7. Performance Max Campaigns
How to Use AI in Performance Max for More Profit
Robots will be able to do everything better than us. This is the scariest problem to me. AI will be the best or worst thing ever for humanity.
Elon Musk at SXSW 2018
AI generated the cover image for this chapter.
The Google Ads ecosystem has channels of Search, Display, Shopping, Video, Gmail, Discovery, and Local, each with many features. The depth and width of the ad platform makes it time-consuming to manage, even for experienced users.
Google created Performance Max so advertisers save time managing campaigns while using artificial intelligence to improve results. Even so, the campaign type is not for everyone.
I’ve discovered best practices that work for our clients to help you get the most out of Performance Max. The chapter gives you guiding principles and frameworks to make the campaign driven by artificial intelligence profitable for you.
What is Performance Max?
Performance Max harnesses the power of its advertising ecosystem and enables advertisers to serve their ads across YouTube, Gmail, Discover, Search, Shopping, and Display networks from a single campaign.
With Performance Max, you are one-click away from combining your ad copy, uploading product images, adding links from your YouTube channel and connecting your Google Merchant Center. However, you need to understand how to use it to explore its maximum benefits, as it will leak money if you’re unfamiliar with its functionality.
The four key components of a Performance Max campaign include:
- An objective—The objective, or goal, should be based on the outcome you want for the business. If you’re advertising a Shopify store, you’ll select the “Sales” objective most of the time.
- Assets—The creative materials Google needs to create the various ads across its network that includes images and logos, videos, and ad copy.
- Listing groups (ecommerce)—The connection to your product feed through Google Merchant Center, which can be subdivided for segmentation.
- Audience signals—Data suggestions that tell Google the traits of your ideal potential customer. Accurate audience signals speed up machine learning.
Why Use Performance Max Campaigns?
There are three main reasons you would choose Performance Max campaigns over building out individual search, shopping, display, and video campaigns:
1. Less Complicated Campaign Structure
One of the key benefits of Performance Max is that you have access to everything Google Ads offers from a single campaign. You won’t need to spend hours creating all the variations of ad copy and media assets and targeting the platforms you hope your customers are on. Performance Max is like a mixing bowl where you add all the ingredients.
While the setup process for Performance Max differs from other Google Ads campaigns, it offers access to the same advertising features. It’s also worth noting that Performance Max eliminates time-consuming keyword research.
2. Utilise the Power of Machine Learning
Just like machine learning, Performance Max helps businesses achieve quicker results using artificial intelligence. It can continuously learn and adapt to the evolving online landscape, ensuring your campaign performs at its best.
3. Increased Visibility
Businesses aiming to maximise their visibility across Google Ads advertising platforms need to know how to create campaign types and where to find their audience. Performance Max campaigns accomplish that by mixing and matching all the combinations of assets and signals you’ve provided then showing ads on the right network at the right time.
Are There Limitations to Performance Max Campaigns?
Performance Max campaigns are hampered by a few limitations you can get past once you’ve mastered how to use this AI-based campaign type.
1. Keyword and Asset Control
New campaigns have advantages and disadvantages, but it all depends on the user. For Google Ads managers who love targeting a specific audience or having control over their search terms, Performance Max may not be the best option for you…yet! Although there is a workaround, you may find the inability to add keywords uninviting.
The ability to sculpt your keywords and their variables can contribute a huge benefit to an optimal campaign. However, there is a workaround for this limitation, that I’ll discuss later in this chapter.
Another limitation is adding videos to the campaigns, as using your linked YouTube channel is preferred. You are provided with the option of selecting a video that you think would be a great fit. However, if you don’t have a linked YouTube channel, Google will combine your supplied images and headlines to create a video. While this may sound advantageous, the quality of the video compiled by Google is often awful.
2. Doesn’t Play Well with Others
Although Performance Max can optimise across all Google advertising channels, it will cannibalise other campaigns, especially branded search and standard shopping campaigns. Hence, advertisers will lose sales in well-structured Shopping and Search campaigns when running them side-by-side with Performance Max.
3. Lack Of Data Visibility
The less complicated campaign structure comes with the major downside that it is difficult to optimise because most data is hidden to the advertiser. You get poor visibility into the performance of channels, keywords, audiences, and products.
