How should I structure “works with” data so AI can answer “will this work with my model?”

Why “works with” questions are hard for AI

Compatibility questions look simple: “will this filter fit my fridge?”, “will this toner work in my printer?”. They're hard to answer from most listings because model names are ambiguous:

• One product family can contain dozens of model numbers, and not all of them take the same part.
• Generations share a name but differ in size, connector or firmware.
• Suffixes on a model number can mean a different color or a different internal part.

An assistant can only confirm compatibility when the listing names the shopper's exact model, or gives specs precise enough to rule it in or out.

Pick the matching key for your product type

How to structure compatibility data, step by step

1. 1

   Start from the supplier's model list, not a summary

   Ask for the full model-level list. A family name like “fits Series 5 printers” is a summary, not compatibility data.

2. 2

   Write one model per entry, exactly as labelled

   Include suffixes and generation markers as they appear on the device's label. Don't merge two models into one entry with a slash unless that's how the manufacturer writes it.

3. 3

   Add the replaced part number

   Shoppers often search by the number on the part they're replacing: the cartridge number, the OEM filter number or the original battery part number.

4. 4

   Add the deciding specs

   Voltage and connector for batteries and chargers, dimensions for size-based filters, yield and color for toner. These let an assistant answer when the shopper doesn't know their exact model.

5. 5

   Name what it doesn't work with

   If a similar-sounding model or a newer generation isn't compatible, say so. That one line prevents a whole class of returns.

6. 6

   Publish it in the page and in structured data

   Show the list as text or a table in the page HTML, and add it to Product JSON-LD: isConsumableFor for consumables like ink, toner and filters, or isAccessoryOrSparePartFor for accessories and spare parts (schema.org). Keep both generated from the same source so they never disagree.

Example: a replacement filter

The same filter, described two ways. The model numbers are invented.

In Product JSON-LD, the same list can look like this (trimmed):

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Refrigerator Water Filter",
  "mpn": "WF-100",
  "isConsumableFor": [
    { "@type": "Product", "model": "RF-2840A" },
    { "@type": "Product", "model": "RF-2840B" },
    { "@type": "Product", "model": "RF-3100A" }
  ]
}

Where Nightshift fits

Nightshift's templates today cover vehicle and machine parts, from powersports to appliance parts. The principles on this page apply to any compatibility-critical product. For the vehicle-parts version, read how to make parts listings AI-ready, and for what incomplete compatibility costs, see how to reduce wrong-fit returns.