The ag & turf attributes buyers filter on — and most catalogs miss
The ag & turf attributes buyers actually filter on, why missing specs like bore diameter drop SKUs from search, and how to structure them for humans and AI.
A buyer looking for a mower spindle doesn't search "spindle assembly." They search by bearing bore, shaft length, and bolt pattern, because that's what determines whether the part fits. If your catalog only has a title and a marketing blurb, that buyer never finds the SKU — not because it's out of stock, but because the facet never fires.
Ag & turf buyers don't shop on descriptions, they shop on tolerances
Most ag and turf categories are interchange-driven. A distributor selling spindles, PTO shafts, hydraulic cylinders, or blades is competing against OEM part numbers and a dozen aftermarket cross-references, and the buyer already knows the spec they need before they land on the page. Faceted navigation only works when every one of those specs exists as a structured value on the product record — filters populate from the attributes that are actually present in the data, and platforms explicitly hide a facet entirely when no products in the result set carry that attribute value, per BigCommerce's own product filtering documentation. A blank field doesn't just weaken the listing. It removes the SKU from the filter path altogether.
The attributes that actually drive fit-and-function in ag & turf
The specific attribute set varies by part family, but the pattern repeats across the category:
| Part family | Attributes buyers filter on |
|---|---|
| Spindle assemblies | Shaft diameter, shaft length (pulley face to blade mount), bearing type/size (for example 6205-2RS), bolt pattern (3, 4, or 6-bolt), blade mount type (star, keyed, round), housing material, deck size compatibility |
| PTO / driveline shafts | Series size (1000, 1 3/8-6 spline), yoke type, tube telescoping length, shear bolt vs. slip clutch, shield type |
| Hydraulic cylinders | Bore diameter, rod diameter, stroke length, mounting style, port thread size |
| Blades | Length, center hole diameter, thickness, lift type (high-lift, mulching, gator), bolt hole spacing |
None of that is exotic information — it's exactly what a mechanic already has written on the old part before they call to reorder it. The interchange logic aftermarket suppliers publish for mower spindles makes this explicit: shaft length alone (roughly 6.5"–6.9" on a 42-inch deck vs. 8" or more on a 54-inch deck) is often the deciding factor between two parts that look identical in a photo, and bearing spec (sealed ball bearings such as 6205-2RS vs. greasable open bearings or taper rollers on older models) is the other. If either value is missing, the part is functionally unsearchable to anyone who isn't already staring at the box.
Before and after: a mower spindle assembly
Here's what a typical supplier feed looks like next to what a buyer — human or AI — actually needs to act on it.
Raw feed description:
"Heavy duty spindle assembly, fits most 42-54 inch zero turn mowers, replacement part, high quality construction, easy install."
Enriched attribute table:
| Attribute | Value |
|---|---|
| Part type | Mower deck spindle assembly |
| OEM cross-reference | 918-04822B, 187292, AM125172 |
| Shaft diameter | 0.75 in |
| Shaft length (pulley face to blade mount) | 6.9 in |
| Bearing type | Sealed ball bearing, 6205-2RS equivalent, double-sealed |
| Bolt pattern | 4-bolt, 3.62 in spacing |
| Blade mount | 5-point star |
| Housing material | Cast aluminum |
| Pulley diameter | 5.5 in, single groove |
| Deck compatibility | 42 in–48 in decks |
| Fits (brand cross) | Craftsman, Cub Cadet, MTD, Troy-Bilt |
The raw version reads fine to a human skimming a page. It fails completely as data — there's no bore, no bearing spec, no bolt pattern, so it can't be indexed against any of the facets a buyer actually filters on, and it can't be matched against a competitor's cross-reference number either.
Ask an answer engine
This is also where the AI-search gap shows up. Buyers are increasingly typing full-spec questions into ChatGPT or Perplexity rather than a search box — something like "spindle assembly with 0.75 inch shaft and 6205 sealed bearing for a 48 inch zero turn deck." Google has confirmed that structured product data with required properties like name, sku, and offers is what makes a listing eligible to be surfaced and cited at all, and the same principle holds for AI answer engines: they can only recommend a SKU if the attributes that answer the question exist as machine-readable values, not as a sentence buried in a paragraph. A part with a real bore diameter and bearing spec in its data gets matched. A part with only "fits most mowers" doesn't — it's invisible to the exact query it was built to answer.
Why this keeps happening
It's not that distributors don't know these specs. It's that the specs live in a PDF spec sheet, an old ERP field, or a supplier's flat file — never mapped into the fielded, filterable attributes the storefront or the AI crawler actually reads. Fixing it one SKU at a time doesn't scale across a catalog with thousands of spindles, shafts, cylinders, and blades, each with a slightly different attribute set.
That gap — supplier documents full of the right values, catalogs missing the structured fields to use them — is exactly the layer Anglera works in. It plugs into whatever PIM or flat file a distributor already runs, extracts values like shaft length and bearing spec straight from source documents, and quality-scores each one so gaps get flagged instead of shipped blank. The PIM still stores the record. Anglera just makes sure the attribute that decides whether the part fits is actually there to be filtered on.