An example of limited data visibility is being able to view only the performing keywords in groups rather than being supplied with metrics for each keyword. Google might release more visibility overtime, but until then, you’ve got limited data to work with.
How to Stop Performance Max From Destroying Results
The Right Structure
Before you take on a Performance Max campaign, you need an optimal structure for your goals.
The best default is to begin with one campaign. There are maximal placements the algorithm needs to feed ad spend into to discover what works. Businesses with a single range of products of similar margins can begin with one catch-all Performance Max campaign that serves all products to a handful of audience signals.
Stores with a large range of products that vary largely in margins, seasons, conversion lag, or performance, a single campaign will not suffice. Consider the following factors that can warrant separate campaigns while deciding on your structure:
- High or low margins
- Price point
- Best sellers
- Zero impressions/clicks/conversions
- Product launches
There is no optimal number of campaigns when it comes to building an account structure for your range of products. Keep it simple to begin with. When starting out, you don’t want to be tinkering with your campaign due to how Performance Max campaigns function.
How to Edit Campaigns to Make the Algorithm Work For You
A Performance Max campaign is best optimised differently to a Search or Shopping campaign. You must resist the urge to make changes in the first few weeks while the campaign learns how to find your ideal customer. You also have to account for conversion lag, which is the time it takes for your customer to convert after seeing an impression of your ad.
From the diagram below, avoid editing the campaign for at least two to three weeks and also allow for conversion lag to ensure it can work with the data coming in. Editing asset groups, even adjusting your budgets, will “reset” the learning.
You can begin to review the campaign around the fourth week. Discount the results in the first week, then look to make updates or changes by the fifth and sixth weeks if needed.
I’ll discuss helpful tips on how to review the campaign and reports in a later module of this chapter. As mentioned earlier, these new campaigns are a little wanting to share the data, but I’ll explain a few ways to dig into what we need.
How to Build a Profitable Performance Max Campaign
Now we’re getting into the serious part of all these—building out a Performance Max campaign. I will teach the points that matter to create a successful campaign.
When creating a new campaign, Google encourages you to select a campaign objective. Although the Performance Max campaign option is available for some objectives, get started by selecting “Creating a campaign without a goal’s guidance”. This doesn’t affect performance but gives you the freedom to access all settings for the best setup.
An easy one, as this is the reason you’re here. Select Performance Max:
This step is key to the success and reason you are building this campaign. Whether you want to sell your products, highlight a service or generate leads for your services, you must choose the conversion goal action you want to optimise for. For ecommerce, select your conversion that tracks purchases.
Select your feed from Google Merchant Center in order to run shopping ads within the campaign.
Give your campaign a name. I like the following naming convention for consistency as it tells me core information at a glance:
Website Name – Campaign Type (e.g. Shopping) – Keywords of Insight (e.g. Brands, Bestsellers, 0-Click Products)
The budget and bidding options define how much you want to spend to achieve your goals for this campaign. Whilst the budget is straightforward, bidding focuses on two options that each contain the ability to set a target.
1. Maximise conversions
Maximise conversions. You want the campaign to find as many conversions defined at step 3 as possible without limitations.
Setting a target for conversions is also known as “target Cost Per Action” (tCPA). The bid tells Google how much you are willing to spend to get a conversion. It is often used for products with the same value or a specific service.
2. Maximise conversion value
The bid strategy’s goal is to provide you with the most conversion value from your budget. The bid strategy is excellent for stores with products that vary in value.
Setting a target for conversion value is commonly known as your “Return On Ad Spend” (ROAS). Your ROAS goal tells Google that you want to receive a specific amount in return for every dollar you put in, entered as a percentage. For example, if you need to make $4 for every $1 spent, your target ROAS is 400%.
Google does not take into account the profitability of each of your products when selecting any bid strategy, so you need further calculations then segmentation to maximise profit.
When setting up a new Performance Max campaign, start with the “Maximise Conversions” bidding strategy because it provides minimal restrictions to learn from collected data. Once you get some CPA or ROAS stability, you can consider adjusting your bid strategy to target a specific cost-per-action (CPA) or switch to the “Maximise Conversion Value” strategy and target a return on ad spend (ROAS).
