The channel’s biggest brands. AI can’t see them.
Every week we compare two realities: which brands the major electronics distributors actually carry, per region — and which brands the AI models of that same region can name as components companies. The gap is measured, alias-screened, and domain-verified before anything appears below — 20 manufacturers here are confirmed canonical and genuinely invisible to AI. (A broader re-screen of the long tail is in progress; we publish only what we’ve verified.)
Can AI see you?
Can AI see you?
Carried by 17 distributor catalogs with 52 category placements. AI mentions as a components company, any region, all history: 0. Yet ask AI who makes its categories and 16 of 17 models name it — the story is retrievable, but not attached to the brand.
Ask AI who makes EMI suppression — it names them. Ask AI about the brand — silence. The knowledge is in the model; the brand identity isn’t attached to it. That gap is recoverable, which is exactly why it’s worth knowing about.
Invisible whether you ask by category or by name. The distributor shelf says these are significant manufacturers; the models simply don’t carry them. 4 are carried by Asian distributors and invisible even to the five Chinese AI labs — home-market blindness.
The blind list
| Brand | US shelf | US AI | EU shelf | EU AI | Asia shelf | Asia AI | Placements | Direction |
|---|---|---|---|---|---|---|---|---|
| 3M3m.com | 88% | 0 | 50% | 0 | 80% | 0 | 52 | one-way+ home-blind |
| Schneider Electricse.com | 75% | 0 | 67% | 0 | 60% | 0 | 45 | one-way+ home-blind |
| Honeywellhoneywell.com | 88% | 0 | 33% | 0 | 60% | 0 | 37 | both ways+ home-blind |
| Abraconabracon.com | 75% | 0 | 50% | 0 | 20% | 0 | 20 | one-way |
| Panduitpanduit.com | 88% | 0 | 0% | 0 | 0% | 0 | 15 | one-way |
| Weidmüllerweidmueller.com | 63% | 0 | 50% | 0 | 20% | 0 | 14 | both ways |
| Ohmite Manufacturingohmite.com | 75% | 0 | 17% | 0 | 0% | 0 | 13 | one-way |
| DFRobotdfrobot.com | 25% | 0 | 0% | 0 | 40% | 0 | 12 | one-way+ home-blind |
| Tektronixtek.com | 50% | 0 | 50% | 0 | 20% | 0 | 12 | both ways |
| EBM-Papstebmpapst.com | 75% | 0 | 50% | 0 | 0% | 0 | 11 | one-way |
| Siemenssiemens.com | 13% | 0 | 33% | 0 | 0% | 0 | 11 | one-way |
| HellermannTytonhellermanntyton.com | 50% | 0 | 33% | 0 | 0% | 0 | 10 | both ways |
| pomonaelectronics.com | 63% | 0 | 17% | 0 | 0% | 0 | 10 | both ways |
| XP Powerxppower.com | 50% | 0 | 50% | 0 | 20% | 0 | 10 | one-way |
| IDECidec.com | 38% | 0 | 0% | 0 | 20% | 0 | 9 | both ways |
| Schurterschurter.com | 25% | 0 | 33% | 0 | 20% | 0 | 9 | one-way |
| Vicor Corporationvicor.com | 50% | 0 | 17% | 0 | 20% | 0 | 9 | both ways |
| Bivarbivar.com | 38% | 0 | 0% | 0 | 20% | 0 | 8 | one-way |
| Rohde & Schwarzrohde-schwarz.com | 25% | 0 | 50% | 0 | 0% | 0 | 8 | one-way |
| NIC Componentsniccomp.com | 13% | 0 | 17% | 0 | 0% | 0 | 7 | one-way |
This is not a quality judgment on any brand. It measures one thing: whether the AI models buyers increasingly consult can connect a brand’s name to the components the channel says it supplies. Suspected alias splits are screened out before any blindness claim is made; brands we don’t yet collect first-party data for are bucketed separately and never presented as “blind.”