The two most overused words in the world of digital marketing and the world of PPC in 2022 have to be:
Performance Max.
The whole PPC community is buzzing about it, and it's pretty understandable. This campaign type will hugely affect many businesses. So we, PPC specialists, have to take it seriously.
This 2 part guide will showcase a few best practices, so you can get your Performance Max campaigns up to the max. This first part will cover the best practices for eCommerce, and the second part will focus on Lead Generation.
Let's dive right into it!
Table of contents
What are Performance Max Campaigns?
Performance Max is Google's massive step towards automated PPC advertising. We have seen every year that Google is shifting more and more towards automation, with DSA campaigns (dynamic search ads), smart-bidding, RSA's (responsive search ads), or Smart Shopping.
But, Performance Max is a very different animal.
This campaign type takes automation to a different hemisphere, and the PPC world had one-or-two things to say about it.
We can dwell on it, but this is here to stay, so we better start figuring it out.
How does Performance Max works?
The campaign type can serve the ads on every Google network. The advertiser will give the input for the campaign: photos, videos, ad copies, and, if available, a product feed.
After setting up the base of the campaign, we can choose from 2 bid strategies: Maximize Conversion Value or Maximize Conversion. We can ultimately set a target ROAS or a target CPA (cost per acquisition).
Based on the input, Google will ensure that our ads are served to the right customer and reach our desired KPIs.
If you are a business owner, you might think at this point:
"Wow, nice, I can do this myself. I don't even need to hire a PPC specialist."
Here comes the catch! (There is always one. Sorry! :) )
Performance Max campaigns give a false sense of being too easy. However, when we dive deeper, we realize there are many possibilities in this campaign type.
Let's see how an eCommerce store can fully utilize the Performance Max campaigns.
1. Finding the correct structure
Everything starts with a solid foundation. The same applies to the Performance Max campaigns. With a good structure, we are already halfway to the winning formula.
However, this is easier said than done.
The world of PPC is slightly divided on this topic, and many suggestions and solutions exist. I don't want to judge any of them. I will only share the structures I tried myself and utilize them successfully.
Therefore, I consider three more significant structure types:
1.1 One campaign, multiple asset groups.
This structure is appropriate for smaller shops/businesses or advertisers who want to test the waters before diving deeper.
The idea is to create one Performance Max campaign and add numerous asset groups. Then, we will use the asset groups to split the feed based on product group, brand, or other unique elements that are relevant to you.
It is a straightforward layout and somewhat easy to operate. If you have a low conversion volume, I suggest taking this approach.
I'm not particularly eager to complicate things just for the sake of it. Most of the time, less is more.
Focus points:
When you utilize this approach, divide your feed logically (based on product group, brand, Etc.). Add unique assets to each group to be more relevant to the shoppers.
Keep one generic asset group where you have all your products and more generic assets. It comes in handy if Google can't connect the search term with one of the more specific assets.
For example, You sell clothes and split the campaign based on product type (tops, t-shirts, Etc.). What if someone uses the search query "beautiful clothes"? In this case, Google can nicely use the generic assets.
Reminder: I don't recommend using the "One campaign - one asset group" structure. Start with this structure to learn your feed or give time to gather assets. As soon as you have enough data/assets move away from this structure.
1.2 The BCG matrix structure
The BCG matrix feed structure is an excellent step toward the margin buckets and gives incredible feedback about product performance.
I can highly recommend the Producthero Labelizer tool. It makes managing this approach easier and gives further valuable insights into your product feed.
Let's break down the diagram using the names given by Producthero.
Heroes - A small percentage of the products that generate the most volume and revenue at a great ROAS.
Sidekicks - They generate a small amount of revenue at a great ROAS, but they don't have much volume
Villains - The bad boys generate much traffic and little revenue at a low ROAS.
Zombies - The sleepers in your product feed. They get little to no impressions. Time to wake them up!
I recommend this method for eComm shops with more conversion volume. You will use a higher number of campaigns and don't want to limit the campaigns with low conversion data.
The below-showcased layout it's a great starting point:
Fallbacks: The newly added products to the feed don't have one of the four labels. We use the 5th campaign to activate the new products and give them a label. In this campaign, one asset group is enough!
I like this format because it allows you to push your best products while giving a chance to other products in the feed to become a "Hero" product. In addition, it adds a healthy rotation to your product feed.
1.3 The margin buckets
I would consider this approach if you are a more advanced retailer and want to take your Performance Max game to the ... Max (hehe).
I won't go into how to find your breakeven ROAS or your margins. I think you already got this. Nonetheless, if you need some help, there are numerous excellent resources out there that can help you with this.
As you can see, we will split the product feed into three campaigns: High, Mid, and Low Margin. With this layout, you will focus your attention on where it matters.
Inside the campaigns, try to keep it simple. For example, I would split the asset groups based on product type or brand.
Then repeat the same thing in the Mid and Low Margin campaign.
Note: I don't think it's necessary to split the asset groups too much. At this point, the algorithm should do that for you. However, in some use cases where you have a lot of unique personas, you will need to do that.
You can use many other structures, and if you Google a little, you can find a few of them. However, I can highly recommend the retail structure developed by Mike Ryan.
Furthermore, the guys at AdChieve also have a nice piece of content about the almighty Performance Max
After we found the proper structure for the Performance Max campaigns, there are a few other things that you have to keep in mind.
2. Feeding the algorithm with data
In the age of automation, when the AI will do everything for us, bidding, listing the products, and choosing the audience, you might ask how we can distance ourselves from the competition. How can we become unique?
The answer is to feed the algorithm with data that Google doesn't have. Use your first-party audience lists, use high-quality, unique assets and use solutions to fill the data gaps (exp: consent mode)
2.1 Assets
The easiest way to stand out from the crowd is to provide high-quality assets to the Performance Max campaign. I know it's easier said than done, but this can make a huge difference.
Note: Create your videos and add them to the asset groups. If you don't add a video, Google will create one for you from the provided assets. I recommend avoiding this since those videos are awful.
2.2 Audience signals
These are part of every Asset group inside the Performance Max campaign. The audience signals are no more than signals, as the name suggests.
It's a starting point for Google to get information about your audience and find people similar to them.
It's an excellent way to feed the algorithm with correct data.
I recommend not going too broad here. Instead, stick to your first-party audience lists or maybe a few custom segments. The more exact and specific you are, the better and more relevant data you will provide to Google.
2.3 Fill the data gap
Google provides excellent options to fill the data holes in a privacy-safe way, and I think we should use them.
I have to say that consent mode has its flaws and bugs, so use that with caution and after a few tests, but the Enhanced Conversion works nicely.
If you don't know how to implement the Enhanced Conversions, I created a step-by-step guide, so check it out.
Start small and improve.
No single method is better than the rest, so start with a simple structure and improve from there. See what works for you and when you find the right direction go from there and improve.
Keep in mind :
Don't overcomplicate things - Start with one campaign and go from there.
Feed Google with data - Use first-party data and give information to Google that they don't know (stock level, margins, Etc.)
Use unique assets - Use relevant and high-quality photos and videos.
Conclusion
As in every marketing field, there is no correct answer to which structure is best.
It depends.
It's essential to test and improve our approach constantly. I don't believe in best practices since the best practice is what works for you.
It's an entirely new campaign type, and there is a lot of noise about it in the PPC world. Try it for yourself and decide.
If you liked the article, please consider sharing it with others if you have any questions feel free to contact me.