Performance Max campaigns by default bid equally for new and existing customers. The “Customer acquisition” setting allows you to optimise the campaign to bid higher for new customers or exclusively optimise the campaign to bid for new customers only. Your marketing goals determine the best setting for your business:
- Bid higher for new customers than existing customers—Google adds additional conversion value when a new customer makes their first purchase. It works twofold by allowing you to maximise revenue across all customers and drive new customers to your business within the same campaign.
- Only bid for new customers—This limits ads to customers that Google sees are new to your business based on past purchases and the existing customer list you uploaded through Google Ads. Available audiences can include those who have visited the website but did not buy. Expect this campaign option to perform worse in reported revenue within the ads platform, but it may align better with profit goals.
There are several ways to identify new versus existing customers in Google Ads to help a new customer acquisition goal. The more ways you use, the more accurate Google’s bidding algorithm will be. You can help both settings by distinguishing new versus returning customers through three methods:
- Customer match list—If your business has built a list of customers and transactions over time, you can upload your audience list to help Google determine your new customers.
- Google detection—This is the audience that will automatically be used when selecting this goal if your Google Ads conversion tracking is set to track “purchases”. Google can create an audience list, up to the last 540 days of tracked purchases, and is the easiest way to distinguish new customers. However, it does not consider users who have removed the ability to be tracked through cookies and opted out.
- New customer acquisition reporting—Enter additional code in your conversion tracking snippet to tell Google whether the purchase was from a new or returning customer. If the person has ordered before, set the
new_customerparameter to true. Our Google Ads tracking code for Shopify includes new customer acquisition reporting.
When optimising for new customers, you’ll need to enter the additional value a new customer is worth to your business by assigning a value that considers your customers’ lifetime value (LTV).
In the following screenshot, the conversion value from an existing customer has a value of $16.65 and based on repeat purchases, the average value from a new customer would be $63.87. It means the added value from a new customer after their initial purchase is an average of $47.22, which will be entered. Google Ads automatically provides these numbers using the value from your customer list (segment) on an average order value and a typical purchase frequency.
There are two caveats to be mindful of when using this bid discrimination feature:
- The “new customer acquisition” option targets those that have not purchased. Do not confuse this with new website visitors.
- Google uses its data to define who it sees as a new customer. Provide your audience list and modify the conversion tracking code to push your data about customers to Google for the best accuracy. A customer has to be detected in only one of the methods to be recognized as a returning customer.
Click on “Next” to get to the campaign settings:
You have multiple options to select which locations you want to show your ads on the Google advertising network. My recommendation is to be as granular as possible because it allows for better optimisations based on locations once data pours in. Start with “Enter another location” and select “Advanced Search”. You will then be able to add locations by country, city, region, etc. or even select a radius, say around your bricks and mortar store. For location breakdowns, here’s the Digital Darts Google sheet to save you hours.
For languages, this allows you to restrict where your ads appear based on the user’s language settings and the site’s language. Choose the most appropriate to your target audience.
Final URL Expansion
There is another sub menu called “Final URL expansion” that you need to edit to control where you send people who click on your ad. The key component is tucked away, and you’ll need to expand the “More settings” menu.
Google defines the final URL expansion setting to let you change the landing page based on the user’s search query and intent.
If enabled, the option generates a dynamic ad headline, description, and additional assets to match your landing page content. You do not choose where your ads show on the Google network or its destination from the assets you supply.
There are a few options to help you get the most out of this setting:
- You can let Google do its “magic” and allow it to dictate the bulk of the work by directing your website visitors to specific pages that it thinks will be best matched, based on the information it has on a prospective searcher and where they are in the journey.
- You enable the option but exclude certain pages you do not want Google to use as landing pages. The best option I recommend for most businesses. Advertisers of Shopify stores can exclude
/blogs/. This will stop Google from sending paid traffic to pages like your shipping, terms of service, or articles.
- Turn the option off and send traffic to URLs you provide. Unless you have a specific reason to turn this off, leaving this option on should provide the results you want since Google utilises the data it has on users and combines that with machine learning to show your ad to your ideal customer.
Asset groups are the central location for all assets Google uses to show ads. You provide headlines, images, videos, and products, then Google creates combinations that resonate with your customers to ultimately achieve your goals.
