Reading time: 6-8 minutes
* This article about working with multiple CSS partners is a direct translation from the original, Dutch version which can be found on the Emerce website. In this article, you can find references to Kiesproduct, which is Shoparize’s Dutch brand name.
During the search for a Google CSS partner, you come across terms like discount and cashback. Usually, they are talking about a Google CSS partner’s self-service model. With a self-service model, merchants can save Shopping ads costs directly due to the 20% bid advantage
But what is the managed service model through which shopping activities are partly outsourced to a CSS partner? Does this have added value for your client? In most of the situations, it can even lead to more revenue and a higher ROAS from Google Shopping, as it did for Conrad as well.
What is Google CSS managed service?
There are several ways in which a Google CSS partner and a merchant can work together. One of the well known strategies is that merchants advertise in Google Shopping through a self-service model, whereby merchants stay in control of their Merchant Center and the Merchant Center is linked to a Google CSS partner. At all times, the merchant manages their Shopping campaigns and billing from their Google Ads account.
Most of the time, merchants also use a Managed service model which runs in parallel with their in-house Google Shopping activity. With this service model, the Google CSS partner manages the Google Shopping campaigns and billing on behalf of the merchant.
Since the Google CSS partner also finances the campaigns, they are reimbursed based on CPC (Cost-per-Click), CPS (Cost-per-Sale) or Adspend model. The most popular is CPS, because this no cure no pay model is the most transparent collaboration as it does not involve any risk for the merchant. This collaboration on CPS is usually set-up through affiliate networks and can lead to additional shoppers and sales without competing with the merchant’s current campaigns.
You don’t want cannibalization
It is assumable that you would compete against yourself with an extra Google Shopping campaign. That is not desirable. That is why we examined how Google CSS partner’s campaigns influence their own Google Shopping performance.
Does bidding against each other really take place when working with multiple CSS partners? Are clicks and sales being hijacked by Google CSS partners, or in other words, was there cannibalization?
We had to put this to the test. There are enough theoretical articles about this subject, but we could not find any concrete customer cases that support this. That is why we, together with Conrad, launched Google Shopping campaigns via our Premium CSS partner Shoparize. These campaigns ran beside the Shopping campaigns from Conrad via Google CSS.
You never compete with yourself
In the official source documentation, Google shows how the Google Shopping auction introduces Google CSS partners in the equation. Google designed a second-price auction in which a merchant can never compete against itself with the same bid and working with multiple CSS partners. Google also answers the question: Do I have to pay more for Shopping ads if multiple CSSs advertise on my behalf?
Google’s answer is: “A merchant will never be second-priced against itself in the auction for any offer, and that rule holds irrespective of the number, or identity, of the CSSs used by the Merchant. Therefore, if two CSSs place bids on behalf of the same merchant, the winning offer and the price paid by the winning CSS will be the same as if those bids had been placed by one and the same CSS”. “Google support”

Besides Conrad’s Google Shopping campaigns, Shoparize managed a campaign being the only Google CSS partner for Conrad, based on identical product feed. For a period of eight weeks, we tested the following assumptions:
- Does Google Shopping CSS managed-service model result in incremental sales?
- What happens to the ROAS when two Google Shopping campaigns advertise for the same website?
- Does the performance of your own Google Shopping account suffer from working with a Google CSS managed-service?
The test is evaluated based on the customer’s KPI’s: revenue and ROAS (where the costs are equal to the own Google Ad Spend plus the CPS fee paid to Shoparize CSS).
On top of that, we evaluated the transactions based on Google Analytics 360’s last click. To assess the results, we provide insight into revenue and ROAS per week.

