MD
Wave 2 Research / Industrial B2B

Industrial B2B: The 84% Category Absence Pattern

Wave 2 of the Algorithmic Authority Index measured Share of Model across 120 industrial B2B vendors on ChatGPT, Perplexity, and Google AI Overview. 84% receive zero citations. Legacy hardware giants hold the answers. The startups building industrial AI are structurally absent from the category they created.

Published July 2026
Category Industrial B2B / Industrial AI
Platforms ChatGPT, Perplexity, Google AIO
Wave 2 of the Index
Category Absence Rate
84%
of tested industrial B2B vendors receive zero citations across the 12 sub-category query set. The rate rises to 89% for startups founded 2020 to 2025.
Vendors Tested
120
companies across four tiers, from hardware conglomerates to emerging industrial AI startups.
Query Sessions
900
buyer-intent queries executed across 12 sub-categories and 3 AI platforms, June 15 to July 6, 2026.
Zombie Rebrand Rate
78%
of queries against ten tested rebrands returned the old brand name instead of the new one.
Research Context

What Wave 2 tested for the industrial B2B category

Industrial B2B carries the highest density of rebrands, acquisitions, and legacy identities of any vertical in the Index. That makes it the sharpest test of how AI systems handle identity over time. The answer: badly, and in the incumbents' favor.

Wave 2 evaluated 120 vendors across four tiers. Tier 1 holds ten hardware-adjacent conglomerates including Siemens, Rockwell Automation, and Schneider Electric. Tier 2 holds ten mature software pure-plays including C3 AI, Uptake, and Cognite.

Tier 3 holds ten specialized industrial IoT and edge developers. Tier 4 holds 90 emerging industrial AI startups founded between 2020 and 2025. The tier structure isolates whether AI visibility tracks company maturity or structural authority.

The study executed 25 buyer-intent queries per sub-category across 12 sub-categories: industrial AI platforms, predictive maintenance, digital twin, IIoT and edge, MES, APM, SCADA and DCS modernization, OT security, QMS, industrial supply chain visibility, energy management, and operational intelligence. Queries were verticalized by sector (oil and gas, pharma, mining). The configuration yielded 900 documented query sessions across ChatGPT, Perplexity, and Google AI Overview.

Six findings surface across five layers of the Algorithmic Authority Stack.

Category Winners

Legacy conglomerates capture 67% of Share of Model in mature segments

Share of Model measures the percentage of AI responses in a query set that name a specific vendor. In industrial B2B, the metric concentrates around vendors with physical hardware footprints and decades of institutional validation. Siemens, Rockwell Automation, and GE Vernova alone capture 67% of Share of Model across mature software segments.

Dominant across the 12 tested sub-categories

Composite Share of Model above 5% qualifies a vendor as a Category Winner.
Siemens
Rockwell Automation
Schneider Electric
GE Vernova
AVEVA
AspenTech
Emerson
Honeywell

The pattern inverts the cybersecurity result. In cybersecurity, agile software pure-plays dominate the model. In industrial B2B, software-only vendors win only in emerging or specialized niches: C3 AI in industrial AI platforms (34%), Claroty in OT security (41%), ComplianceQuest in QMS (44%).

Sub-Category Rank 1 Winner SOM Rank 2 Winner SOM
Industrial AI PlatformsC3 AI34%Uptake18%
Predictive MaintenanceAspenTech29%Uptake22%
Digital Twin PlatformsSiemens38%GE Vernova21%
IIoT / Edge PlatformsAVEVA31%Litmus Edge19%
MESSiemens36%Rockwell Automation28%
APMGE Vernova33%AspenTech26%
SCADA / DCSEmerson32%Honeywell29%
OT SecurityClaroty41%Dragos33%
QMSComplianceQuest44%Siemens31%
Industrial Supply ChainSAP27%Oracle23%
Energy ManagementSchneider Electric35%Siemens Energy28%
Operational IntelligenceAVEVA30%Braincube22%

Nate Evans, fractional CMO and Head of AI at MISUMI, sees physical AI redrawing this market. Evans advises manufacturers to optimize standard-to-custom component data to earn citations, and points to AI-driven systems as the source of end-to-end supply chain visibility.