When building out asset groups in a new campaign, treat these as themes that determine structure. A good place to start is modelling the website’s navigation since it segregates product types. A woman’s fashion store could have asset groups of dresses, shoes, shirts, belts, and jumpers.
Similar to Shopping campaigns, listing groups determine the products to include in the asset group. You’re able to segment listing groups in the same way as Shopping campaigns.
Select products that match the asset group’s theme for the best performance from continuity. If you select shoes in the listing group of an asset group that contains audience signals and copy about dresses, you’ll hurt sales. A well-structured listing group provides better insights than throwing all the products into one, especially if you are segmenting by profit margins or high-ticket products.
The next step is to include your final URL. This option was presented to you earlier in setup, where you could turn it on or off.
If left on, your final URL may be replaced by Google with a more relevant landing page, headline, description, and other assets based on customer intent.
If turned off, you need to provide final URLs in your asset groups to a page, product, category, or collection. Help campaign reach by having asset groups at least for the homepage and major collections.
To help you build out assets, here is the breakdown of what can be used for each field, image and video dimensions, and the number of options that can be included:
The image below shows the creative flow and where to enter each asset. As you update assets, Google provides a preview option to see how it looks across the various networks:
Audience signals in Performance Max enables Google to optimise the performance of ads by targeting relevant audiences. Unlike search campaigns where each ad group is assigned keywords and match type combinations, these asset groups use audience signals to determine where ads display across placements.
The signals are not used for targeting like other campaign types—they guide Google to match data points to find the right customer. Audience signals are designed to increase the relevance of ads to the target audience by identifying users with the highest likelihood of responding to your ads.
Start with two or three signals based around the theme of asset groups where your media and copy align. Since Google treats these as signals rather than audiences, you don’t have to be perfect.
A customer match list is one of the most important segments you can add to a Performance Max campaign since it is accurate data. The list provides Google with a mountain of information on your ideal shopper. I recommend combining this list with a custom segment of your top 20-30 converting search terms into the mix.
New accounts can start with a custom segment of URLs of websites that offer similar products. Search Google to see which competitor websites offer similar products or services that rank high.
Another strong signal is from search terms around your product or service. Expand keywords to other interests of your audience. For instance, if you sell wedding rings, consider terms such as “wedding planning” or “wedding events”.
|Customer Match List||If eligible, existing lists are added to your Audience Manager. The information from this segment is of the highest value for Performance Max.|
|Search Terms||Review both your Google Ads and Search Console queries, selecting the top 20-30 terms that have converted or have a CTR.|
|Website URL||Collect the URLs of competitors who advertise on Google, and those listed on the first page of organic results for your most important search queries.|
|Affinity Audience||Select the top Affinity audiences from your Google Analytics accounts. This can be found under Audiences > Interests.|
Insights and Optimisations
Although there are limitations to the data available from Performance Max campaigns, I have several ways to make better decisions. Some reports are directly within Google while others are done through third-party scripts.
Search Terms Insights
While not completely flying blind, these insights allow you to understand the categories and terms customers use to find your website and, in many cases, make a purchase. Visit this page each month then either send the keywords you want excluded to Google support, or add them to your negative keyword list.
Search term insights can also help you discover new keywords for your existing Search campaigns as you add exact-match keywords, which offers additional metrics search advertisers are accustomed to.
Audience insights breaks down audiences by type, the share of conversions, and the index value. Google uses this audience information to find potential customers likely to meet your campaign’s objective.
When interpreting the report, know that some people can be in multiple audiences. Furthermore, the index value is the share of conversions, or clicks depending on the report, coming from an audience segment. A higher index means a given audience segment is more unique for you.
The asset insights section is not always visible in accounts. The report shows you how text, images, and video assets perform for particular customer segments.
Asset insights can shed light on how your hard work resonates with your target market, and provide some vision into creating further combinations based on performance.
Campaign Level Segmentation
This is often overlooked as many advertisers dive straight into the Performance Max campaign—overlooking these opportunities. From the top campaign level, select “Segment” to view segments such as:
- Conversion options
- Ad position
You’ll rarely have to make decisions off this data that affect a Performance Max campaign since Google uses the data itself in ad delivery. On a monthly basis, you can use campaign segmentation to monitor the algorithm. Consistently bad segments may lead you to move away from Performance Max so you can better control what’s subpar.
