AMP

Automated Bidding: What to Do When Smart Bidding Does Not Perform

Published on

AI-driven automation in big ad platforms like Google and Facebook is inevitable.

It is slowly but surely becoming a pivotal facet of digital marketing that SaaS marketers need to know how to do – and do it successfully.

But what happens when your campaigns are failing, and who is to blame?

Let’s take a closer look.

How Does Automated Bidding Work?

Automated bid strategies use machine learning to maximize results based on campaign goals. To do this, they analyze data such as a user’s demographics, device, operating system, and location.

With automated bidding, you rely on Google to adjust your bid in real-time, instead of manually selecting your maximum CPC for all of your ads.

Google’s automated bid strategies are portfolio bid strategies, grouping together multiple keywords, campaigns, and ad groups. Consequently, this allows you to apply your chosen strategy across all of your keywords, campaigns, and ad groups.

Is Smart Bidding the Same As Automated Bidding?

Smart bidding is a buzzword that comes up a lot when discussing automated bids.

While it’s often used interchangeably with automated bidding, it is not quite the same.

In a nutshell, smart bidding is a subset of automated bidding. Currently, it includes four conversion-based strategies – Target CPA, Target ROAS, Enhanced CPC, and Maximize Conversions.

To maximize conversion results, smart bidding tracks and analyzes data from search and clicks. Then, it adjusts the bid amount based on whether or not the algorithm believes a click will result in an actual conversion. Simply put, the bid will be set higher for searchers who are more likely to make a purchase/convert.

When Should You Use Automated Bidding?

For many SaaS marketers, automated bidding is a way to improve performance.

Google Ads’ algorithm uses signals present at auction time to tailor bids to a user’s search context, which frees up time from complex tasks such as manually establishing intent and setting bids.

Which automated bidding strategy will work for your business depends on a number of factors, including your end goal.

For SaaS companies with tiered, licenses- or usage-based pricing models, shifting to a value-based bidding strategy can be especially beneficial. This can transform your conversions into almost e-commerce products, allowing you to optimize for a specific required return on each individual conversion based on its expected value.

Automated Bidding Pitfalls to Watch Out For

When it comes to automated bidding, one of the most important things you should keep in mind is past data.

To best optimize for your goal and predict future bids, Google’s automated bidding strategies rely on historical data. In fact, for smart offers to be effective, Google recommends having at least 30 conversions (50 for Target ROAS) in the last 30 days or longer.

It’s also worth noting that switching to a new strategy will force the algorithm to re-learn the nuances of your market and collect data from scratch. So make sure you are giving your campaigns enough time.

Remember to pay attention to the Cost Per Click (CPC) of your Google Ads campaigns, too.

Although paid search is fantastic at delivering fast ROI and sustaining your business, if you’re in a very competitive PPC space with a low entry barrier, your CPCs will inevitably get higher.

To protect your business from the ‘race to zero’ in paid search, consider moving up the marketing funnel and investing in supporting the user journey.

What to Do When Your Automated Campaigns Fail

So what happens if your automated campaign doesn’t yield the results you were hoping for?

It might seem logical to blame the machine learning algorithm. But where many smart bidding campaigns actually run into trouble is often the data set itself.

A lot of times the solution is quite simple: further segmentation of the audiences.

If your automation is failing, consider looking at different ways that you can further shrink and segment the audience data that you’re feeding into the algorithm.

As one part of the audience may look completely different than another, running an AB test can prove very useful, as well. AB testing can help you discover key findings, such as whether a variant should be deleted, or whether one of the segments should be further segmented.

Audience Segmentation Is Key

AI-driven automation in digital advertising is becoming increasingly complex and powerful, but also more accessible for all of us to benefit from. Still, many businesses are struggling to attain a satisfactory return on their PPC spend.

One of the best ways to tackle a failing smart bidding campaign is by (re)defining your audience.

Ultimately, further segmenting your audience based on identifying elements can significantly increase the likelihood that searchers will click through and take action.

 

Curious what PPC best practices will look like in the future and which trends you should continue to implement? Head to our blog to discover some of the most important PPC insights to keep top of mind.