Google-Shopping-logo

In this post I will look at the best practices when using different product groups within a Google Shopping campaign, plus two bonus methods to increase return on advertising spend even further.

Why you need to divide a Google Shopping Campaign

The top product group level contains all of the products that can be divided by different attributes such as brand, category, product ID, colour, etc. We can set the maximum cost-per-clicks for each product group but the average conversion rate of those products and the average costs of the products are out of our hands within Google AdWords:

new_used (5)
New and used products will have different statistics and should use different maximum cost-per-click bids
many_levels (2)
Differently coloured products may have different conversion rates, so there is a need different Maximum CPC bids

Without product division, every single product added to the campaign would have to have the same maximum cost-per-click bid level. This means that without product divisions the entire Google Shopping account will be based off the average performance of all the products bunched together, which isn’t good for optimisation.

Let’s pretend that you had only two types of products to sell on Google Shopping; barbecues which sell for £300 each and barbecue accessories which cost £50 each. If the Google Shopping feed was not divided then all products would be combined within the same product group:

Not dividing a Google Shopping feed results in general average statistics and a lack of control
Not dividing a Google Shopping feed results in general average statistics and a lack of control

When setting the bid levels we want a good return on investment, comparing the average revenue generated from product sales versus the average advertising costs. As in the example above, having such wildly different priced products within the same group creates an issue while setting bid levels:

  • The lower priced item will have a relatively high cost-per-click and may have higher advertising costs than it generates in terms of profit – Advertising costs would be wasted
  • The higher priced item will have a relatively low cost-per-click and therefore will have a low ad position, resulting in very few clicks overall – Exposure is reduced and potential sales would be lost to other advertisers

Having separate bids for the two separate product groups would work as shown below. We now can set the higher and lower bids appropriately according to the conversion rate and the average product cost:

bbq (4)
Large bids can now be placed on the highly converting & priced products and vice-versa

Dividing all products down to a product level

The furthest the shopping campaign can be divided down by is the individual product level:

You can go as far down to individual product IDs whilst creating product groups
You can go as far down to individual product IDs whilst creating product groups

Dividing down to the individual product level gives the greatest control on bid levels and the clearest statistics.

Using ecommerce tracking with Google AdWords allows you to see the exact costs and revenue generated from individual products. This allows you to target an ideal return-on-investment level based off the relationship between profits & loss.

Best and worst selling products can also be individually singled out within a product group sub-division. Products that aren’t individually extracted will be catered for with a bid placed for “Everything Else” within the product group:

All other brands of smartphone will be contained within the "Everything Else" product group as they haven't been specified
All other smartphone brands will be contained within the “Everything Else” product group as they haven’t been individually specified

For many eCommerce sites it may not make sense to reduce the shopping campaign down all the way to the individual product level. There are a few big reasons for this:

  • There may be hundreds or even thousands of products to manage, and not enough time each month to manage them all
  • Each product may get so few clicks that it’s impossible to get any significant statistics from it in a sensible time frame
  • Products may be in and out-of-stock so quickly that there isn’t enough time to gather any significant statistics from them before they are gone

In these cases, you would have to rely on more generic product groups in order to optimise the campaign fully.

 

Dividing all products into product groups

Product groups are useful as they split products into meaningful segments in which they can be separately optimised with different bid levels.

Product dividers can be used be in any set order on Google AdWords. For example, you could split all products by brand and then by product category, or vice versa:

Products divided by category and then by brand
Products divided by category and then by brand
cat_brand (2)
Products divided by brand and then by category

It’s important to note that the Maximum cost-per-click bids are only set at the furthest level down the product divisions. If the child product group does not have a set Max. CPC then the Max. CPC set at its parent group would then count.

When setting up product groups, it’s best to divide the products as if they were divided inside a traditional brick and mortar shop. For example, a shoe store would typically sell men’s shoes and women’s shoes in different areas of their store. Within each section, there will be different sub-sections offering different categories of men’s or women’s shoes. Repeating this store layout within Google Shopping product groups gives the ideal segmentation shown below:

shoes (1)
Split up the products as you would in a real-life store. Dotted lines signify further hidden product groups

The product group order could easily be set the opposite way around, but this can cause issues when analysing the statistics taken from parent group levels. Whenever a split is made, the statistics for all the products within the split are revealed:

shoes2 (2)
Now the “Trainers (Sneakers)” product group has been split once more into different brands

In the example above, we can see that when thinking of a maximum cost-per-click bid for “Women’s + Trainers + Nike” we can base the bids off the similar parent product group. In this case, the parent product group is “Women’s + Trainers”, a very similar topic.

 

Dividing All Products with Custom Labels

If you are able to customise your Google Shopping feed then custom labels can be used to divide products in to even more precise groups for further optimisation.

Custom labels allow you to add an extra piece of information to each item within a Google Shopping feed. The information can contain anything you wish and there can be up to five sets of custom labels within a shopping feed.

 

Cost Bracket Custom Labels

One of the most important custom labels you can implement is one for different price brackets:

The product group "Laptops" is then split up by product price bands
The product group “Laptops” is then split up by product price bands

It no longer matters how many products there are within the shopping feed, as we can now easily place higher bids for items with higher profit margins and lower bids for items with smaller profit margins. This promotes the items with higher profit margins and saves over-spending advertising costs on the smaller items, a double positive effect overall.

It’s best to use cost bracket custom labels at the very end of the product group divisions. This allows for bids to be based off the average statistics of the parent category if needed.

 

Special Offer Custom Labels

Seasonal promotions and limited time offers can play havoc with Google Shopping campaigns. Often a discounted product would require a lower bid amount to be set to offset the lower profit margin, but at the same time, it will be more likely to sell, so it will require a higher cost per click to take advantage of the increased conversion rate!

Special Offers = Higher chance of selling but lower profit margins per sale

The advantage of higher conversion rates may not equal the disadvantage of lower product prices, so it’s important to separate special offers so bids can be raised or lowered accordingly:

In this example the sofa beds on sale need a higher Max. CPC
In this example the sofa beds on sale benefits from a higher Max. CPC due to the high conversion rate despite lower revenue from the product sale

 

Conclusion

Play around with splitting up your Google Shopping campaign into different product groups to discover what works for you best. There is no penalty for re-arranging a Google Shopping campaign’s product groups, and product statistics will be kept intact within the same campaign on AdWords.

The main goal is to split the products up into very similar groupings so bids can be placed appropriately to maximise profitability.

Please let me know if you have any questions in the comments section below.