Shopify's built-in fields cover a product's title, description, price and images. Everything a real catalogue also needs — fabric composition, care instructions, a size chart, the certification a product carries, the designer behind it — has to live somewhere else. That somewhere is metafields and metaobjects, and choosing the wrong one is the single most common reason a store's data model has to be rebuilt eighteen months in.
The distinction is not complicated once you see it, but Shopify's own naming does not make it obvious.
What is the difference between a metafield and a metaobject?
A metafield is a single custom field attached to one existing record. "This product's fabric is 80% merino." It belongs to that product and nothing else. Shopify defines the field once (a metafield definition) and then every product can carry its own value.
A metaobject is a custom record type of your own. It is not attached to anything by default — it exists independently, with several fields of its own. A store location, a designer profile, a size chart, a shipping-policy block: each is a small structured record you want to define once and reuse.
The two connect through a reference. A metafield on a product can point at a metaobject entry. That is how a product says "my size chart is this one" without the chart being duplicated on every product that shares it.
The one question that decides it
Ask whether the value is ever shared. If the answer is no — this fabric composition belongs to this product and no other — use a metafield. If the answer is yes — twelve products share the same size chart, and when it changes it should change everywhere — use a metaobject and reference it.
This is the rule most stores get wrong in the same direction. They start with metafields because they are simpler to set up, then copy the same size chart text into ninety products. The day the chart changes, someone has to edit ninety records, and in practice they edit sixty and miss thirty.
Where the rebuilds come from
Three patterns cause the expensive kind of rework.
Using plain text where a typed field belongs. A metafield defined as single-line text will happily accept "80% merino", "80 percent merino wool" and "Merino (80%)" for the same thing. Once a few thousand products carry inconsistent free text, you cannot filter or template against it. Shopify supports typed definitions — number, dimension, list, product reference, file — and the type is what makes the data usable later.
Storing content in the theme rather than in data. A shipping-information block written straight into a Liquid section looks fine until the store needs it in three languages and on two templates. Metaobject definitions support translation and storefront publishing as first-class settings; theme markup does not.
Not planning for the storefront. A metafield or metaobject is not visible to the storefront unless the definition exposes it. Teams routinely build a full data model in the admin, then discover at template time that nothing is readable from Liquid or the Storefront API. The visibility setting belongs in the initial definition, not in an emergency migration later.
A model that holds up
For a mid-sized catalogue the shape that tends to survive looks like this. Product-specific facts — composition, weight, country of origin, care code — are typed metafields on the product. Anything editorial and reusable — size charts, care guides, designer bios, certification explainers — are metaobject definitions, referenced from products by a metafield of type metaobject reference.
Collections get their own metafields for the things a category page needs and a product does not: an intro paragraph written for search, an FAQ block, a hero image sized for the template.
The test is whether a merchandiser can add a new product without touching the theme, and whether changing one shared fact changes it everywhere. If either answer is no, the model is not finished.
Metafields and SEO
Structured product data earns its keep twice. It drives the on-page detail that shoppers compare — and it feeds the structured data that search engines and answer engines read. A product schema built from typed metafields stays accurate as the catalogue changes; one hand-written into the theme drifts within a season.
The same applies to collection metafields. Category pages carry most of a store's search demand, and an intro written into a collection metafield can be edited by the merchandising team without a developer in the loop.