The day Google introduces a fully-automated campaign type that requires minimum setup and serves across multiple inventories has finally come! A couple of months ago, the so-called Performance Max Campaign was made available for testing and lately it has fully replaced its older sister – Shopping Ads.
Does this mean Google Ads specialists’ work will start fading away? Well, not necessarily.
A full automation doesn’t mean zero optimization work for the advertisers. Quite the contrary – since any type of automation needs to be able to recognize the context well, it is then our job to supply Google’s algorithms with the right advertising content, clear goals and valuable user data, so that they can operate effectively.
Automation has great capabilities, but it also has some limitations too. It’s essential to build the right framework and constantly test what creates the best context for the algorithms. In other words, automation is a tool that professionals use, it can’t self-operate.
Here is what we have learnt at Monsoon about the Performance Max. Discover the tactics to effectively manage Performance Max across eCommerce and lead generation businesses.
✓ Tip: Google rewards simplicity and always encourages advertisers to simplify the account structure. The idea behind it is simple: aggregating as many conversions per single campaign accelerates the learning phase of the algorithm and allows the tool to achieve the desired goals more effectively. This applies to Performance Max, as well.
✓ Tip: if you run Performance Max for a big commerce store with multiple products available, start with a limited number of the campaigns. When having a sufficient volume of conversions, you can work on a more granular campaign setup. Start evaluating metrics such as ROAS, CPO or CR% per product to decide how to segment your Performance Max campaigns. Make sure you do this, especially if you notice some of your products (eg, with high average basket value) aren’t getting enough impressions. They can fall in terms of priority – the algorithm prefers to boost the visibility of other, top performing products. Simply filter them out of the existing campaign and target them within the campaign with the new setup.
✓ Tip: An exception from the above rule is when, from the start, you want to preserve separate budgets, language or geo-targeting settings for your campaigns. Or if you know that certain product groups differ in terms of ROAS or margin – then you can build fresh new Performance Max campaigns for each segmentation from the start. Under the Listing Groups, choose which products you want to target in your PMax campaigns. If you know that certain groups of products have a different value for your business or different ROAS, create separate campaigns or asset groups for them by selecting and grouping the desired products (you can choose product type or Product ID filters).
This is great for advertisers who want more control!
Conversions / Bidding
✓ Tip: Be sure to track all important conversion actions in the system. When setting up a new Performance Max campaign, we recommend you clean up the setup and remove any irrelevant conversions, so that they aren’t taken into account for bidding. If you happen to run the campaigns with different business objectives, split them for optimal optimisation. You can also use Conversion Value Rules to value conversions differently, depending on specific business insights in real-time at auction time bidding.
✓ Tip: Google advises for performance Max campaigns to have a high enough budget when running on tCPA. When starting the campaign, keep the budget at least 20x the avg CPA. In already running campaigns, budget to be set to at least 10x the avg CPA (last 30 days of the account’s avg CPA).
✓ Tip: If the goal is to acquire new users only, Performance Max gives you an option to focus on this action specifically. When creating the new campaign, you can adjust this setting and the ads will be shown to new customers only, regardless of the selected bidding strategy. Loyalty-oriented goals will be soon added in the system – right now they are available as a beta only.
✓ Tip: Performance Max is complementary to Search campaigns using keyword targeting, that means it should be run alongside your search campaigns. The exact keywords added in your Search campaigns will take precedence over Performance Max, but if you see too much of an overlap, you can always ask your account manager for adding keyword exclusions. That will guarantee that Performance Max ads won’t serve. The negative keywords will block all creatives including advertiser-provided creatives from serving for those user searches.
✓ Tip: The Final URL Expansion feature is turned on by default when you create a new campaign. The setting allows the machine learning to select the best landing pages that will drive the best performance based on customer intent. Here, you can use this option to achieve 2 things: remove non-transactional landing pages or to only send the traffic to a single page, as you know that this is the only path to conversion.
✓ Tip: Create unique asset groups if you want to tailor messages for different audiences and be sure to group them under a common theme. Create unique asset groups for different final URLs if the assets are final URL specific. Make sure that the assets closely match the content of the Landing Page.
✓ Tip: The more assets you provide, the more ad formats the system can create and the more inventory the system can target. Utilise all formats: text, graphic and videos.
✓ Tip: Refresh the creatives frequently! Ad fatigue is real, so be sure to develop a habit of frequently testing new creatives and new messaging to customise the campaign for the relevant audiences. If your campaigns have enough data volume try testing new creatives at least once per week.
✓ Tip: Ad strength can be found during ad creation and also it is ranked under “Performance” columns in Google Ads panel. Make sure you frequently review this section to identify all poorly performing assets. If you see that the performance is “low” consider changing the asset into something else.
Also, take notes of which ads are/aren’t performing well so you can consider those factors when developing new creatives in the future.
✓ Tip: Adding Audience signals is optional, but it provides Performance Max with an important signal for optimization, which will help machine learning to find the right audience for your business. Add higher intent and more specific audiences like custom intent search query audiences to give the algorithm a strong signal about your targeted audience.
✓ Tip: Possibly the most powerful option is using 1st-party data (Customer Match) and remarketing lists. If you want to increase the chances of the algorithm finding the right users, feed the system with the actual data about your customers.
Lastly, remember to let the Performance Max campaign ramp up at least 1-2 weeks and bear in mind the conversion lag time. Take that into account when evaluating the performance of the changes you’ve applied.
Let us know about your best practices regarding Performance Max!