Skip to content

Improve catalog campaigns: what to check before raising budget

A practical diagnosis tree to separate measurement, inventory, feed, creative and offer issues before you spend more.

J

Julia Villegas

· 9 min read · 275 views
Products and paper paths to improve catalog campaigns

Improve catalog campaigns by diagnosing what is failing before you touch the budget. If your dynamic ads on Meta, Google, TikTok or other channels are generating fewer sales, the issue may sit in measurement, inventory, the product feed, creative or the offer itself. This guide helps you review each layer without assuming that spending more will solve it.

What it means to diagnose a catalog campaign

A catalog campaign uses a product feed to show ads for specific items: image, title, price, availability, brand, category and other product attributes. On Meta, you may know them as catalog ads or dynamic ads. On Google, the same catalog logic can support Shopping or Performance Max. Other channels have their own feed formats and requirements.

Diagnosing a catalog campaign does not mean looking only at ROAS and reacting. It means separating signals. A campaign can look worse because measurement changed, because many products are out of stock, because the feed is sending outdated prices, because the creative no longer stands out, or because the products receiving spend have weak margin.

For beginners, it helps to define five terms clearly:

  • Attribution: how an ad platform assigns a sale to an ad. It may not match your real margin or your internal analytics exactly.
  • Inventory: what products are available, at what price, and under which shipping or fulfillment conditions.
  • Product feed: the file or connection that sends catalog data to each advertising channel.
  • Dynamic creative: an image or video built from catalog fields, such as product photo, price, discount or brand.
  • Offer: the full commercial proposition: price, margin, promotion, shipping, availability and product appeal.

Before increasing budget, your goal is to answer a simple question: is this a demand problem, a data problem, a presentation problem or a business problem?

Improve catalog campaigns with a diagnosis tree

Use this diagnosis tree in order. If you start with creative before checking measurement or stock, you may redesign ads for products that are not ready to sell. If you start with bids and budget before reviewing margin, you may scale revenue that is not actually profitable.

Layer Basic question Warning sign First action
Measurement Are the numbers reliable? ROAS varies heavily between platforms Compare attribution, orders and margin
Inventory Can the advertised products be bought? Clicks go to out-of-stock or limited products Exclude or segment unavailable products
Data Is the feed complete and current? Rejections, inconsistent prices, empty categories Review required fields and rules
Creative Does the ad explain the value clearly? Low CTR or fatigue on the same products Test formats, messages and visible fields
Offer Is the product worth scaling? Sales with low profit or high return risk Segment by margin, price and priority

The order matters. First, make sure you are reading the campaign correctly. Then check whether you are advertising what you can actually sell. Next, verify that each channel receives accurate product data. Only after that does it make sense to adjust creatives, product sets or budget.

1. Measurement: do not confuse attribution with profitability

When a campaign drops, it is tempting to open the ad platform, look at ROAS and make a quick decision. But platform ROAS is an attributed number. It does not necessarily include margin, returns, shipping costs, discounts or differences between attribution models.

Start with three checks:

  1. Compare equivalent periods. Do not compare a promotional week with a normal week, or days with very different spend levels.
  2. Cross-check platform and store data. Look at orders, revenue and sales in your ecommerce platform, not only attributed conversions.
  3. Go down to product level. A campaign can look stable overall while spend shifts toward less profitable products.

This is where product-level analysis changes the conversation. AdMotion Meta Analytics is designed to cross Meta Ads metrics with catalog data so you can analyze performance product by product and export reports. It is available on request and uses read-only permissions: it helps you diagnose, but it does not pause campaigns or change bids and budgets for you.

That limitation is important. A report can show which products concentrate spend, clicks or attributed sales, but your final decision should also consider real margin, stock, returns and commercial priorities.

2. Inventory: do not push products you cannot sell well

Inventory is often the hidden layer behind poor catalog performance. If a hero product loses key sizes, if a popular variant is unavailable, or if delivery times become less competitive, the campaign can keep getting clicks while conversion rate drops.

