When most agencies or raw AI tools build a Google Search campaign, they start from a static brief or a prompt. Someone types a short description of the brand, guesses ten keyword themes, and writes generic headlines that could apply to any store in the category. Two weeks later, the merchant discovers that half the ad spend went to a collection where four out of five sizes were sold out, and the ad copy quoted a starting price that changed two months ago.
For a catalog-scale ecommerce store, Google Search campaign construction must start by reading the live store itself: products, collections, variant stock levels, real prices, and on-page proof.
Why static briefs and chatbot prompts fail catalog stores
A store with thousands of SKUs is a moving target. Prices shift when supplier costs change, seasonal collections rotate in and out, and hero sizes sell through unevenly. A language model prompted in a vacuum has no idea whether your best-selling espresso machine is currently $449 or $529, whether your waterproof hiking boot is in stock in common sizes, or which collections actually carry enough inventory depth to absorb paid traffic.
When campaigns are drafted without reading the live catalog first, three expensive errors happen immediately:
- Ad groups launch for products or collections that are out of stock or down to broken end-of-season sizes.
- Headlines and price assets state outdated prices or promotional discounts that no longer match the landing page.
- High-margin, well-stocked collections sit completely uncovered while budget crowds into a handful of obvious category terms. You can see how this gap develops in our guide on search demand with no campaign pointed at your high-volume SKUs.
What a pre-launch store and collection scan must capture
Before choosing keywords or writing a single Responsive Search Ad, an operator or automated system should extract five concrete data points directly from the target store URLs:
First, capture the catalog hierarchy: which collections exist, how many active products sit inside each collection, and which individual SKUs anchor the category.
Second, verify live price ranges and active offer rules. If a collection starts at $68, the ad copy and price extensions must reflect $68, not an invented "from $49" hook that triggers buyer distrust and mirrors the problems seen in price and availability mismatches in Merchant Center.
Third, check stock depth at the variant level. A product that is technically "in stock" because one XXS unit remains in the warehouse is effectively out of stock for paid acquisition.
Fourth, extract real product specifications, materials, model numbers, dimensions, and verified customer review counts from the page so ad copy is grounded in catalog facts.
Fifth, identify non-commercial URLs on the domain so blog articles, shipping info, and return policy pages are fenced off before any campaign structure is assembled.
How Kopa scans your store in Stage 1 of Smart Campaigns
In Kopa, launching Google Search starts by pasting your store or collection URL into Smart Campaigns. Kopa scans your live products, collections, prices, stock levels, and on-site USPs directly from the store before drafting ad groups, verifying keyword demand in Google Keyword Planner, or writing ads. Your team reviews the complete draft before anything publishes to Google Ads.
What good looks like
- Every new Google Search campaign starts from a live scan of target product and collection URLs rather than a static spreadsheet.
- Products with broken size runs or low variant stock are held back from paid Search launch until inventory is replenished.
- Starting prices, review counts, warranty terms, and shipping thresholds used in ad copy are pulled directly from live store pages.
- High-stock, high-margin collections are mapped to dedicated ad groups so catalog coverage matches actual inventory depth.
- Price or availability changes on the storefront trigger a review of active Search headlines and price assets.