If you only have budget for one, put it into collection pages. On almost every Shopify catalogue we audit, the plural commercial query — "cotton kurta sets", "whey protein isolate", "gold plated jhumkas" — is the one with volume and buying intent, and Google almost always wants a listing page for it. Product pages earn their traffic later, on brand-plus-model and long-tail queries, and they do the actual converting. So the honest answer to the Shopify product page vs collection page question is that they rank for different things and the split isn't 50/50. It's closer to 70/30 in favour of collections for organic acquisition, and 100/0 in favour of product pages for revenue per session.
What each page type is built to rank for
Look at the SERP before you decide. Search a plural, category-shaped query in your niche and count the results: you'll mostly see marketplace category pages, brand collection pages and a couple of listicles. Search a singular, specific query — a model number, an exact weight and flavour, "22k gold bangle 8 grams" — and product detail pages start appearing.
That's the whole rule. Queries that imply choosing get listing pages. Queries that imply a decision already made get product pages. Everything else in this post is plumbing.
Where it gets interesting is the middle: "organic cold pressed coconut oil 1 litre". That's a category query with a size filter attached. Sometimes a filtered collection wins, sometimes a single product wins, and which one you should target depends on how many products you actually have that match. Two SKUs? Point it at the better product page. Fourteen SKUs? Build the collection.
The arithmetic, because it settles the argument faster than opinion
Take a store with 800 SKUs and 12 real collections. Assume — and check this against your own Search Console data, not mine — that a typical product page in a catalogue that size pulls 2 organic sessions a month, while a collection page that's genuinely optimised pulls 900.
800 × 2 = 1,600 sessions from product pages. 12 × 900 = 10,800 from collections. The 12 pages beat the 800 by nearly 7×.
Now flip it to effort. Rewriting 800 product descriptions at 20 minutes each is 267 hours. Doing 12 collection pages properly — intent research, copy that sits in the right place in the DOM, filter architecture, internal linking, image weight — is maybe 40 hours. You do not need a spreadsheet to see which one you start with.
The counter-argument is fair: those 1,600 product sessions convert at 3–4× the rate of collection sessions, because someone landing on a PDP from a specific query is further down the funnel. True. Which is why the answer isn't "ignore product pages", it's "stop treating all 800 of them as equal".
Shopify's URL structure makes some of these decisions for you
A few Shopify behaviours change how this plays out, and they trip up people coming from WooCommerce or Magento.
- The same product is reachable at /products/handle and at /collections/some-collection/products/handle. Shopify canonicalises the second to the first. Good. But your theme's internal links may all use the long form, which means your internal link equity flows through URLs that aren't the ones you want indexed. Check what your collection grid actually outputs.
- Variant URLs (?variant=) are common in ad landing pages and shared links. They shouldn't be your SEO targets. If a variant deserves its own ranking page — a distinct flavour, a distinct gram weight, a distinct metal purity — make it a separate product, not a variant.
- Legacy tag-combination collection URLs (the ones with a + in the path) are blocked by Shopify's default robots.txt. Read your own /robots.txt rather than trusting that; stores that have edited it through the theme's robots template sometimes have surprises in there.
- Filter and sort parameters on collection pages usually carry a canonical back to the clean collection URL. Usually. View source on a filtered URL and confirm, because themes get modified and the behaviour differs across theme versions.
None of this is exotic. It's just the sort of thing that costs you six months of wondering why a collection won't rank while eleven parameterised versions of it fight each other.
How to make a Shopify collection page actually rank
Most collection SEO advice stops at "add a description". That's the least valuable part.
Put text above the grid, but keep it to two or three sentences. Google reads it, shoppers ignore it, and a 400-word wall pushes your products below the fold on a 6-inch screen. Long-form supporting copy goes below the grid where it does no conversion damage.
Fix the H1. Half the themes we open have the collection title in an H1 and a hero banner heading also in an H1. Pick one.