Review these cases:

  • Products active in campaigns but with insufficient stock.
  • Important variants sold out, even if the parent product still appears available.
  • Items with the correct price in the store but an outdated price in the feed.
  • Products restricted by country, shipping method or season.
  • Categories with strong traffic but weak real availability.

Not every out-of-stock product needs the same treatment. You may exclude it, move it into a separate product set, or keep it if the channel and your strategy allow it. The key is not mixing unavailable products with ready-to-sell items without control.

The AdMotion Feed Manager connects your store or any product file, maps fields, normalizes data with rules and lets you create filtered sets. Each feed is published through a URL generated on every download. Still, the frequency at which your catalog is imported and the frequency at which each ad platform reads the feed are different things; you should not assume instant synchronization inside the advertising channel.

3. Product feed data: a messy catalog limits your options

A product feed is not just a product list. It is the base that helps the channel understand what you sell, who may be interested, and how products can be shown. If the data is missing or poorly structured, your optimization options shrink.

Review these core fields:

  • Product ID: stable and unique, so reporting history is not broken.
  • Title: clear, searchable and not overloaded with irrelevant details.
  • Price and sale price: consistent with your ecommerce store.
  • Availability: current and consistent across variants.
  • Category: useful for segmentation, not just compliance.
  • Brand: especially important for multi-brand retail.
  • Main image: clean, readable and suitable for the channel.
  • Custom attributes: margin, season, bestseller, clearance or commercial priority if you use them internally.

Example: a sports retailer sees its catalog campaigns decline in ROAS. After reviewing the feed, the team finds that the running shoes with the strongest stock do not have a “discipline” attribute, while outlet products are fully tagged. The platform can still show products, but the marketing team cannot clearly compare running, trail and fitness items or build useful product sets.

The solution is not simply “spend more on running.” First, the retailer needs to fix the data: map or complete the discipline field, separate outlet from current-season products, review stock and create comparable sets.

4. Creative: do not decorate a product problem

Creative matters, but it should not be used to hide an offer problem. A template can make a product easier to understand, highlight price, organize your brand system or adapt formats for feeds, Stories, Reels or display placements. It cannot turn a product with no demand, no margin or no stock into a strong business opportunity.

Review creative with practical questions:

  • Is the product image easy to read on mobile?
  • Is the price or discount visible when it matters?
  • Does the message change by category, season or promotion?
  • Does the format fit the channel: 1:1, 4:5, 9:16 or 16:9?
  • Does the creative respect the policies and specifications of each platform?

The AdMotion Templates Editor lets you design a template using catalog fields such as product photo, title, price or brand, then generate one image per product. It supports visibility rules and 1:1, 4:5, 9:16 and 16:9 formats. When price or stock changes in the catalog, the creative can reflect those dynamic fields when it is regenerated from the available data.

For products with inconsistent backgrounds, background removal can help standardize presentation, with human review. And if you work across multiple channels, do not assume one image works everywhere. Cropping, hierarchy and visible fields may need to change by channel and placement.

5. Offer: separate sales from profit

A campaign can sell more and still be worse for the business. That is why the diagnosis should end with the offer. At this point, you are not looking only at clicks or purchases; you are deciding which products deserve more visibility.

Create simple product groups:

  • Products with strong margin and strong stock.
  • Products with strong margin but low visibility.
  • Products with high spend and weak conversion.
  • Products that sell but leave little profit.
  • Seasonal products that need urgency or delivery cut-off messaging.

You do not need complex segmentation on day one. For a mid-sized store, three or four well-maintained sets can be more useful than twenty rules nobody trusts. For example: “high priority,” “clearance,” “new collection” and “temporarily exclude.”

How to run this workflow with AdMotion

If you want to diagnose with catalog data and prepare changes without editing products one by one, the platform can support the workflow. The point is not to replace your commercial judgment. It is to organize data, create sets and review creatives in a more controlled way.

