A measured view of how ChatGPT, Gemini, Google AI Overviews, and Google AI Mode answer buyer questions in the commercial products category, and the specific gaps costing meridiancommercial.com recommendations right now.
Meridian Commercial holds a defensible position where buyers already know the brand: ProBin, SlimLine, SteriClean, and utility carts all score 70+ AI visibility. The exposure is everywhere else.
Your monthly AI audience has fallen from a ~16M peak in February to 8.4M today. Your weakest engine is Google AI Overviews (38/100), the surface with the largest buyer traffic. And across 535 category topics worth 4.4M monthly searches, AI engines recommend a national foodservice competitor and Competitor B while never mentioning Meridian once.
The findings below are ordered by revenue exposure, and every one maps to a fix in Section 08.
Your visibility is riding your parent company's coattails. Only 1 of your top 25 performing prompts cites a Meridian page as a source. 14 of the 25 are really about the parent company (Apex Consumer Brands), not Meridian's own products. AI knows the portfolio; it doesn't treat your content as the authority.
Google AI Overviews is your weakest engine at 38/100 vs. 65–70 on ChatGPT, AI Mode, and Gemini. AI Overviews sits on top of standard Google search, the highest-traffic surface your buyers use. This is the single largest gap between where you're visible and where buyers actually look.
Your AI audience is in decline: 8.4M/month, down ~48% from the February peak. Mentions are flat while the reachable audience shrinks, meaning share is being absorbed by sources AI engines trust more in your category.
Competitors own 535 topics where you don't exist. a national foodservice competitor is cited 727 times across these gap topics; Competitor B 208; Meridian zero. Restaurant Supply & Equipment alone carries 317K monthly searches at 47% commercial/transactional intent.
Your citations are diluted across 601 pages with no authority hubs. The top 10 cited pages carry only 13% of all citations. AI engines cite pages that answer whole questions; Meridian's depth is spread thin across product pages instead of concentrated in answer-shaped resources.
Your strongholds are real and worth defending. ProBin (84), Baby Changing Stations (81), Mop Buckets (80), SlimLine (76). These prove the brand can win AI recommendations when content depth exists. The playbook that built them hasn't been applied to the category topics above.
Each AI engine builds answers differently. Gemini and AI Mode lean on Google's index and structured data; ChatGPT leans on its crawl plus cited web sources; AI Overviews rewards pages that answer questions directly and rank organically. Your scores show exactly which system is underfed.
Mention share by engine: Gemini 42.9% (1.2K) · ChatGPT 26.2% (716) · AI Mode 22.9% (625) · AI Overviews 8% (219). The engine with the most buyer reach contributes the fewest mentions.
These are your highest-visibility brand topics. They share a pattern: deep, well-structured category pages that AI engines can extract answers from. This is the template the gap topics need.
| Topic | AI Visibility | Monthly Volume |
|---|---|---|
| ProBin Waste Containers | 81–84 | 6,099 |
| Brooms & Commercial Cleaning Tools | 71 | 3,284 |
| SlimLine Recycling Bins | 76 | 1,631 |
| SteriClean Equipment | 71 | 994 |
| Mop Buckets & Accessories | 80 | 355 |
| Baby Changing Stations | 81 | 266 |
Most-cited pages confirm it: the ProBin hub, utility service carts, SteriClean, and brooms category pages carry the citation load. Product-level pages are almost never cited on their own.
Across 535 topics adjacent to your catalog, AI engines answer buyer questions by citing your competitors. These are the highest-value gaps, filtered for commercial intent and product-line fit.
| Gap Topic | Monthly Volume | Buying Intent | Who Gets Cited | Meridian |
|---|---|---|---|---|
| Restaurant Supply & Equipment | 317,086 | 47% | Competitor A | 0 mentions |
| Disinfectant & Cleaning Wipes | 113,421 | 39% | Competitor A | 0 mentions |
| Mold & Mildew Removal Solutions | 325,050 | 31% | Competitor B | 0 mentions |
| Hospital & Disinfectant Wipes | 20,875 | 38% | Competitor B | 0 mentions |
| Food Contact Surface Sanitation | 33,029 | 22% | Competitor A | 0 mentions |
| Utility & Rolling Carts brand topic · vis 16 | 34,952 | — | Mixed | Near zero |
Competitor citation totals across all 535 gap topics: competitor-a.com 727 · competitor-b.com 208 · meridiancommercial.com 0. Buying intent = commercial + transactional share of prompt volume.
The pattern: these are problem-state topics ("how do we sanitize food contact surfaces") rather than product-state topics ("ProBin container"). Buyers start with the problem. Competitors publish answer content for the problem; Meridian publishes catalog pages for the product. AI engines cite whoever answers the question.
When an AI engine cites meridiancommercial.com as a source, it names the Meridian brand in its answer only 28.9% of the time, the lowest conversion rate of any Meridian-owned domain we tracked. The engine trusted the page as a fact source but gave the recommendation anonymously. The fix is direct: explicit brand and product naming in the first two sentences of each cited page, plus complete Organization and Product schema.
