One of the most common approaches to setting up Shopping campaigns is to segment everything by brand.
By doing it this way, retailers can improve their control over ad spend while also maintaining a more streamlined approach to managing bids and other campaign optimizations.
There are, however, quite a few pitfalls that we see retailers falling into when attempting this setup, especially if it is their first time setting up campaigns.
We wanted to share 3 top tips to help you avoid these common issues and improve your effectiveness of managing your campaigns by brand.
Tiered Shopping Structure
The more products you have, the more important it is to develop a manageable structure for both bidding and analysis. This is best achieved by combining methods to take advantage of Tiered Shopping Structures.
It’s important to note, this is best suited for anyone with over 200 products in their campaign.
With this method you will first take one “top-level” or Primary grouping structure for the entirety of your catalog, and then break that down once or twice again until you achieve the structure you wish to base your bid management and product analysis on.
For example, here’s how that works if I were to start with Brand as my first grouping structure when in Google Ads:
- Check the top-most box to select all brands
- You can leave your default bid which you set at the ad group level earlier so all you have to do is click SAVE WITHOUT EDITING BIDS
Before we move on scroll down to the bottom and take note of the Everything else in ‘All products’ group. Google Ads has created this automatically and it essentially acts as a catch-all for any products that did not match your original subdivision criteria.
In this example, it could be products that were submitted without a brand. If all your products have a brand, this everything else group should have nothing in it. For that reason, you can exclude this product group as it is unnecessary.
Never Forget Granularity
From here you have two choices:
- RECOMMENDED: Subdivide again to achieve product-level granularity in each brand group
- NOT RECOMMENDED: Don’t subdivide any further and set brand-level Max CPC bids
I don’t believe there is any one right way to handle subdivision (the process of structuring a Shopping campaign). All structures come with their own inherent strengths and limitations.
However, the most important element to campaign structure lies in maintaining granularity.
Getting granular allows us to:
- Set unique bids for each product or build very small groups of products and set the same bid for items with similar qualities or shared characteristics such as price or category.
- Exclude (pause essentially)products that maybe we don’t want to advertise due to razor thin margins or a lack of inventory.
- Monitor product performance at a much deeper level so you can understand exactly which of your items are driving most of the conversions versus those that are merely wasting your budget.
Now, there are a lot of ways to structure a Shopping campaign. Every marketer or agency has an approach that they prefer. Some structures are very simple whilst others are incredibly complex.
Even if you are grouping products initially by brand, one of the worst things you can do is keep the default grouped structure without any further subdivision.
Without granularity, you risk overspending on products that don’t convert. Other phenomena such as Impression Crowding can occur as well, especially when you have thousands or more of the same brand of product lumped together with the same bid set across the board.
So, for every brand that you carry (which is one product group each) always remember to break down each group even further to the product level:
Brand-Specific Campaigns or Ad Groups
We’re going to shift gears here a bit and, using what we’ve learned already, take a slightly different approach.
Previously, my suggestions above all revolved around using a single campaign for all your brands. In that, each brand would be its own Product Group. If you followed further, you would then have each product broken out individually for each group.
While maintaining our constant theme of granularity, you can also consider creating specific campaigns per each brand OR separating brands by ad group. Doing this provides some additional benefits:
- Budget Allocation: Not all brands perform equally, and not all brands have the same margins. By separating brands into individual campaigns, you can apply a unique budget to each and monitor the brand performance at the campaign level versus having to dive further into the campaign to ad groups and then product groups.
- Device Modification: There are no “product group bid adjustments” by device (Mobile, Computer, Tablet). These modifications are made at either the campaign or ad group level. With separate campaigns, you can implement and monitor device performance per campaign. If you were to separate by ad group, you can do very much the same but also keep one campaign for all products.
- Search Query Management: This would work for either the campaign or ad group approach. Having brands broken out in either option will give you additional levels of control over your search query analysis and usage. This comes especially handy when you want to build a negative keyword strategy that focuses on the ad group level. In a single campaign approach, with no ad group distinction (pure product level), negative keyword management must be far more strategic. It can be done a bit more easily with the segmented ad group approach.
With the benefits come a few caveats too though.
Splitting up your brands into multiple campaigns forces you to also split your budget between them or you’d have to use what’s called a shared budget across all campaigns. If you have a smaller budget to work with then this may not make the most sense.
If you sell a lot of separate brands, an ad group specific approach can make things a bit time consuming to manage as you must analyze each brand, then open up each ad group, find top performing SKUs, etc.
Still though, both offer a very organized approach that can be easier to manage in their entirety as you can take things a bit slower and be more systematic in your implementations.