And, despite our expectation of an overall drop in volume, the additional (and unplanned) $2,659 in cost resulted in a jump for both impressions and clicks, of 189% and 71%, respectively.

So, Does Maximize Conversions Work?

Maximize Conversions certainly worked for this particular campaign. Of course, we would be remiss if we didn’t point out that the 135% increase in conversions and the 12% decrease in cost-per-conversion came coupled with a 108% increase in cost. If our budgets weren’t as flexible as they were, we might have had to have an awkward conversation with our client.

We also didn’t report on the effects on other potentially important KPIs, such as cost-per-click or impression share, for example. The purpose of Maximize Conversions is to… well, maximize the number of conversions you’re getting. If cost-per-click is paramount, then using an automated bidding strategy that essentially gets rid of your maximum bids is likely not the best idea.

The Takeaway

There are many machine-learning powered bidding strategies in Google Ads; bid strategy can have a significant impact on how (un)successful any given campaign is. Ultimately, your bid strategy should be determined by the top priority KPI. And if that KPI is the total number of conversions, Maximize Conversions may be the right bidding strategy for you.

Just remember to do your research and set expectations before you get started.

The post Testing Automated Bidding with Google’s Maximize Conversions appeared first on Portent.


Leave a Reply

Your email address will not be published. Required fields are marked *