Brand named per citation, by owned domain. The flagship US site under-converts its own citations into brand credit; the smaller regional sites outperform it.
Retail and community domains carry most of the citation load in your category, and they rarely name you either. Your presence on these domains is an off-site visibility lever you don't currently manage.
| Cited Domain | Mentions | Brand Named | Signal |
|---|---|---|---|
| homedepot.com | 632 | 18.6% | Low conversion |
| walmart.com | 434 | 17.5% | Low conversion |
| meridiancommercial.com | 427 | 28.9% | Low conversion |
| lowes.com | 360 | 17.3% | Low conversion |
| amazon.com | 355 | 20.6% | Low conversion |
| reddit.com | 322 | 18.8% | Low conversion |
| youtube.com | 305 | 24.4% | Moderate |
These are actual prompts where AI engines named 16 to 22 brands in the answer and Meridian wasn't one of them. Each is on-catalog, English-language, and high buying intent. We filter the raw opportunity list by hand: this run, 9 of 25 machine-suggested topics were cut as off-catalog noise, because chasing AI volume on irrelevant topics wastes content budget.
| Buyer Prompt | Engine | Brands Named | Meridian |
|---|---|---|---|
| "Where can I buy high-quality scrubs and wipes for professional use?" | Gemini | 22 | Absent |
| "What are the popular brands of restaurant plates?" | AI Mode | 18 | Absent |
| "Where can I buy high-quality plastic or eco-friendly brooms online?" | Gemini | 19 | Absent |
| "Can you recommend effective contamination control products for pharmaceutical cleanrooms?" | ChatGPT | 17 | Absent |
| "Where can I buy high-quality hardwound paper towels and dispensers at a good price?" | Gemini | 19 | Absent |
| "Can you recommend popular brands of lidded trays and platters available online?" | AI Mode | 18 | Absent |
Sample of 25 tracked prompt opportunities. Each maps to a content build in the roadmap below.
Visibility starts with access. The full audit checks AI crawler permissions (GPTBot, ClaudeBot, PerplexityBot, Google-Extended), CDN and bot-management settings, structured data coverage, entity consistency, llms.txt, and whether critical content renders without JavaScript. Two flags surfaced in this sample scope:
Duplicate canonical citations. AI engines cite both meridiancommercial.com/ and meridiancommercial.com as separate sources, splitting citation authority for the homepage. A canonical and redirect cleanup consolidates it.
Citation concentration is inverted. Healthy profiles concentrate 40%+ of citations in 10–15 authority pages. Meridian's top 10 pages hold 13%, a signal that no page is the definitive answer for any category question. This is a content architecture issue, not a crawl issue.
Every item is tagged: things your team can do this week, and things Scylla builds. We tell you which is which because trust converts better than scope creep.
Consolidate homepage canonicals and audit AI crawler access across robots.txt, CDN, and bot management. Confirm GPTBot, ClaudeBot, PerplexityBot, and Google-Extended aren't throttled.
Your teamRewrite the top 15 cited pages for brand conversion: explicit Meridian brand and product naming in the first two sentences, so the 71% of citations that currently go anonymous start crediting the brand.
Scylla sprintShip llms.txt and complete JSON-LD coverage (Organization, Product, FAQPage) across the top 20 category pages, so every engine resolves Meridian products as clean entities.
Scylla sprintAdd direct-answer blocks to the six stronghold category pages to defend ProBin, SlimLine, and SteriClean positions before competitors target them.
Scylla sprintAnswer hubs for the top 10 gap topics by revenue exposure, starting with Restaurant Supply & Equipment (317K/mo, 47% buying intent) and Disinfectant & Cleaning Wipes (113K/mo). Problem-state content, structured for extraction, interlinked to product pages.
Scylla sprintAI Overviews recovery program: target the 170 brand topics with high volume but sub-40 visibility. These already have brand demand; they lack extractable answers.
Scylla sprintCitation source expansion: earn presence on the third-party domains AI engines already trust in this category (industry publications, spec databases, review sources) where Competitor A currently appears and Meridian doesn't.
Scylla + your PRQuarterly re-measurement against this baseline: score, share of voice vs. Competitors A and B, and gap-topic closure rate, reported to leadership in this format.
Scylla retainerEntity and knowledge graph maintenance as product lines evolve, so new SKUs inherit visibility instead of starting from zero.
Scylla retainerOn our own domain, this exact roadmap moved AI Visibility from 41 to 62 and tripled monthly AI mentions. We measure it because we sell it.
"They accomplished more in 30 days than we did in the previous 2 years. I never heard 'You can't...' since working with them."
— Olivia Clemmons, Manager of Digital Marketing, USA for UNHCR"Lots of agencies can build websites, but very few can drive digital strategy and transformation."
— Jim Smith, North American Marketing Director, tesa SEThe 60-minute readout covers every finding above, what to fix in-house, and what it takes to close the gaps that matter. No obligation past the hour.
Book Your Readout Call →