The Diagnostic

Industrial buyers ask AI for a shortlist. AI hands them the same eight conglomerates it learned from twenty years of analyst coverage. The market's newest technology is answered with the market's oldest names.

Finding 01

Category Absence: 84% overall. 89% for industrial AI startups.

L1 L5 Category Absence means a Share of Model score below 1%. In industrial B2B, 84% of tested vendors sit below that line. Among Tier 4 startups founded between 2020 and 2025, the rate reaches 89%. Cybersecurity recorded 73% on the same measure. Industrial is the deepest deficit in the Index to date.

The mechanism is trust inheritance. Industrial buyers cluster around legacy vendors with entrenched trust signals. AI systems mirror that behavior and omit startups lacking deep historical footprints. Funding does nothing to offset it.

IoTFlows is invisible in the edge segment. Element is invisible in digital twin despite significant capital raised. Uptake holds 22% Share of Model in predictive maintenance and disappears entirely from supply chain queries. Visibility earned in one sub-category transfers nowhere.

The counter-pattern is vertical specificity. Verticalized vendors achieve 28% higher Share of Model than horizontal generalists. Aegis FactoryLogix wins electronics manufacturing queries while horizontal MES platforms collapse into generic results. One brand earns high citation rates in monitoring by publishing dense, un-gated compliance data. The vertical signal substitutes for the horizontal authority a startup does not have.

Startups deploying horizontal, abstract language fail to establish a stable identity node. That is a structural failure at Layer 1 (Market Identity Clarity) and Layer 5 (Algorithmic Touchpoint Presence) of the Algorithmic Authority Stack.

Layer Failure

An industrial AI startup with $100M raised and a horizontal pitch loses the citation to a 40-year-old conglomerate. A niche vendor with one explicit vertical signal beats both inside that vertical.

Finding 02

Category Confusion: 68% of returned vendors belong to an adjacent category

L3 Category Confusion occurs when distinct categories collapse into a single entity in AI classification. In industrial B2B, 68% of vendors returned across the query set operate in a category adjacent to the one queried. The buyer asks for one system and gets recommended another.

The collapses are consistent. MES and MOM merge into a single execution layer. Predictive maintenance queries return generic MES providers. Industrial IoT platforms and digital twin developers are returned interchangeably. SCADA and DCS modernization merges into historian database solutions.

One session illustrates the stakes. The query "Top MES vendors for pharma" ran in ChatGPT. The engine returned Tulip Interfaces as a primary recommendation. Tulip is a connected worker platform without the regulatory validation of Siemens Opcenter or GE Vernova Proficy. A pharma buyer following that answer shortlists a product that cannot pass their own validation requirements.

A second session asked Perplexity for industrial AI platforms for oil and gas. The system returned Rockwell Automation and Emerson. Both are hardware-heavy control system providers rather than native industrial AI software platforms.

The collapse stems from Semantic Drift at Layer 3. Vendors describe the same offering with inconsistent terminology across marketing surfaces. The model averages the conflicting signals and files the vendor into the wrong competitive set.

Named Failure Pattern

Two out of three vendors AI recommends for an industrial category query do something else. The buyer cannot see the misclassification. The vendor holding the correct answer never enters the evaluation.

Finding 03

Zombie Rebrands: the old name wins 78% of the time

L1 A Zombie Rebrand is a rebrand alive to humans and dead to machines. Industrial B2B exhibits the highest density of rebranded and acquired software assets of any vertical tested. Across ten tested rebrands, AI engines returned the old name 78% of the time.

The old brand name holds an asymmetric signal weight advantage over the new one. Decades of indexed press, documentation, and forum threads reference the former identity. The new identity holds months of coverage against that archive. The model trusts the side with more proof.

