I estimated 33%. The real number? Also 33%. But I got everything else wrong.

Executive Summary

• Brand: Apple-ecosystem tech accessories — USB-C hubs, docking stations, keyboards, chargers, adapters. $49.99-$399.99. Apple Store placement. San Diego, founded 2005
• AI visibility score: 50/150 tests surfaced the brand (33%)
• The pattern: Strong on Gemini (62%) and Copilot (38%) but ChatGPT is a complete blank (0/50). Spec-rich descriptions drive Gemini visibility; marketing-first copy does not
• Key competitor gap: Anker owns chargers; CalDigit owns Thunderbolt docks; Logitech owns keyboards editorially
• Root cause: Descriptions are marketing-first, spec-second. Tags only cover device compatibility — zero feature, spec, or use-case tags. Trustpilot 2.5/5 (33 reviews) is a liability
• Fix complexity: Low-Medium — the EX1 keyboard proves they can write spec-rich copy. That approach needs to be the standard, not the exception

The brand

Satechi is a consumer electronics accessories brand based in San Diego, founded in 2005. The brand builds its identity around the Apple ecosystem — designing USB-C hubs, keyboards, chargers, and docking stations that match Apple's aluminium aesthetic. Products are stocked at select Apple Store locations, a meaningful credibility signal.

Satechi sits between cheap Amazon-brand hubs and enterprise-grade docking solutions (CalDigit, OWC). They compete with Anker on price-performance and with Logitech on keyboards, while positioning as more Apple-native than either. Strong Amazon and DTC presence.

The test

We ran 150 automated browser-based tests using Playwright — 10 repeats × 5 queries × 3 platforms (ChatGPT, Gemini, Copilot). Queries targeted Satechi's positioning: USB-C hub for MacBook, slim wireless keyboard for Mac, Thunderbolt dock under $400, compact USB-C charger for travel, and best desk accessories for Mac setup.

The results

ChatGPT 0% is the headline finding. Not a single surfacing across 50 runs. The old directional estimate was 50%. Complete miscalculation. ChatGPT has no awareness of Satechi in any query category tested.

Gemini rewards spec-rich descriptions — the data proves it. The EX1 keyboard (112 words with actual specs: Bluetooth count, battery life, EU replaceability standard) scored 9/10 on Gemini. The OntheGo charger (56 words of lifestyle copy) scored 1/10 on Gemini. Same brand, same structured data quality, same platform. The only difference is description density.

Copilot 38% is one of the best Copilot scores in the audit set. USB-C hub queries hit 8/10 on Copilot — near-universal surfacing. Copilot's product carousel format favours Satechi's product range.

The most expensive product has the worst data and the worst results. The CubeDock at $399.99: 72 words, 5 tags, no aggregateRating. Thunderbolt dock queries: 5/30. CalDigit's spec-dense pages dominate.

Charger queries are an Anker monopoly. 3/30 across all platforms combined. With 56 words of marketing copy and no dimensions, weight, or per-port wattage, the OntheGo charger cannot compete.

Why this is happening

Descriptions are marketing-first, spec-second — and in tech, that is backwards. "Power up anywhere with the OntheGo 67W Slim Wall Charger — a fast, ultra-slim charging solution built for your on-the-move lifestyle." An AI agent needs: what is the per-port wattage split? What are the dimensions? What protocols are supported (PPS, PD 3.0)? The keyboard shows they can write specs — 112 words including Bluetooth count, battery life, EU 2027 replaceability standard. That approach is the exception, not the standard.

Tags only cover device compatibility. Tags like "iPad Pro", "MacBook Air", "iPhone 16" are genuinely useful when someone asks "what hub works with my MacBook Air?" But there are zero tags for port types, wattage, connectivity protocol, form factor, or any feature attribute. The charger with 18 tags has 11 device-compatibility tags, some campaign tags, and zero spec tags.

Trustpilot 2.5/5 with 33 reviews is an active liability. On-site ratings are 4.3-4.6/5. Trustpilot is 2.5/5 with only 33 reviews — almost certainly unclaimed. The ChatGPT 0% may partly reflect negative external signals like this.

The CubeDock data gap is widest where the stakes are highest. A $399.99 Thunderbolt 5 docking station with 72 words and no aggregateRating, competing against CalDigit's spec-dense product pages. The data investment is inversely correlated with price.

What Satechi could do, in priority order

Phase 1 (quick wins):

• Claim and address the Trustpilot profile — 2.5/5 with 33 reviews is an active liability. Drive satisfied customers (4.3-4.6/5 on-site) to leave Trustpilot reviews
• Rewrite the CubeDock description — at $399.99, it needs full port list with speeds, display configurations, NVMe SSD specs, thermal design, dimensions, weight, and direct CalDigit comparison

Phase 2 (medium effort):

• Add spec-first descriptions to all products — use the EX1 keyboard as the template. Every product should lead with specifications (wattage, port counts, dimensions, weight, protocols) and follow with lifestyle positioning
• Expand tags beyond device compatibility — port_type (USB-A, USB-C, HDMI, Thunderbolt), wattage (67W, 100W), technology (GaN, PD, PPS), form_factor (desktop, portable, wall), use_case (travel, desk_setup)
• Add aggregateRating to the CubeDock structured data

Phase 3 (longer term):

• Create comparison content — Satechi CubeDock vs CalDigit TS4, OntheGo 67W vs Anker 65W, EX1 vs Logitech MX Keys
• Build Thunderbolt 5 authority content — the category is still forming. Be the editorial source before CalDigit and OWC claim those positions
• Add additionalProperty fields to JSON-LD — structure spec data in schema for maximum AI agent coverage

Close

Satechi is the audit's estimation paradox. The total visibility — 33% — matched the directional estimate exactly. But the per-platform breakdown was completely inverted. ChatGPT was estimated at 50% and scored 0%. Copilot was estimated at 30% and scored 38%. The aggregate number was right. The story was entirely wrong. And within that 33%, the data tells a clear cause-and-effect story: the EX1 keyboard with 112 words of specs scores 9/10 on Gemini. The charger with 56 words of lifestyle copy scores 1/10. Same brand, same platform, same structured data. The difference is description density. In tech accessories, where buyers ask about wattage, port counts, and dimensions, marketing copy is the wrong format. Satechi's own keyboard proves they know how to write specs. They just have not applied it everywhere yet.