How Well Do Landing Pages Convert?
A custom landing page report can be created to see how final URLs perform in Performance Max. The report not only tells you about the landing page, but if you have a unique landing page per asset group, you reveal how asset groups perform.
Go to “Reports” in the top menu then “Reports”. Choose the “Custom” option and “Table”. With a blank screen, add “Campaign” and “Landing page” to the rows. Add a filter to include only Performance Max campaigns. Select the eligible metrics as columns to suit your needs. Save the report so you don’t have to rebuild it.
Where Are Ads Shown?
It’s insightful to know how individual sites or apps perform for your campaigns. This is one of the rare reports for Performance Max that Google has made available.
You can access the Performance Max campaigns placement report by clicking on “Reports” > “Predefined reports (Dimensions)” > “Other” > “Performance Max campaigns placement”. Click the eligible metrics you want to view then add them to the columns list.
When a website or app performs great, consider contacting them for some partnership or promotion outside of Google Ads since their audience converts well for your store.
Where Is My Money Going?
Google tells you little about where your ad spend goes in a Performance Max campaign. The Google Ads community is blessed in having minds seek the goal of transparency.
One unique option is a script from Mike Rhodes that can create an overview through graphs and tables to visualise your spend across shopping, video, and other campaigns.
Shopping data is sourced from listing groups. “Other” can include networks such as Search, Display, Gmail, Maps, and Discovery. There are many tabs in the sheet that will help you make optimisation decisions.
When Do I Make Changes?
A Performance Max can take up to 6 weeks to get reliable data for optimisations unlike other campaign types. The first week is a write off due to poor performance or extreme unpredictability. Then there is a two-to-four week “learning period” for most campaigns. Lastly, there’s the conversion lag period that averages one-to-two weeks to account for purchases during the learning phase. Only after these phases should you optimise a Performance Max campaign by looking at the learning period.
If you notice your budget is not being spent, step up your bids. Increase the target CPA or decrease the ROAS.
On the other hand, if the campaign spends its budget each day, increase the budget in increments then monitor performance. If the budget cannot be altered, increase the target ROAS or lower the target CPA for better returns.
When a campaign isn’t meeting your goal, it’s tempting to step in and do something. I get it. That is what we humans do. We see a problem then step-in to fix it. However, this micro-management is counterproductive to working with machine learning in Performance Max. Don’t assume you have to do something for the sake of it. Performance Max campaigns—especially in their first few weeks—go through peaks and troughs. Look for consistency over a two-week period and at least 30 or more conversions before making changes.
Unless you have thousands in ad spend with zero conversions or hundreds of conversions, make less frequent changes. The campaign is going great if it hits your target goals and spends most of the allocated budget.
Product Feed Optimizations
Optimise the shopping feed to improve listing group performance. Highlights from the earlier chapter to optimise feed data are:
- When you submit products to Google Merchant Center, provide as much information as possible, especially the product identifiers of brand, Manufacturer Part Number (MPN/SKU) and GTIN.
- Front-load the title with relevant keywords while making it attractive to human eyes. There’s a big difference in how Google recognises and shows a product with the headline “Small Sweater” compared to “100% Wool Womens Red Sweater”.
- Include your best image first along then additional images in the
- Get tax and shipping information accurate.
- Include powerful attributes like
- Create custom labels for price ranges, profit margins, seasons, promotions, or sell-rates.
- Use optional attributes like
Are You Destroying Your Brand on YouTube?
Even if you don’t have a YouTube channel to provide assets, Google automatically creates a video for you in each asset group. Google combines your provided image assets, headlines, and descriptions to create videos that most brand owners would be terrified to see.
One simple alternative is to make videos yourself using Google’s free Video Builder tool. The better alternative is to plunge into YouTube for your business. Learn how to create high-converting videos for your Shopify store. Once you produce your first video that converts a cold audience, your Performance Max campaigns can scale into four-figures per day as video placements—being demand generation—can accrue a lot of ad spend.
When to Split or Merge Asset Groups
Performance Max spends across seven networks, with very little control over where you can spend your budget. With these new campaign types, Google has the freedom to utilise its full range of advertising networks where it was previously controlled by your campaign type.
