Producthero Labelizer
Automated daily product segmentation for Shopping and Performance Max campaigns to label and prioritize SKUs for campaign structuring.
What it does
Producthero Labelizer automatically segments a retailer’s Shopping catalog into performance-based labels and makes those labels available for Google Merchant Center and Google Ads workflows. Marketers use it to group SKUs by volume and profitability so they can structure Shopping or Performance Max campaigns around those groups. Daily automated labels Labelizer classifies SKUs each day into four rolling categories (Heroes, Sidekicks, Villains, Zombies) based on click volume and an internal Producthero target ROAS. Labels update automatically as product performance changes so teams see current priority groups without manual recalculation. Push labels into Google The tool provides a workflow to expose generated labels to Google Ads via the Merchant Center (supplemental feed or custom label). Marketers map those labels into campaign structures or asset group rules so different labels can be treated with separate campaign groups or bidding strategies. Manual tagging alongside automation You can add manual tags (for example, Seasonal or High margin) that are stored as separate custom-label values. This lets teams override or augment the automatic segmentation for specific promotions, clearances or strategic priorities. Strategy controls and thresholds Label assignments depend on two configurable thresholds: a clicks (volume) threshold and Producthero’s internal target ROAS. These controls let advertisers tune how aggressively products are categorized and adapt the segmentation to catalog size and business goals. Scope and data considerations Labelizer’s model uses Shopping ad streams for its analysis; conversions or traffic that originate from non-shopping assets inside Performance Max or other channels are not included in the classification. Also, Producthero’s internal target ROAS is used for labeling and is not passed through to Google Ads as a bidding signal, so labels are a segmentation signal rather than an automated bid input.

