Google Shopping: Focus on 10% of your products which make 75% of your revenue [how to]
Most online shops treat all products as equal, but, as a retailer, you probably have products which sell better than others. Within each brand or product type there are fast and slow selling products. For exactly this reason Google Shopping allows you to divide products by attributes like brand, product type, etc.
We assume you have some basic knowledge on how to setup a Google Shopping campaign using products feeds. If not, for Shopify users you can create a data feed using Simple Google Shopping Feed App (free) and use Google Shopping Setup Guide to start.
In this article we dive into the product portfolio management. How to find best selling segments and how to focus on 10% of the products which generate 70% of the revenue. We also show how to put these segments into Google Shopping campaigns.
Products segmentation in Google Shopping
In the example below we divided a Shopping campaign as follows: All products (L1) by Brand (L2), and each brand by Product Type (L3). We still have 34 product within each group "chanel foundation". Using this feature we can change bids for each of the groups separately. If one of the brands is more popular, you can spend more per click to get more traffic and more revenue for that brand. If one of the products sells slow you can stop advertising on it by bidding $0.01 and cut costs.
Best performing products
In most cases dividing products by brand or product type is not enough. Not all products with a brand or a category perform in the same way. As experience showed us, only a handful of products from the same brand will sell and the rest of the products will never sell at all.
The AdWords Audit shows us an ABC Analysis of the shopping data. Take a look at the product sales data for past 30 days of an online retailer. We see in A Class that 90 out of 7986 products  accounted for 57.1% of all orders . Plus those 90 products made 74.9% of the monthly revenue . Here is the best part, getting that revenue required only 31.8% of the total adspend .
As a rule of thumb only 10% of your products will generate 75% of your revenue.
Structuring Google Shopping campaigns
We saw that less than 10% of your inventory generates up to 75% of the revenue. These money makers should be treated differently, but so are the products which do not sell at all.
Let's look at the pyramid below. Simply said, there are 3 product segments: (A) products which generate most sales, (B) products which individually are not performing amazing, but are part of a popular brand or a popular category and (C) products for which you just have no idea, because they don't get enough traffic to make a decision.
The top of the pyramid will get roughly 30-40% of the total clicks, the middle part about 10-20% and the bottom part will get 40-60% of the total clicks.
It is a best practice to have a separate Google Shopping campaign for each of the levels. This way you can set a separate budget for each of the levels and the same for bids.
So the question remains which products should go where. What is a top product? What is a middle brand?
Different ways of segmentation
The main reason for product segmentation is to be able to focus on what is important and spend most of the time managing that – thus getting an 80/20 effect. There are 4 common ways of product segmentations.
1) Segmentation by Best Sellers
In most businesses there are only a handful of products which generate most revenue. As a rule of thumb 5-10% of your items will generate 75% of your revenue. Here you concentrate on products which have sold best in the past 30-90 days.
2) Segmentation by Traffic
Another way is to segment products by their search volume or demand. This method lets you concentrate on 20% of products which gain 50-80% of clicks and thus large part of the cost. With this approach you can control your spend better by monitoring and settings your bids on products which actually have the potential to spend more.
3) Segmentation by ROI (Return On Investment)
This method allows you to create different shopping campaigns containing products which have high, average and low ROI. By using this you can allocate different budgets for each and have better control of the cost per click, like lowering your bids on low performing products. Furthermore you can see if there is more impression share (number of time your ads are shown) to gain for the high ROI products - which should get more budget.
4) Segmentation by Gross Margin
In some cases you earn more by selling a certain product or a certain brand. Maybe you have a downloadable products or a products which you produce yourself. $100 dollars earned from those products have a high profit margin and thus can have a higher budget spent for marketing. You can split products with high, medium and low profit margin to have control over your spendings.
Building segmented Google Shopping Campaigns
You can start building segments using your historical data, which will you can use as a guideline.
Let's dive into some statistics.
Step 1. Finding the best performing brands, categories and products
You will first have to look at the Shopping data that you already have, and gain a broader view of the current performance levels.
This can be easily done using the ‘Dimensions’ tab in AdWords.
Under the ‘View’ button there are multiple option to choose from when categorizing the products.
When splitting your data into categories, almost always the best way of doing it is from the bigger categories to the smallest, just to have a better understanding of your account data. This means that first, we will look at the ‘Product type’ category and scale it down from there to Brand, and Item ID.
To see the ‘Product type’ dimension, click on the ‘View’ button, go to ‘Shopping’ and select ‘Product type’. The ‘Brand’ and ‘Item ID’ levels are just underneath it.