Get the internal links right. A collection with 200 products and no filters gets crawled shallowly and converts badly. Filters solve both: they let shoppers cut a large grid to something choosable, and if you expose a handful of high-intent combinations as their own crawlable URLs, you get pages that match queries you'd otherwise never rank for. This is the one place where tooling earns its keep — we usually reach for FilterPro when a catalogue is past a few hundred SKUs and the native filters aren't holding up.
Watch the LCP. Collection pages load 24 to 48 images. They are the slowest templates on most stores and they're the ones your paid traffic lands on. Getting a 3.1s LCP on a collection template down under 2s is often worth more than any copy change; we cover the mechanics in Shopify speed optimization.
Which product pages deserve real work
Sort your catalogue by revenue over the last 12 months. The top 30 to 50 SKUs get proper treatment: original description written by someone who has held the product, specs in a structured table, real photography, review schema, an FAQ block answering the questions your support inbox actually receives, and internal links from the relevant collections and blog posts.
Everything below that gets a clean template, correct product schema, a unique title tag and nothing else. Do not pay a freelancer ₹150 a description to generate 700 pieces of filler about how this saree is perfect for every occasion. It will not rank, and it makes the pages that could rank harder to find.
Two exceptions worth naming. If you sell branded third-party products, your product pages are competing against every other retailer using the manufacturer's copy, and originality there is the only lever you have. And if you're a single-product brand, this entire debate is moot: your product page and your homepage are your SEO strategy.
Where filters create 4,000 thin pages instead of ten good ones
The failure mode is predictable. Someone reads that filtered collections rank, opens up every facet combination to indexing, and ends up with URLs for "blue + size XL + under ₹999 + sorted by price descending" — hundreds of near-empty pages that dilute the parent collection and burn crawl budget.
The line we use: a filtered URL becomes an indexable page only if it has at least 8 to 10 products, a query behind it with real search volume, and a title and H1 that were written by a human. Everything else stays canonicalised to the parent. On a 5,000-SKU store that usually means 15 to 40 curated filter pages, not thousands.
Also worth knowing before you commit: automated collections built on tags break in irritating ways when your tagging is inconsistent across suppliers. We get product taxonomy wrong on the first pass reasonably often, and the fix is always upstream in how products are tagged, not in the collection rules.
Two India-specific wrinkles
Out-of-stock handling matters more here because of how catalogues churn. If a product page has ranked for two years and the SKU is discontinued, do not 404 it and do not bulk-redirect it to the homepage. Redirect to the closest collection. That preserves the ranking signal and lands the shopper somewhere they can buy — which is exactly the argument for collection pages in the first place.
Price queries behave differently in India too. "Price" and "rate" modifiers are common, and for categories where the price moves daily, a collection page with live pricing beats a static product page every time. Jewellery is the sharpest example: shoppers search by metal, purity and weight, land on a category listing and expect today's rate, not last month's.
Then there's festive load. Collection templates carry the traffic spike during a Diwali or Ramadan sale, so if you're going to stress-test one template before a peak, test that one.
A first-pass checklist we run on new accounts
- Export Search Console queries for 12 months. Split into plural/category and singular/specific. Note which page type currently ranks for each.
- Find category queries where a product page is ranking. That's a missing or under-built collection.
- Find collection pages ranking for a single-product query. Usually means the product page is thin or blocked from internal linking.
- Crawl the site and count indexable URLs against SKU count. If indexable pages exceed SKUs by more than roughly 1.5×, you have a parameter problem.
- Check the collection template's LCP on a throttled 4G profile, not on your office wifi.
- Confirm every collection has a unique title, one H1, and a description that isn't the same 40 words with the category name swapped.
Point five catches more revenue than points one through four combined, which says something about where SEO effort tends to go.
Pull your top 50 organic queries and mark each one C or P. If more than half your C queries are landing on product pages, your collection architecture is the constraint, and no amount of description rewriting will move it. If you want a second pair of eyes on the split, our free audit covers template structure and indexation, or read how we approach SEO and content for catalogue stores.