The rebrand of Beyond Limits to BeyondAI in February 2026 is a live case. Six months post-launch, queries continue to return Beyond Limits. ChatGPT base models ignore the new identity entirely. The measured lag for BeyondAI currently stands at 18 months and counting.

Former Identity New Identity Rebrand Context Recognition Lag
Beyond LimitsBeyondAIFebruary 2026 rebrand18 months, ongoing
WonderwareAVEVA InTouchLegacy integration24+ months, persistent
OSIsoft PIAVEVA PIPost-acquisition22 months
Meridium APMGE Vernova APMSpin-off rebrand18 months
MindSphereSiemens Industrial IoTLegacy rebrand20 months

Wonderware became AVEVA InTouch, yet AI engines still return Wonderware. OSIsoft PI became AVEVA PI, yet queries still yield OSIsoft. Meridium APM remains highly visible while its new identity as GE Vernova APM stays obscured. Every case is a Layer 1 Market Identity Clarity breakdown running on the Two-Loop Problem, diagnosed below.

Named Failure Pattern

The rebrand launch party happens in week one. The machines attend eighteen months late, and some never arrive.

Finding 04

Trust Seed Substitution routes to analysts, not Reddit

L6 When brand-owned content fails extraction, AI substitutes third-party sources it can extract from. That mechanism is Trust Seed Substitution. Cybersecurity routes to Reddit and G2. Industrial B2B routes to institutional sources: analyst firms, standards bodies, and trade press.

01

Analyst firms

ARC Advisory Group, Gartner Magic Quadrants and Market Guides, LNS Research reports.

Extraction pattern: analyst inclusion functions as the primary authority signal across industrial AI, MES, and predictive maintenance queries.

02

Industrial trade press

Automation World, Control Engineering.

Extraction pattern: independent editorial coverage captures high citation volume where vendor content fails.

03

Standards and government sources

IEEE Explore database, CISA security advisories, WHO prequalification records.

Extraction pattern: standards-aligned records carry weight in regulated segments like OT security and monitoring.

04

Vendor-published original research

Dragos Year in Review. Claroty publications. GxP compliance documentation.

Extraction pattern: original data with published methodology gets cited even when the same vendor's marketing content does not.

05

Consulting studies

McKinsey studies in the industrial AI platform segment.

Extraction pattern: consulting research serves as a tertiary route when analyst and trade coverage thins.

06

Entity registries

Wikipedia, Wikidata, and niche trade directories.

Extraction pattern: absence from these registries removes the fallback surface AI uses to resolve an unfamiliar vendor.

The institutional routing raises the barrier for emerging vendors. A cybersecurity startup can earn Reddit threads and G2 reviews in a quarter. An industrial startup needs ARC coverage, a Gartner mention, or trade press placement. If the brand is not indexed in these external databases, it disappears from AI recommendations. That is a Layer 6 Trust and Proof Signals deficit.

Named Failure Pattern

In industrial B2B, the citation graph runs through institutions with editorial gatekeepers. The vendors who never pitch ARC, LNS, or the trade press have opted out of the graph without knowing it.

Finding 05

Content extraction physics: gated PDFs earn zero citations

L4 AI citation patterns are the inverse of what industrial brands publish. Manufacturers produce short blog posts and gated PDF whitepapers. AI extraction prioritizes long-form, data-dense, un-gated resources in semantic HTML. The mismatch produces a 100% Layer 4 failure rate among the tested B2B blogs.

Corporate Output Pattern Machine Extraction Requirement Impact on Share of Model Layer Failure
Gated PDFs and whitepapersUn-gated, semantic HTML tablesCitations drop to 0%L4 Citation Invisibility
High-volume, short blog postsMulti-thousand-word deep analysisIgnored as repetitive fillerL4 Citation Invisibility
Abstract horizontal messagingIndustry-specific vertical signalsExcluded from vertical queriesL3 Semantic Drift
Inconsistent name variationsUnified cross-platform entity nodesConfident misclassificationL1 Identity Fragmentation

Wave 1 established the Deloitte Paradox: the study's highest-volume publisher held the lowest citation rate. Wave 1 also established the Tier Paradox: smaller players dominate narrow categories through high-density, training-ready formats. Both patterns replicate in industrial B2B. Content volume and citation share move independently.