The red curve shows revenue and the blue bars show ROAS.
The results are very positive. Both the revenue and the ROAS increased significantly in the test period. At first glance, there is no question of cannibalization.
The positive test results can be explained in different ways. Did Shoparize focus on the end of the funnel to, for instance, advertise EAN codes and product names? Did Shoparize focus on high-priced products or rather on products with a low order value, like the wide range of Conrad’s LED lights?
The answer to these questions is “no”.
The assisted conversions from Shoparize’s campaigns were even higher, despite leaving out “Conrad” in the search terms.
The assisted conversions were an important factor in the results because the results were initially evaluated on the last click, which does not always tell the complete story.
How does Conrad (Shoparize CSS) vs Conrad (own Shopping campaign) relate in the Google Shopping auction?
To get a realistic view of signs of cannibalization, we further investigated the auction statistics from the Shoparize CSS Google ads account.
This shows that Shoparize CSS is responsible for an impression share of 16,02%. The impression share of Conrad’s campaigns is higher, namely 45,26%. You would think that Conrad has the highest impression share, but that is not the case. It appears Conrad has multiple competitors who take on a considerable impression share in the auction.
The outranking share is only 11,97%
Conrad is represented in 34,43% of Shoparize’s Impression share in the same auction. Shoparize’s outranking share in comparison to Conrads is only 11,97%. The outranking share shows how many times your ad is placed in a higher position than another contestant, how many times your ad is shown while the contestant ad is not, divided by the total amount of ad auctions where you participated in.

Blue shows “overlap rate”, light green shows “outranking share”, and dark green shows “without overlap”.
Based on these statistics, we can draw the following conclusions:
- About 53,60 – 65,57% of the Shoparize CSS traffic from Google Shopping for Conrad could never be generated by Conrad itself. Simply because they were not part of the auction. The sales resulting from this are completely attributed to Shoparize’s shopping campaigns. We can speak of verifiable incremental sales.
- From the overlapping Shoparize traffic of 34,43%, the vast majority did not go to Conrad. However, it is hard to give an exact percentage because numerous big players were active in the auction. With Shoparize, Conrad increased visibility in the auction because Conrad would never obtain the same amount of clicks when entering the auction only with Google CSS.
- Because Shoparize CSS realized higher positions than Conrad, you would assume Conrad missed out on sales due to Shoparize. This is not the case because multiple contestants in the auction also obtained higher positions than Conrad. If Shoparize had not participated in the auction, this traffic probably would not have gone to Conrad, but a competitor.
- Of the 34,43% Shoparize CSS traffic, little overlap occurred with Conrad. Because multiple advertisers show similar impression shares / outranking shares with Shoparize CSS, it is highly unlikely that Shoparize competes more with Conrad than with other advertisers. Furthermore, there is no indication that Conrad’s and Shoparize’s ad positions are directly related.

Conclusion: Does cannibalization happen, or can you safely profit from managed CSS service?
Conrad enjoyed the full benefits of Shoparize’s managed CSS service and, for the most part, Shoparize CSS did not cannibalize with a second Google Shopping account. Based on the test results, we can conclude that Shoparize CSS generated incremental sales and an increase in revenue. Furthermore, Conrad benefitted from a higher ROAS from Google Shopping, more Conrad.nl visitors, and extra visibility in Google Shopping.
How can you apply this case to your webshop?
Shoparize CSS is easy to set up through your affiliate network as you only need to accept Shoparize as publisher to your affiliate program. A test period will show how much incremental revenue Shoparize CSS can generate for you on top of your current Shopping activities. For the performance, factors such as commission percentage, cookie length, and attribution are very important.
The attribution model you choose, last click or linear, will significantly influence your final evaluation.
The commission percentage you choose also influences your performance. Increasing your commission percentage results in more competitive bids in the Google ads auction, resulting in higher rankings, more traffic, and ultimately more sales.
In preparation for a test, it is advised to execute an EAN-overlap analysis. This way, you can map competitors that enter the auction with a similar product feed.
Take advantage of these benefits with Shoparize CSS. Allow Shoparize to help grow your business by improving the performance of your Google Shopping campaigns. Make the switch today, by following the steps in this article or sending an email at css@shoparize.com. To learn more about working with multiple CSS partners, read our article on this topic.
This case study was realized with the help of Conrad, Budgetverf, Hunkemoller, Daisycon and Shoparize. You can find the original, Dutch version of this case study on the Emerce website.