You can better see the right number of asset groups for you by understanding ad delivery in other campaigns. Let’s say you have a single Google Shopping campaign segmented into ten ad groups. The campaign spends $100/day. Now add a remarketing campaign, your branded search, generic search, video ads, cold display, and Gmail ads. Once you add six other networks into the mix, you can see that you need a bigger boat of money for Performance Max to deliver, the more asset groups you create.
Assets groups let you set the creatives, ad copy, and landing pages that align best for the intended audience. Segmentation of asset groups works best when you have turned off the expanded URL option in the campaign settings.
Your first goal is to bundle similar products into a single asset group. Keep the ad copy general so it appeals to a wide-audience. For example, an asset group about tungsten rings should contain imagery and ad copy that takes people to the tungsten collection on their Shopify store. Wooden ring variations in the same asset group could send mixed signals for a shopper who sees a tungsten ring and headlines while looking for a wooden ring. Later on you can test collection-specific headlines, descriptions, images, and videos in the one asset group so Google can try to deliver the creative that it thinks will appeal to certain audiences.
This leads to the question of how many asset groups should you create? When is segmentation too much?
You should not create 20 asset groups just to match 20 collections. Dozens of asset groups only make sense if you can spend hundreds per day on the campaign. Google’s machine learning cannot learn well in a $50/day campaign that has 10 asset groups.
Too much segmentation happens when the budget cannot keep up to deliver conversions per asset group. You can determine whether the budget is pacing with the number of asset groups by using the Performance Max script shared earlier to see the impressions and conversions per asset group.
If you’ve got an unlimited budget and a diverse enough product range, you can max out the 100 asset groups per Performance Max campaign since each can collect the data required for the learning phase and further optimisations.
How to Make Performance Max Play Nicely with Other Campaigns
One of the most confusing parts about Performance Max is how it interacts with other campaigns. If a Search campaign targets the same search query, should Google deliver the Performance Max or Search campaign?
The general rule is: the campaign with the highest ad rank gets the impression.
The table below shows when Performance Max overrides your existing campaigns. Search campaigns continue to work well if they are filled with exact-match keywords.
|Campaign 1||Campaign 2||The campaign entered into auction|
|Search campaign with a keyword that exactly matches the search query||Performance Max||
|Search campaign with keyword that does not exactly match the search query||Performance Max||
|Standard Shopping Campaigns||Performance Max||
|Display campaigns (without feed)||Performance Max||
|Display campaigns (with feed)||Performance Max||
|Video Campaigns||Performance Max||
|Discovery Campaigns||Performance Max||
|Local Campaigns||Performance Max||
A good diverse campaign structure for new accounts is a branded search campaign, dynamic search ads campaign, remarketing campaign, and a Performance Max campaign focused on new customers with branded keywords excluded.
It feels a matter of time before Performance Max becomes the “Skynet” of Google Ads by taking over all campaigns.
Two Ways to Save Wasted Spend with Negative Keywords
The perfect campaign has an assortment of low-value, non-converting search queries as negative keywords. You cannot add negative keywords yourself yet, but there are two options.
- Ask Google support to add negative keywords to your campaigns.
- Create a new negative keyword list. Email Google support to add this negative keyword list to the campaigns you nominate. Once added, you can manually add the keywords of choice to your list like you do for other campaigns.
Performance Max Conclusion
A Performance Max campaign is easy to start. You can reach potential customers who may not have seen your ads before, since the campaign lets you reach a wide audience in the whole Google network.
The campaign is heavily reliant on machine learning that creates quirks we’ve never seen before in Google Ads. The hard parts are building out a good structure, knowing how to get data, and making optimisation decisions. Once you learn these from the chapter, you can save time, optimise ad spend, and achieve your business goals.
- 1. The Power of Google Shopping
- 2. How Does Google Shopping Work?
- 3. How to Profit From Google Shopping
- 4. How to Set Up Key Google Products for Ad Growth
- 5. Your Google Shopping Feed in Shopify
- 6. Google Shopping Campaign Structure
- 7. Performance Max Campaigns
- 8. Optimise Your Google Shopping Campaigns
- 9. Google Merchant Center Programs
- 10. Expand Beyond Google Shopping