  1. Connect your catalog. Start from Shopify, BigCommerce, Tiendanube or any product file or feed that fits your operation.
  2. Map the important fields. Connect ID, title, price, availability, image, brand, category and any internal fields you use for margin, season or priority.
  3. Normalize with rules. Clean inconsistent values, complete categories where appropriate and create rules or sets. Rules can also be written in natural language and reviewed before they are applied.
  4. Cross performance and catalog data. With Meta Analytics, available on request, you can analyze Meta metrics product by product and export reports to review spend, attributed sales and behavior by group.
  5. Create actionable sets. Separate products to push, products to monitor and products to exclude or review. Keep the naming simple.
  6. Prepare the channel feed. With Export Channels, you can publish feeds for Meta, Google, TikTok, Pinterest, Snapchat, Criteo, Microsoft Advertising, Thribee, ChatGPT or a custom channel, and review each feed against the channel specification.
  7. Design creatives by set. Use templates with dynamic fields to show price, brand, category or promotion when it helps the user decide.
  8. Copy the feed URL into the channel. Remember that the feed URL is generated on each download, but every platform has its own reading frequency.

This turns “the campaign is getting worse” into more useful hypotheses: “spend is concentrated on low-stock products,” “the feed does not distinguish margin,” “outlet creatives do not show the discount,” or “the problem is one category, not the full catalog.”

If your priority is to understand which products are helping or holding back Meta campaigns, you can review Meta Analytics and request access when you need product-level diagnosis.

What to do this week

  1. Pick one campaign and a clear period. Do not audit your entire history at once. Start with one important catalog campaign and compare it with an equivalent period.
  2. Export a product performance list. Include spend, clicks, attributed conversions, attributed revenue and, if possible, margin or internal priority.
  3. Add inventory status. Mark products with no stock, low stock, sold-out variants or shipping restrictions.
  4. Review ten extreme products. Look at the five products with the highest spend and weakest result, and the five with the strongest result. Check feed data, product page, price, image and availability.
  5. Create three decisions. For example: temporarily exclude, improve feed data or test a specific creative. Avoid changing budget, feed and template at the same time if you want to understand what changed performance.

Final check before increasing budget

Before you raise spend, answer these questions in writing:

  • Are conversion numbers comparable with store data and margin?
  • Do the products receiving the most spend have enough stock and reasonable buying conditions?
  • Does the feed contain the fields you need for useful segmentation?
  • Do creatives show correct and helpful information for the buyer?
  • Does the offer you want to scale make sense for the business, not only for the platform report?

If one answer is “I don’t know,” you may not have a budget problem yet. You have a diagnosis problem. Fixing that layer first will help you make better decisions, even if you later decide to invest more.

A large catalog is not optimized by averages alone. It improves product by product, by separating measurement, inventory, data, creative and offer so you can act where the real bottleneck is.

Frequently asked questions

When should I increase budget in a catalog campaign?

Increase budget only after checking that measurement is reliable, key products have stock, the feed is accurate and the offer has enough margin. If the real issue is data, inventory or creative, extra spend may amplify the problem instead of fixing it and make later diagnosis harder.

Can a new template fix a catalog campaign that sells less?

A new template can help when the issue is presentation: unclear image, hidden price, weak format or generic message. It cannot fix a product with no demand, no margin or no stock. Review measurement, inventory, feed data and offer before redesigning every creative.

What is the difference between attributed performance and real margin?

Attributed performance is what the ad platform assigns to your ads through its measurement model. Real margin includes costs, returns, discounts, logistics and product-level profit. For budget decisions, you need to compare both and connect performance data with the catalog, not rely only on platform ROAS.

Sources

  1. Meta Analytics de AdMotion — AdMotion
  2. Feed Manager de AdMotion — AdMotion
  3. Editor de Plantillas de AdMotion — AdMotion
Share:
J

Written by

Julia Villegas

Performance Marketing

Julia Villegas works in performance marketing at AdMotion.

Related posts