July 22, 2022
Andrew Lolk, Founder & CEO at SavvyRevenue, shares how you get to achieve better performance for Google Shopping whether you run Performance Max (Smart Shopping) or Standard Shopping campaigns. Read everything about it in the article.
Most of the people I speak to look at their Performance Max or Standard Shopping setup and wonder what they can do to improve it.
You can’t control what keywords you show for. Audiences are limited. Negative keywords are misleading. Manual bidding is not fair. So what can you actually do to improve performance?
Improving Google Shopping, and much of Google Ads in today’s world is less about direct tactics and more about nudging the machine in the right direction.
There are three components to Google Shopping:
Today, I will show you how to make changes to each component and get better performance for Google Shopping — regardless of whether you run Performance Max (Smart Shopping) or Standard Shopping campaigns.
It all starts with your feed, and it’s not as confusing as you might think.
Over the years, I have learned that feed optimization is an ambiguous word. It means a lot of different things, so I have distilled it down to two specific optimization tactics:
An attribute is often also called a field — it’s the titles, descriptions, product types, etc in your Google Shopping Feed.
Before you start optimizing certain parts of your feed, you need to add missing attributes. This is a step that most PPC managers miss in their work with feeds. They jump straight to optimizing titles (which is also important), but first, you need to get all the right attributes in the feed.
Here is a list of required and recommended attributes to add:
Once you have added the relevant attributes, you should start optimizing the individual attributes — the first being your title.
Once you have added all relevant attributes, it’s time to optimize the most important ones in the following order:
We will review custom labels in the next section, but let’s first take a look at each attribute here.
Here’s a quick overview of how you can optimize titles:
Once you have finished titles and product types, you can proceed to apply custom labels.
Optimizing your entire feed is a topic in itself, and I recommend our article on Google Shopping feed management
A Custom Label is a way to group products together across brands and product types without adhering to any set list of values. An example is aggregating conversion rate data for all products on sale vs not on sale.
But you can do anything with custom labels. What we use custom labels for the most is to find performance outliers. Some of the custom labels we apply are:
Here’s a further explanation of each custom label:
Discount or not
New product vs existing
In-stock status (across sizes, colors, etc)
Private label products
Custom Labels can be applied in one of two places:
Some of the values can be built using your feed optimization tool even if the data isn’t in there directly. Here’s a simple example of creating a custom label for whether a product is on sale or not:
Once you have applied your first custom labels, you need to analyze the data and find product segments that are worth separating from your existing Google Shopping setup.
I will start off by saying that it’s not always the case that there are performance differences across any of the product segments listed above.
But there are times when you should use them to:
A typical use-case is bestsellers, but this has been talked about for years, so let’s take another example: New products vs existing products.
Let’s say you create this custom label, and find that you spend 95% of your budget on products that have been active on your eCommerce store for more than a year.
But your last year’s strategy has been to expand your product portfolio a lot, so you have actually added 25% more products to your eCommerce store. Spending 5% of your budget on your new products can indicate that:
As PPC managers, we can’t do anything about A or B. What we can do is figure out if C) is a limiting factor. To do that, we need to segment the campaigns to change the budget and ROAS targets.
Segmenting campaigns is a cornerstone tactic with Performance Max (and standard Google Shopping). With Performance Max, you don’t have many options to improve the campaign performance, but you should use the options that are available to you.
Two of the options to improve Performance Max are:
You can use this for a couple of different scenarios:
By combining campaign segmentation, you have options to create different campaign structures that you can use to nudge Google Shopping:
Create a campaign each for:
This will nudge Google Shopping to spend more money on private label products yet still maximize exposure for the external products you carry on your site.
Create a campaign each for:
This enables you to set higher budgets, lower ROAS targets, and overall tell Google it’s time to push the spring/summer products when the season starts — and lower the effort on fall/winter products.
The sky's the limit and the best way to segment your campaigns depends on your business.
Performance Max (Shopping only) is a great way to get started with a baseline Shopping campaign, but sometimes you need to use the standard Shopping campaigns to get the best results.
With Standard Shopping you get more insights and an ability to influence the performance:
This only applies if you actually have something that differentiates your products — and it can’t be directly correlated with the performance (revenue, ROAS, etc.) that you're experiencing right now. Because then Smart Bidding will take care of it for you.
Too many Google Shopping setups are created and left alone. Often advertisers, or agencies, will optimize a bit of the feed, change the ROAS target now and again, but mainly leave the campaigns alone.
That’s a mistake.
As outlined above, there are several ways to improve the performance of your Google Shopping campaigns, and if you feel overwhelmed, they can be narrowed down to a few key aspects:
The only option that shouldn’t be on the table is not doing anything.