Demand Gen Audience Breakdown

There has been some confusion about Demand Gen and mechanics of how audiences work. I thought it would be helpful to share a resource on how audience signals function and how to understand each of the settings. For examples, I used the new Dragon Age game, Veilguard to represent a fictional luxury resort suitable for families and romantic getaways. All rights belong to EA and Bioware.

First it's important to explain what demand gen campaigns are: these campaigns serve on Youtube, Discover, and Gmail. They are designed to inspire interest/demand, though can be held accountable for sales/leads if you choose the right goals and apply the right creative + audiences. They will blend clicks and interactions so depending on how you have your reporting set up, you may need to change which columns you're using to see their full value. Check out Jyll Saskin Gales 's excellent post for more info there.

These ads will have different algorithm rules based on the goals set (much like Meta), so it's important you ask for what you want.

What is the difference between an Audience Signal and an Audience Segment?

Audience signals are inputs from advertisers that help early stage campaigns know who to target, but ultimately do not lock in any specific targets.

On the other hand, audience segments are hard targets that specifically/exclusively target who you choose.

If an advertiser opts into "optimized targeting" they will turn their "segments" into "signals". This means that specific targets will no longer have exclusivity on the budget.

It's important to note that when you're adding in audiences, you'll be shown estimated reach and reach with Optimized targeting.

What we see on the right is a breakdown of the audiences chosen and which are eligible. Below the 100M to 500M, there's a >2.7B number associated with "Optimized Targeting". If the advertiser would opt into that setting, the segments would change into signals, and the available market would grow exponentially based on the chosen locations, devices, and schedule.

Should You Use Optimized Targeting?

The short answer: it depends.

The long answer is it really comes down to how much exclusion work you're prepared to do and whether you've budgeted for that expansion.

If you're going to turn expansion on, it's likely a bad idea to let your ads run all day everyday. While it's true there can be value at all times of the day, you'll likely want to save budget for higher probability of value times.

For those just starting out, I recommend splitting your ad schedule into six hour segments so you can see when traffic is happening and then make edits from there.

Additionally, you will likely want to limit which devices you include in your campaign settings. This will ensure that you don't end up on placements you didn't build creative for. This step should be taken anyway, but absolutely if you leave audience expansion on.

It's important to note that the above screenshot shows optimized targeting off and impressions are forecasted at 31 million (much lower than the 100M-500M or 2.7B). If you see larger numbers than you expect, your campaign has likely turned Optimized targeting on and you're using signals. If you see the same or lower, you're likely "safe" with the segments you chose.

A final note about optimized targeting precautions: exclusions. Google respects exclusions, however many don't take the time to do them. If you don't want certain people, placement, or concepts getting access to your budget, take the time to exclude them. This is a way to get the benefit of Google's bias free expansion, while still protecting budget and brand sensitivity

General Demand Gen Thoughts & Strategies

When Optmyzr looked at ad creative we found that sentence casing tended to outperform title casing. However the hidden finding in the demand gen portion is that we can hold the campaign type accountable for ROAS. The key is to choose the right goals and targets.

What's interesting is that Google's algorithm doesn't always pick up on how their default audience might be a mismatch for the stated intent. For example, in the above screenshot, I asked for "luxury family all inclusive resorts" and Google wanted to serve me an audience that's 62% "non-parents". This is where more specific settings like detailed demographics can help correct the system.

It's critical to test different creative, and confirm that you're happy with your creative in all the formats it's going to render. Be sure that your messaging aligns with your audiences and that you're mindful of any expansion.

It's important to keep current with the rules of engagement in Google Ads. As of this writing, the above highlights how audiences work for each type of campaign.

Hopefully you found this helpful and if there are other questions you'd like me to do a deep dive on, let me know in the comments!