The Cost / conv. column divides total cost by the number in Conversions. For a store owner, a high Cost / conv. in one group still does not name the product or the buyer. You first need to see how the group itself is built.
At an online retailer of heating and thermal equipment in Ukraine, our team found insufficient separation of products and audiences in the shopping campaign. The ad group without an audience signal showed a high conversion cost. That pointed to a problem with the campaign structure.
Where the signal sits in the account is covered in the article on Performance Max asset groups. This page is one audit: products and audiences were not split, and the group without a signal had a high cost per conversion.
What we checked
I am not naming the company. Our team checked whether products and audiences were separated in the shopping campaign, and what conversion cost the ad group without an audience signal showed. For the owner, that is how you tell whether a high conversion cost sits on the whole catalog at once.
I have run ads since 2006, and I often see the whole catalog left in one group, with the audience signal empty or shared across different products.
What the signal and the product groups mean
An audience signal is a suggestion, not a limit on who sees the ads. You add audience suggestions so that Google AI can guide a Performance Max campaign toward the goals you select. Adding a signal is optional. The campaign may show ads to people outside the signal when they are likely to convert. That is what Google’s page on audience signals says.
The signal usually sits on the asset group, in the Audiences section beneath it. The group gathers the ad materials and this signal around one theme or one audience. The path is on the pages about adding audience signals and about building an asset group.
Conversion cost in the report is the Cost / conv. column. It is the average cost of a conversion: total cost divided by the number in Conversions. The definition is on the page about conversion tracking data. “High” here is the team’s reading of that column.
Products are split apart from audiences. In a standard Shopping campaign, product groups live in the ad group. At the start, the help guide gives you one All products group for the whole catalog and recommends smaller groups. The same page also mentions the Product Groups tab for Performance Max. That is product groups in a Shopping campaign.
In Performance Max, products are set with listing groups inside the asset group. How to split a retail catalog is in the article on shopping product groups and priorities.
How we checked
Our team looked at whether products and audiences were separated, and at what Cost / conv. showed for the group without a signal. The high cost on that group was read as a problem with the campaign structure.
What we found
Our team found insufficient separation of products and audiences in the shopping campaign. The ad group without an audience signal showed a high conversion cost, indicating a problem with the campaign structure. The team recommended creating a separate campaign with product groups and matching audience signals.
Why it matters
A store owner assesses the budget by orders, not by the click. When products and audiences sit in one pile, a high conversion cost does not say which part of the heating catalog produces it, or for which buyer. Adding or cutting budget then rests on the mixed group.
A heating catalog is not one demand all year. How to split seasonal ads for this kind of equipment is in the article on ads for pools and heating. Season was not a separate finding in this audit.
What to check in your account
- For Performance Max: in the Campaigns menu, open Audiences. Find the asset group and the Audiences section beneath it. Note the groups that have no signal. In a standard Shopping campaign, check the product group split inside the ad group instead of the signal. The help guide describes the audience signal for Performance Max.
- For Performance Max, open a saved signal in the Summary of that asset group, using the pencil next to the signal. Sources include customer lists, website visitor lists, and custom segments. A new signal is built with Select an Audience or Create an Audience, then Add your data and Save Audience.
- For Performance Max, add the Cost / conv. column next to Conversions and read it on the asset group without a signal. In a standard Shopping campaign, read the same column on the ad group. In this audit the conversion cost on the group without a signal was high.
- For products in Performance Max, go to Campaigns, the campaign, Asset groups, Listing groups, and the pencil next to All products. If the catalog is still one group, the help guide recommends a split. Attributes include Product type, Brand, Item ID, and Custom labels, then Save. Subdividing by Category in Performance Max is only for a campaign that targets the US, the UK, Australia, Germany, France, Japan, Italy, the Netherlands, Brazil, Norway, Sweden, or Turkey. Ukraine is not on that list. In a standard Shopping campaign, the product group split inside the ad group includes Category. For that campaign the help guide does not give this country list.
- If products and audiences are not separated, and the group without a signal is expensive on Cost / conv., the question is structure. In this store the team recommended creating a separate campaign with product groups and matching audience signals.
If you see the same pair in your own account, a high conversion cost on the group without a signal and products that are not split, start with the steps above. The step-by-step version is in my book on Google Ads for business owners.
My e-book “19 Google Ads Secrets”: how to get the most out of Google Ads without draining your budget, drawn from real campaigns and tests.
See the book →When this does not apply
This conclusion is not about your account if products already sit in separate groups and each of those groups has an audience signal for those products.
This conclusion is not about your account when the group without a signal does not show a high conversion cost. The help guide allows an empty signal. The problem in this audit was insufficient separation together with a high conversion cost on the group without a signal.
This is an audit finding, not a result after changes. The team recommended a separate campaign. This audit does not say the campaign was launched, or what happened to the cost. It also has no comparison with the group that had a signal: no sums and no difference.
Questions and answers
What did the group without an audience signal show?
It showed a high conversion cost in the shopping campaign of an online retailer of heating and thermal equipment in Ukraine. The column is Cost / conv.: cost divided by the number in Conversions.
Why is this a campaign structure problem?
Products and audiences were not separated enough, and the group without a suggestion had a high conversion cost. The team read that as a campaign structure problem: in a group that is not split, you cannot see which part of the catalog produces that cost. The team recommended a separate campaign with product groups and matching audience signals. This audit does not compare that recommendation with a split inside the same campaign.
How do you separate products and audiences?
In a standard Shopping campaign, split product groups inside the ad group. In Performance Max, split listing groups inside the asset group. The starting group is called All products. Add the signal in Audiences under the asset group those products belong to, and save it with Save Audience.
What about the group that had a signal?
This audit did not show that group. The write-up has the group without an audience signal and insufficient separation of products and audiences.