TPE #46: advanced pMax campaign structure (high impact)
Jun 12, 2023Read time: 5 minutes
The difference between failing and winning Performance Max campaigns is how you set up your campaign structure.
It can truly make or break your performance.
In this issue of The PPC Edge, we share an advanced feed-only pMax campaign structure that works really well for ecom accounts.
It also works for Standard Shopping.
Let’s dig right in!
The problem with automation: Google is the boss
The biggest issue with automated campaigns like pMax, is that Google decides which products get pushed.
News flash: it’s not always the best performers.
Here’s proof from one of our accounts: 66% of the budget was spent on underperforming products:
And only 33% of revenue comes from best performers…
Note: we got this data from the Flowboost Labelizer script. Your products get a performance label based on historical data:
- Index (products that perform on target)
- Over-index (overperforming products)
- Near-index (products that perform near target)
- Under-index (underperforming products)
Check out this guide for a complete breakdown.
Let’s move on!
So… These insights show us Google is not always efficient at pushing our best-performing products.
The solution: Performance-Based Bucketing
Luckily for us, there’s a solution. We call it ‘Performance-Based Bucketing’.
It works particularly well for feed-only pMax (and Standard Shopping) campaigns.
Important note: this is an advanced strategy that needs a lot of conversions (50-100 per month in total).
These are the benefits:
- Increase spend on best products
- Decrease spend on worst products
- More control and better scalability
And it’s 100% data-driven.
Full credit to ProductHero for popularizing this strategy with their Labelizer tool that labels products as ‘Heroes - Sidekicks - Villains - Zombies’.
You can set up Performance-Based Bucketing with a tool like ProductHero Labelizer, or a script like Flowboost Labelizer.
For the sake of simplicity, we’ll only show how the strategy works with ProductHero Labelizer. If you want, you can also use another tool or script.
The main goal: increase spend on your best products, and decrease spend on the worst
With Performance-Based Bucketing, you spend the majority of your budget on your best-performing products, and less on your worst-performers.
Here’s how it works:
Based on historical data, your targets, and CVR/clicks thresholds your products will get a specific performance label:
- Heroes (your best-performing products)
- Sidekicks (products that perform near target)
- Villains (your worst-performing products)
- Zombies (products with low impressions)
ProductHero Labelizer then creates a supplemental feed with performance labels for every product. You need to connect it to your product feed in Google Merchant Center.
Then, for every performance label, you’ll set up a campaign.
After that, your products will automatically rotate between campaigns once they receive a new label based on recent performance.
For example: one of your best products suddenly drops in ROAS and CVR and performs below target. The label will change from ‘Heroes’ to ‘Villains’, which automatically puts the product in your ‘Villains’ campaign.
Let’s briefly look at each performance label and campaign.
1: Heroes — your best-performing products
These are your best-performing products.
- Budget: +/- 50% of daily budget
- Bid strategy: optionally lower tROAS to push harder
The main goal: maximize spend on your best products.
2: Sidekicks — products that perform near target
These are products that show potential, but aren’t there yet (for example: high ROAS but low amount of clicks).
- Budget: +/- 20% of daily budget
- Bid strategy: variable ROAS target
The main goal: dedicate a portion of your daily budget to products with high potential. Once these products get more spend, they will likely shift from Sidekicks to either Heroes or Villains.
Note: if your Heroes and Sidekicks campaigns have low conversions, you can combine them to provide more data for the algorithm.
3: Villains — your worst-performing products
These are your worst-performing products.
- Budget: +/- 10% of daily budget
- Bid strategy: optionally higher tROAS to scale down
The main goal: decrease spend on worst-performing products.
If a product performs very badly over a long period of time, you could remove them from your campaigns altogether. However, we like using Villains campaigns, because some products might perform decently with low spend.
4: Zombies — products with low impressions
These are products with low impressions. Google often prioritizes products with historical data, resulting in low impressions on a big percentage of your products.
- Budget: +/- 20% of daily budget
- Bid strategy: variable ROAS target
The main goal: increase visibility on products with low impressions to find new winners.
Easy technical implementation with ProductHero Labelizer
If you want to test Performance-Based Bucketing in your accounts, we suggest using ProductHero Labelizer.
It’s one of those tools that instantly pays for itself (and it’s not even expensive — around €99 per month).
How you set it up depends on your account data and goals.
Luckily for us, ProductHero has an easy-to-use setup wizard that guides you through the process and helps you determine your thresholds.
TLDR: Take back control with this advanced Performance Max campaign structure
Performance-Based Bucketing is an advanced strategy to take back control and scale your best-performing products.
ProductHero Labelizer is the best, easiest-to-use tool to help you set it up fast.
And that’s all for today.
Happy testing and see you next week!
Cheers,
Bob & Miles
Disclaimer: this newsletter is NOT sponsored by ProductHero but we did include affiliate links. We only recommend tools we actually use ourselves. ProductHero is a fantastic company that has a big, positive impact on the PPC world and we’re 100% sure you will improve your Shopping/pMax performance with their Labelizer.