Now, you’ll just have to look at the ‘best performers’, ‘low performers’ and ‘everything in between’. No matter at what level are you looking at, use filters to rule out the item/brands you are not interested in including. The key is to use the targets in your filters (regardless if you’re looking for target ROAS, target CPA, or any other targeted metric).
The basic idea behind the three cases is:
Best performers - are hitting or even exceeding your target metrics
Worse performers - are below your target metrics and losing money
Everything in between - are not losing money, but you are not totally satisfied with their performance
Step 2: Getting best items, brands and product types
Use Scraper Chrome extension to get the data from Dimensions which you can later use for creating the shopping campaign.
For example, if you are scraping the ‘Best performers list’, go to the (already) sorted dimensions list, just right click on one of the items, and select ‘Scrape similar’.
In the new window that will pop-up, you will find a syntax in the upper left corner. You will have to replace it with this one: //div/table/tbody/tr/td/div/table/tbody/tr[*]/td
After you run the scrape tool with the new syntax, it will only show the values contained in the first cell of the column for all the rows you have in the table. In other words, you will have all the products ID’s that fit your filter.
With this done you can build a shopping campaign only targeting your best- or worst performers.
Step 3. Creating the Shopping campaigns
To create a Shopping campaign you will have to:
Step 1: Go to your AdWords account
Step 2: Select the campaign level view and click on the ‘+ campaign’ button
- Step 3: From the dropdown menu, select ‘Shopping’
- Step 4: After you fill in the specifications you would like to use for the Shopping campaign (locations, daily budget, bids, etc.), you will have to work at ‘Ad group’ level.
Important: when creating the campaign, you should select the priority for each campaign:
High: for best performers
Low: for low performers
Medium: for everything in between
To do this, you will have to click on the ‘Shopping settings (advanced)‘ button and select the campaign priority.
Bare in mind that you will have to create a campaign for each level of performance (best, low, and in between). Let’s use the best performers as an example.
Step 4. Structuring the campaign
To visualize the ad group, you’ll have to click on the newly created campaign, and click on the ad group (in this example it’s named ‘All products’)
Now, click on the ‘Subdivide’ button, select ‘Add bulk values manually’, and paste the list that you have previously scraped with the tool, and click save.
After you saved it, you are done...at least with this one.
Using the same steps you can create different shopping campaigns covering different segments. Don't forget to set the budgets and bids for each campaign so those reflect desired performance.
Automating product segmentation and enriching the shopping feed data with custom labels
Finding the right segments and keeping them up-to-date can be a time consuming process – especially if you do it on monthly basis. For this reason we automate this process by enriching the data feeds with sales data which is part of the AdWords Robot management services.
Custom labels are the perfect tool for this and the labels can be used in AdWords to segment your product inventory in different campaigns or product groups.
Custom Label 0: Defines the price range. Using this label we can bid more on products which are not too cheap or too expensive.
Custom Label 1: Identifies the cheapest variant (or the most expensive variant) of all variants of the same product.
Custom Label 2: We divide all revenue in 4 quarters (Q1..Q4). Q1 contains all products which generated first 25% of the revenue. Q2 contains all product which generated next 25% of the revenue and so on. Basically products tagged with Q1..Q3 generated 75% of all revenue. Those are the best sellers.
Custom Label 3: This label shows top 3 best selling variants (of the same product). Can be useful to promote most popular variants.
Custom Label 4: This label shows top 3 variants (of the same product) which generate most revenue.
ABC classes (based on product revenue)
Using the ABC-analysis you can segment each product in one of the classes. Class A contains the money makers. This small selection contains 5-10% of the products, but earns 75% of the revenue. Class B earns about 15% of the revenue with only 10-15% of the products. Class C contains the rest.
By concentrating on Classes A & B in your campaigns you can focus on 10-25% of your product inventory which generates 90% of the revenues. This way your budget and your time best spent.
Below we see that Class A contains 90 (1.1%) products, and generates 57% of all conversions. Furthermore this class consumes 31.8% of all cost and generates 74.9% of all revenue. Class B contains 73 products and generates 15% of the revenue against 7.3% of the costs. Class C contains 98% of all products and consumes a staggering 61% of the costs and generates only 10% of the revenue.
We can conclude that concentrating on the important few products can be a good strategy to get most bang for your buck in Google Shopping. Often products in Class A are the stars of your assortment – maybe because of the price, maybe because of the availability or maybe because not that many merchants sell them.
Note here that products move from class to class during the year. Maybe a new competitor is having a huge sale or due to seasonality. For this reason it is recommended to recompute all classes every 1-4 weeks, depending on the number of sales per week and the number of products.