Category: Positive Sparks News
Platform: WordPress
What do you do when a health supplement brand has 200 products in Google Merchant Center and still captures 0% impression share on competitive queries? You rebuild the feed from the ground up.
That was the starting point for one supplement brand we worked with. On paper, the catalogue was live. In reality, Google had very little confidence in the data quality, relevance signals, or commercial priorities behind it. The account was generating about $12,000 per month in Shopping revenue at 1.8x ROAS, but it was nowhere near its potential.
The starting point
Why would 200 products produce so little visibility? Usually, it comes down to weak structure rather than weak demand.
In this case, the feed had a 34% completeness score, inconsistent naming conventions, thin attribute coverage, and no segmentation layer for bidding logic. Some titles led with internal SKU-style phrasing. Others buried the product benefit. A few variants had size details, while others didn’t. You know how this goes: Google sees incomplete context, and the auction gets harder.
Actionable takeaway: Pull a sample of 25 titles from your feed and check whether they follow one format. If they don’t, your catalogue is probably underperforming before bidding even starts.
The feed rebuild
So, what did we change first? We standardized every product title using one consistent formula:
Brand + Key Benefit + Product Type + Size/Quantity
That meant product names were written for both search relevance and shopper clarity. Instead of vague or inconsistent titles, we gave Google a reliable pattern it could interpret at scale.
We also added 7 custom labels across the catalogue:
- Profit margin tier
- Bestseller rank
- Seasonality
- New arrival
- Subscription-eligible
- Customer segment
- Inventory priority
Then we filled all 18 recommended attributes in Merchant Center. That included fields many brands leave half-finished, such as product type depth, GTIN coverage, shipping details, condition, availability accuracy, image consistency, and variant-level identifiers. Bear in mind, a feed isn’t just a product list. It’s a decision engine for Google’s auction system.
Actionable takeaway: If you can’t segment products by margin, inventory pressure, and merchandising priority inside the feed, you’re forcing campaigns to optimize in the dark.
The data
What happened once the rebuild was live?
Before the rebuild:
- 34% feed completeness score
- $12k/month Shopping revenue
- 1.8x ROAS
After 6 weeks:
- 94% feed completeness score
- $29k/month Shopping revenue
- 3.4x ROAS
That’s a 140% increase in Shopping revenue in just six weeks, alongside a major improvement in efficiency. We didn’t get there by increasing budget blindly. We got there by improving the asset feeding the machine-learning system.
The single most impactful change
If we had to isolate one lever, which one mattered most? The answer was the profit margin custom label.
By telling Google which products carried 55%+ margins, we gave campaign structure a commercial signal most brands never pass through. Once those products were grouped properly, the algorithm naturally favored more profitable inventory. Over the six-week period, that improved blended margins by 12%.
Let’s face it, ROAS alone can hide bad economics. Two products can both hit a 3.0x return, but if one has a far stronger contribution margin, that’s the one we want more exposure on.
Actionable takeaway: Build a custom label for gross margin bands this week. Even a simple structure like 0–30%, 31–54%, and 55%+ gives you far more control over bidding and reporting.
The takeaway
What should we take from this? Your feed is your most valuable ad asset.
Most supplement brands spend weeks debating creatives, campaign types, and budget splits while the product data underneath everything stays messy. In our experience, 50%+ of a feed’s potential is often left untapped because nobody has turned it into a strategic asset.
If your Shopping performance feels stuck, start with the feed before you touch bids. Clean titles. Complete attributes. Commercial labels. Better data in usually means better results out.
What’s your current feed completeness score, and do you know which products deserve more budget based on margin, not just revenue? If you want, we can compare notes on how your catalogue is structured today.
