What does LM do?
Label Manager merges product performance data from Google Merchant Center and Google Ads campaigns (Shopping/P.Max) to define how well-performing each product is.
By labeling your feed of products based on performance, it becomes much easier to allocate more budget to sales-driving products and cut costs on those that don’t bring much to the table.
Five different labels correspond with the performance of the products, from best to worst: Winners, Prospects, Jumpers, Sleepers, and Wasters. The picture below explains their meaning in detail:
How does LM help me to reach my campaign objectives?
The main advantages that Label Manager delivers are:
1. Quick control over budgets and ROAS* for Shopping and Performance Max campaigns.
2. Efficient Product Management as an optimal balance between granularity of per-product analysis conducted for you and speed of turning the segments ON or OFF.
3. Automatic Budget allocation gives more to best-performing products and cuts losses on the worst-performing ones.
Based on these advantages, LM users may reach one of the following objectives:
1. Increase ROAS* while keeping the sales volume.
2. Increase sales volume while maintaining the same costs.
3. Decrease costs on demand while keeping ROAS* and sales volume stable.
*ROAS stands for Return on Ad Spend, calculated as Conversion Value / Cost
Are there any requirements for using LM?
As of early 2024, we require the following conditions to be fulfilled before Label Manager will start running:
- A completed Cobiro CSS setup (either created or switched MC) or a standalone configuration, which is described below.
- The Google Ads account is connected to Cobiro CSS MCA and stays connected;
- The Shop runs on at least 300 products;
- Campaign for Label Manager should've generated at least 1000 clicks AND 30 conversions in the last 30 days;
- Label Manager Settings are completed.
In addition to all of the above, it’s recommended to use Label Manager on a campaign that is
- big enough to generate 100 sales within 30 days and
- that advertises products not present in any other active campaign on the same account.
What are the Label Manager Settings
On the first launch, we ask LM users to provide three values to finish the setup. These values allow the tool to understand the overall performance of a given store and produce results most relevant to the business objectives.
- Break-even ROAS: represents the ROAS value, calculated as Revenue / Cost, which the user expects to get out of the Google Ads campaign on which Label Manager will be used.
We suggest multiplying the value by 0.75 to allow LM more flexibility. For example, if the expected ROAS is 400%, after the multiplication, 300% should be put into the setting.
If Label Manager is to be tested on multiple campaigns, we suggest calculating the average ROAS expected from those multiple campaigns.
- Lookback Window: This represents the number of days that a Google Ads campaign needs to generate 100 sales conversions. We suggest setting this value at around 30 to 60 days.
- Custom Label: LM needs to use one of the 5 available custom_label attributes to connect to Google Merchant Center, so the user points out which one can be used to make the connection.
4a What if I only want Label Manager, without Cobiro CSS?
We allow a standalone use of Cobiro Label Manager, which requires a few additional steps upon purchasing a Cobiro Label Manager subscription:
1. Open Cobiro Label Manager
2. Fill in the form of 4 fields. This information will be used for Label Manager configuration
3. Next, we need to get Standard access to your Google Merchant Center to be able to retrieve your products data to our system and Google Ads to retrieve their performance. Please proceed through 5 steps before clicking “Continue”
After that, please continue the Configuration process as normal. If you have any doubts, please check the next section “What are the Label Manager Settings” below.
What if I want Label Manager on multiple campaigns at once
We distinguish multiple scenarios related to this case. Below you’ll find the suggested approach for each of them:
a) Multiple campaigns will be advertising non-overlapping products. When Label Manager is applied to the campaigns separately, there's no interference and Breakeven ROAS, Lookback Window settings should be set as the Average value of these campaigns.
b) Multiple campaigns will be advertising the same products. Then the Label Manager may interfere between campaigns making the results impossible to compare and likely worse in performance. We suggest to keep just one campaign for the test period.
How do I know I used the right settings?
There are a few ways to define if Label Manager settings are optimal. Below you’ll find some of them with our suggestions on how to adjust the settings to improve the results.
Start with a simple observation of segmentation results inside the Label Manager dashboard, which is available right after the analysis is completed:
a) Winners and Prospects segments are too small.
Reason: Breakeven ROAS is set too high.
Recommendation: Lower break-even ROAS to let more products appear in top-2 segments. Ensure that the initial value is about 75% points lower than the actual Expected ROAS (eg. 425% instead of 500%) and keep decreasing by 25% and re-check the results.
Please keep in mind that lowering the value here doesn’t necessarily correspond to its drop in the campaign.
b) Too many products in the Sleepers segment
Reason: The Lookback window is too small so products do not have enough data
Recommendation: The sweet spot here is to set a timeframe in which the campaign generates 100 sales conversions, usually around 30 days. Extend the Lookback window by another week or two. This will allow the Label Manager to get more data for each product to analyze.
How does LM work?
Label Manager connects to Google Merchant Center and Google Ads accounts to combine various data from these two sources. Upon conducting an analysis, a supplemental feed is created and shared through the URL that the user sees in the Label Manager Dashboard.
Adding this supplemental feed to the user’s Merchant Center allows to assign segment values to each product to later be available for selection inside Google Ads.
How to connect LM to Google Merchant Center
This one-time setup requires the user to go through the following steps:
1. Go to the Google Merchant Center Account, which is used for LM/CSS setup.
2. Select “Feeds” on the left sidebar
3. Find the section “Supplemental feeds” at the bottom, and click “Add Supplemental Feed”
4. Add the name you like and select “Scheduled fetch”, then “Continue”
5. Add the name you like at the “File name”
6. Set Fetch time to 12:00 AM (EST/GMT+1)
7. Paste the URL of Labelling Results from the LM dashboard, and click “Continue”
8. Select the primary Feed and a Country to apply Labels to, then click “Create feed”
9. Click “Fetch now” to push the latest labeling results to the Merchant Center.
How to use product segmentation in Google Ads
As soon as product segmentation becomes visible in Google Ads (usually within 1-2 hours from the setup), the user should follow the steps below to include it in the Listing Group/Product Group of the Google Ads Campaign:
1. In your campaign, navigate to "Listing group" (Performance Max) or “Product group” (Shopping)
2. Find the "Edit subdivision" button by the pencil icon and click it.
3. Select the Custom Label used for LM from the list to see the names of the product segments that the LM provided:
4. Select all the checkboxes to see every segment and click "Save".
5. Based on our recommendations in the section “How to split Product Segments between campaigns”, set the appropriate product segments to ON and the rest to OFF. This will ensure that the products do not appear in multiple campaigns simultaneously.
How to split Product Segments between campaigns
Most Label Manager users will be using multiple campaigns to advertise all product segments, which means that the same set of actions has to be repeated for each of these campaigns. It may be faster to apply segmentation to the first campaign and copy it multiple times, so the structure will be identical for all of them.
What’s left is to carefully check segment statuses, as suggested in step 5 above, to avoid overlapping. An example of this tip is demonstrated below:
Scenario 1: Campaign #1 with Winners and Prospects ON, rest OFF
Scenario 2: Campaign #2 with Jumpers and Sleepers, Everything else (optional) ON, rest OFF
Scenario 3: Campaign #3 with Wasters ON, rest OFF
Best performance requires a sufficient amount of products in each campaign. This means that based on your specific store, it’s worth combining several product segments into one campaign to get as close to 300 products in one campaign as possible.
The most common scenario shown above includes the balanced tradeoff between control over budgets for each product segment and the number of items in one campaign.