Snapshot: your product feed is the single biggest lever in Google Shopping, because it decides which searches your products are eligible to appear for. Optimise the titles, attributes, and structured data — and a product that was invisible starts showing up for the queries it should have been winning all along. Done well across a full catalogue, feed optimisation drives measurable lifts in click-through rate, conversion rate, and revenue.
This is the commercial heart of Shopping strategy, and it’s where a lot of South African stores leave money on the table — because their feed is running on whatever their ecommerce platform exported by default. This guide is part of our complete Google Shopping guide for South Africa.
Why the feed decides everything
Google Shopping doesn’t work like text search, where you bid on keywords. There are no keywords. Instead, Google reads your product feed — your titles, descriptions, attributes, and identifiers — and decides which searches each product is relevant to. Your feed is your keyword strategy, whether you’ve treated it that way or not.
This means a product can have real demand and still be invisible, simply because its feed data is weak. If your title reads like an internal SKU name, if attributes are missing, if there’s no GTIN, if the description was written for someone browsing your site rather than for how people actually search — Google can’t match it well, and it quietly loses auctions it should have won.
That’s why a meaningful share of “zombie” products (the ones that get no impressions or no clicks) aren’t demand problems at all — they’re discoverability problems. And discoverability problems are fixable. For the full picture of which products are worth fixing versus which need a different approach, see what is a zombie SKU in Google Shopping.
The feed fields that matter most
Not all feed fields carry equal weight. In rough order of impact:
Product title. By far the most important field — it’s what Google leans on hardest to decide which searches you’re eligible for. A strong title leads with the terms people actually search, in the right order, and includes the attributes that matter for that product type (brand, product type, key specs, size, colour). We cover this in depth in how to write Google Shopping titles that rank.
Product type and Google product category. These help Google understand where your product sits and which queries it belongs in. Vague or missing categorisation means weaker matching.
GTIN and identifiers. Missing or wrong GTINs suppress eligibility and can stop products serving entirely. This is a common cause of zero-impression products — see fixing Google Merchant Center disapprovals.
Description. Less critical for traditional Shopping matching than the title, but increasingly important for AI shopping surfaces, which read and quote descriptive, keyword-rich copy. More on that in optimising your feed for AI search and ChatGPT Commerce.
Images. Poor or non-compliant images kill click-through even when a product ranks. Clean, correctly formatted images are table stakes.
Attributes (colour, size, material, etc.). These both improve matching and prevent the variant-duplication problems that create false zombies — see managing variants so they don’t wreck your Shopping data.
The real problem: scale
Here’s the truth about feed optimisation: everyone knows the title matters. The reason most feeds are still bad is that hand-writing optimised titles, descriptions, and attributes for thousands of SKUs is tedious, slow, and never finished. It’s exactly the kind of work that quietly never gets done — so the feed runs on platform defaults, and the store keeps paying for poorly matched clicks.
A store with a few hundred products can optimise by hand. A store with 5,000, 20,000, or 50,000 products cannot. And that’s precisely the scale at which the zombie problem is worst.
Doing it at scale with AI
This scale problem is why we built ShopRank, our in-house AI-powered feed optimisation and ranking platform. It’s designed to solve the feed-quality half of the zombie problem across a full catalogue:
- AI-generated titles, descriptions, and highlights optimised for both traditional Shopping search and the newer AI surfaces like Google AI Mode and ChatGPT Commerce.
- Real-time ranking tracking so you can see exactly where each product ranks by search term, location, and device — instead of optimising blind.
- Direct Shopify and WooCommerce integration — connect your store via API and sync products in one click, no CSV wrangling.
- ML performance forecasting that predicts impressions, click-through rate, and clicks before you optimise, so you prioritise the products with the most upside.
- Bulk optimisation across 50,000+ products with auto-updating feed URLs that push changes automatically.
The evidence
Feed optimisation isn’t a theory — the impact is measurable. In one 10-week split test we ran, comparing roughly 2,400 optimised products against about 470 non-optimised products from the same retailer:
- Click-through rate rose dramatically on the optimised group versus the unoptimised control.
- Conversion rate lifted meaningfully on optimised product pages.
- Revenue per click improved because the feed was matching higher-quality, more relevant queries rather than broad, low-intent ones.
There was also a clear organic halo: optimised products grew their organic search visibility by +17% while non-optimised products fell 24% over the same period — a 41-point swing — and only the optimised products converted from AI/LLM traffic. Better titles don’t just help paid Shopping; they make products more discoverable everywhere, including in AI answers.
Structured feed optimisation of this kind drives up to a 30% improvement in conversion rate — because you stop paying for clicks on poorly matched queries and finally appear for the searches that convert.
Feed optimisation in the South African context
A few things specific to running Shopping feeds in South Africa:
Everything is ZAR-denominated and price-sensitive. South African shoppers are highly price-aware, so accurate pricing, sale-price annotations, and competitive positioning in the feed matter more than in some markets.
Local search language. Optimise titles for how South Africans actually search — local terminology, brand spellings, and product names — rather than assuming international phrasing.
Multi-channel reach. Google Shopping isn’t the only surface. A well-structured feed can be adapted for other channels South African shoppers use, and increasingly for AI shopping surfaces. Building the feed properly once pays off across all of them.
Comparison Shopping Service (CSS) partners are not a factor here. Unlike the EU, South Africa isn’t a CSS market, so don’t get drawn into CSS-based feed tactics that don’t apply locally.
Crucially, feed optimisation works alongside good campaign structure. Optimise the feed to rescue the products that are fixable; separate genuine low-demand products into their own discovery campaign so they get room to perform; and let the data tell you which is which.
Frequently asked questions
What is product feed optimisation? It’s the practice of improving the titles, attributes, descriptions, and structured data in your Google Shopping feed so that Google matches your products to more relevant, higher-converting searches. Because Shopping has no keywords, the feed is effectively your keyword strategy.
How much does feed optimisation improve results? It varies by how weak the starting feed is, but structured optimisation commonly drives up to a 30% improvement in conversion rate, along with higher click-through rate and revenue per click. There’s often an organic halo too — better titles improve free organic and AI-search visibility.
Which feed field is most important? The product title. It’s what Google relies on most to decide which searches your product can appear for. Getting the title right — leading with searched terms, in the right order, with key attributes — is the highest-leverage single change you can make.
Can I optimise my feed manually? For a few hundred products, yes. For thousands, it’s impractical to do well by hand, which is why AI-powered tools like ShopRank exist — to optimise titles and attributes across large catalogues and keep them updated automatically.
Does feed optimisation help with AI search? Yes. Descriptive, keyword-rich feed data is more “quotable” by AI shopping surfaces like ChatGPT Commerce and Google AI Mode. In our testing, only optimised products converted from AI traffic. See optimising your feed for AI search and ChatGPT Commerce.
Next in this cluster: how to write Google Shopping titles that rank and optimising your feed for AI search and ChatGPT Commerce. Or return to the complete Google Shopping guide for South Africa.
Ready to optimise your feed at scale? See what ShopRank can do.