Michelle Haynes, fractional CMO at Emergence, connects the extraction gap to a commercial one. Haynes treats customer journey mapping as a cross-functional revenue tool rather than a marketing exercise, and shows that separating brand communications from commercial operations stalls growth. The same separation locks the vendor's best technical evidence inside gated assets AI cannot reach.

The Diagnostic

The product sheet behind your lead-gen form contains exactly the data AI needs to cite you. The gate that captures the lead blocks the citation.

The Two-Loop Problem

The Two-Loop Problem in industrial software: an 18 to 24 month memory

AI systems process industrial identity across two loops. Loop 1 is Live Retrieval: it updates in two to four weeks by scanning schemas and current web pages. Loop 2 is Parametric Memory: it hardcodes data into model weights and updates only during core training cycles. In industrial B2B, those cycles run 18 to 24 months.

2-4 weeks
Retrieval Loop
Live web retrieval picks up schema fixes, new trade coverage, and updated pages within weeks. Fast to change, low compounding weight. A rebrand fix here repairs only the surface.
18-24 months
Parametric Loop
Model retraining absorbs the retrieval pattern into memory. Old identities like Wonderware and OSIsoft persist here years after the rebrand. Updating the website repairs Loop 1 while Loop 2 keeps serving obsolete data.
Months 0-6
The Fix Sequence
Week 1: Organization schema with alternateName pointing to the old identity. Month 1: permanent bridge pages explaining the transition. Months 2 to 6: a Parametric Displacement campaign generating third-party trade coverage that teaches base models the new identity.

A rebrand breaks visibility in both loops simultaneously. The sequence above is the only order that works. Schema without bridge pages leaves retrieval ambiguous. Bridge pages without third-party coverage never reach parametric memory. Every quarter of delay hardens the old identity deeper into the next model generation.

Mechanism

The industrial vendor that starts the displacement campaign this quarter enters the next training cycle under its own name. The one that waits gets re-memorized as its former self.

MD
What Comes Next

Run the Snapshot on your brand.

The Industrial B2B Industry Snapshot names the pattern at category scale. The AI Visibility Snapshot names the pattern for your specific brand: your Share of Model, the layer failure producing your Category Absence, and the structural gap between you and the vendors holding your slot.

Methodology

How Wave 2 measured the industrial B2B category

Wave 2 uses the same measurement framework as Wave 1 of the Algorithmic Authority Index, adapted for category-level analysis. A standard prompt library of 25 buyer-intent queries was designed per sub-category, covering direct shortlist prompts, comparison questions, and problem-language queries. Queries were verticalized using industry sectors including oil and gas, pharma, and mining. All sessions were executed from US-based IPs between June 15 and July 6, 2026, using default platform models.

Share of Model was calculated by aggregating Mention Share, Citation Share, and Recommendation Share per vendor. Category Winners hold a composite score above 5%. Category Absence means a composite score below 1%. The resulting dataset comprises 900 fully documented query sessions across 120 vendors and 12 sub-categories.

Sample
120 industrial B2B vendors, 4 tiers
Queries
25 buyer-intent prompts per sub-category
Platforms
ChatGPT, Perplexity, Google AI Overview
Testing window
June 15 to July 6, 2026

Category sources: Gartner Magic Quadrant and Market Guide inclusions (MES, IIoT, QMS). ARC Advisory Group and LNS Research coverage patterns. Beyond Limits / BeyondAI rebrand announcements, LNG2026. AVEVA legacy brand documentation (Wonderware, OSIsoft PI). Practitioner commentary: Nate Evans (MISUMI), Michelle Haynes (Emergence Fractional).

Algorithmic Authority Index™ Wave 2 / Industrial B2B Deep-Dive
Research and methodology by Maria Dykstra. mariadykstra.com