Intro
The basic unit of Product Insights is a product filter. You can filter products by markets, channels, campaigns, product attributes, and performance metrics. The design of Product Insights allows you to easily combine multiple filters into a single query to answer questions about your product inventory performance.
Configuration of Metric Filters
The filtration panel on the left side of the Product Insights menu is the main source of input for composing analytical insights.
To open the metric configuration panel, click "Add first metric" under the "Metrics" heading.
This panel can also be opened by clicking on metrics in the top of the graph view.
From here, there are three UI components important for calibrating your product metric:
List of available metrics (Left); the source icons to the right of the metric name indicate the possible channel(s) (e.g. Meta and/or Google) and type of metric (e.g. analytical metrics are indicated with a chart symbol and calculated metrics are indicated with a 🖩 calculator symbol)
Note: If a source icon is greyed out, hovering over the icon provides a possible solution to enabling source data (such as connecting relevant assets in the Market settings).Metric calibration settings (Middle); The settings can be limited for a deeper dive for more specific results, while unavailable options are greyed out.
Metric description (Right); text definitions clarify any acronyms and also may specify metric components such as the source(s) and/or ad type(s) from which the metric draws data.
Available Metrics
Product ROAS: This hybrid metric unique to ROIH allows you to evaluate the overall performance of each given product and see whether the revenues match the ad investment or not. Available for Meta and Google sources, and calculated using product data from Meta and Google.
Note: This is distinguished from other ROAS metrics (such as campaign ROAS), which is without the product-specific context.Meta metrics: Scoped to Dynamic product ads, and product-level data is taken from Product ID breakdowns of the Dynamic Product Ads. Criteria include:
Amount spent
Impressions
Clicks
Link clicks
Clicks to the associated business page profile or profile picture
Post reactions
Comments or shares
Clicks to expand media to full screen
CTR
Meta Product ROAS
Google metrics: Scoped to Performance Max (PMax) and Shopping campaigns. Criteria include:
Cost
Impressions
Clicks
CTR
Google Product ROAS
Analytical metrics: Scoped from Google Analytics, Branch, or Adobe Analytics that are paired on a product level. Criteria include:
Conversion rate
Custom metrics
Item quantity
Product average sale price
Product detail views
Product revenue
Quantity added to cart
Product ROAS
After calibrating your metric, click the "Use" button in the lower right to apply it. Additional operators and values may be input in the same panel:
Additional metrics can be included by clicking "Add next metric."
These combinations of filters can be saved as Segments be clicking "+ New Segment" at the top of the filtration panel menu:
Usage
Subsequent parameters may be applied in other filters, to narrow the scope of data discovery.
Campaign Filters
This filter targets insights on the campaign level, and therefore defines the scope of subsequent filters.
To apply a Campaign Filter, select the drop-down menu and select any desired campaign(s). Campaigns are separated by channel source, but can be mutually selected.
Product Attribute Filters
These filters characterize your products based on factors such as brand, category, price, gender, size, name, etc. Custom Attributes are also possible to integrate and customize for more fine-tuned insights.
For more information on connecting and using Product Attributes (both conventional and custom), read here: Business Data Integration
To apply a Product Attribute filter, click "Add first attribute":
Next, select the relevant attribute. To further focus the filter, specify from the associated entries (which can be narrowed down by applying conditions; the remaining products filtered in are counted in the bottom right):
|  |
|  |
|  |
| 
|
|  |
|  |
Strategy Through Granularity
The power of effective filtration comes from applying these tools towards a practical outcome, such as identifying a source of inefficiency which can then be resolved to increase product performance in some manner. The goal of assessing a practical outcome is assisted with product granularity; this is essentially recognizing that the same product (e.g. shoes) may have variants (e.g. sandals or sneakers) which may perform differently but unintuitively receive the same advertising approach.
To isolate product variant performance, it's important to understand the difference between product-level data and product group-level data. Their relation to category levels is crucial to understand for successfully applying filters towards insights. Consider the following example:
You are interested in which types of shoes are performing well.
To do so, you have to drill down into the category of shoes (e.g. category level 1) by filtering that category via the filtration panel or the chart.
Only then can you see the performance of its subcategories, like sandals or sneakers (category level 2), on the charts and as individual products in the table.
This interaction of product granularity between the product-level and product group-level leads into the significance of the correct product inventory setup process; to be able to measure the factors that are relevant to your business model, those factors must be represented somehow (e.g. with custom Product Attributes) in the data being filtered.
Effective strategy is important for achieving business goals. Read more about optimizing for revenue or profit here.