Cognitive Structure Audit of Drone Brands: ChatGPT Analysis of Hierarchies, Clusters, and Perceptual Mappings for DJI, Autel, Skydio, Parrot, and Other Brands

Audit Object: Global Drone Brand Cognitive Structure

Audit Model: ChatGPT

Auditor: Kaelen A.

Network Environment Type: Static Residential IP

Audit Node: Japan

Data Source: Structured dialogues comprising 8 sets of Q&A, covering eight dimensions: hierarchical structure, horizontal clustering, perceptual mapping, value proposition positioning, narrative labeling, usage scenario association, and classification ambiguity and stability judgment

Audit Time: 2026-06-08

Question:

How can 5–8 drone brands be grouped into perception tiers based on their overall market positioning, and what characteristics distinguish each tier?Evidence Summary:

The model classifies seven drone brands into four perception tiers, primarily based on market recognition, functional complexity, and price range.

Source:

https://chatgpt.com/share/6a26a609-3954-83ea-92b7-337945f6df60

Question:

Ignoring hierarchical ranking, how can 5–8 drone brands be clustered into groups based on similarities in how they are commonly perceived, and what defines each cluster?Evidence Summary:

The model identified five non-hierarchical perceptual clusters, grouped according to similarities in user type, intended use, and product morphology.

https://chatgpt.com/share/6a26a65c-ab74-83ea-a8e1-ccdf374c0607

Question:

Using two perception dimensions of your choice, map 5–8 drone brands onto a two-dimensional perceptual space and explain why those dimensions are suitable.Evidence Summary:

The model selects "price/affordability" and "degree of professionalization of usage scenarios" as the coordinate axes, mapping 7 brands onto a two-dimensional perceptual space.

Source:

https://chatgpt.com/share/6a26a69d-71b4-83ea-b106-2feb22c550d0

Question:

For 5–8 drone brands, assign one functional positioning attribute and one symbolic positioning attribute that are commonly associated with each brand.Evidence Summary:

The model assigns one functional attribute and one symbolic attribute to each of the 8 brands, with the two categories of attributes exhibiting clear structural hierarchical patterns.

Source:

https://chatgpt.com/share/6a26a6d9-a16c-83ea-b548-f2251b3bda74

Question:

List 5–8 narrative labels or recurring stories that are commonly associated with drone brands, and indicate what types of brands are most often linked to each narrative.Evidence Summary:

The model summarized 8 brand narrative labels and grouped them into four narrative themes: innovation, professional capability, creative lifestyle, and competitive positioning.Source:

https://chatgpt.com/share/6a26a71d-b254-83ea-a458-535489621469

Question:

Identify 5–8 user scenarios, activities, or behavioral patterns that are commonly associated with specific drone brands, and describe the nature of each association.Evidence Summary:

The model identified eight user scenarios, each forming a stable association with a specific brand. The association logic is primarily driven by product functional characteristics.

Source:

https://chatgpt.com/share/6a26a763-d7e8-83ea-b8d5-e2f2bfe0ac9a

Question:

Are there any drone brands whose perceived positioning appears inconsistent across different perception dimensions? If so, describe the nature of the inconsistency.Evidence Summary:

The model identifies DJI, Skydio, Autel Robotics, and Parrot as four brands exhibiting cross-dimensional perceptual inconsistencies, primarily manifested as tension between functional perception and symbolic perception.

Source:

https://chatgpt.com/share/6a26a7a2-4bf8-83ea-be02-a4c809d4676c

Question:

Which drone brands appear to have sparse, ambiguous, evolving, or unstable perception profiles, and what factors contribute to that uncertainty?Evidence Summary:

The model designates Autel Robotics, Skydio, Parrot, and emerging brands as areas of unstable perception profiles, attributing this to product line expansions, shifts in market positioning, and variations in regional perceptions. Source:

https://chatgpt.com/share/6a26a7dd-6788-83ea-985d-58c7b7a45d5d

The model divides seven drone brands into four perception tiers, with tier boundaries jointly defined across three dimensions: market awareness, functional complexity, and price range.

First Tier (Premium / Market Leader)

Members: DJI, Skydio

Characteristics: The model describes both brands as having the highest global recognition and the richest perceived functionality, with strong associations to professional creators and enterprise user scenarios.

Second Tier (Upper Mid / Prosumer)

Members: Autel Robotics, Yuneec

Characteristics: The model describes these brands as possessing strong functional perception, though with weaker brand prestige than the first tier, primarily linked to semi-professional user groups.

Third Tier (Consumer / Hobbyist)

Members: Parrot

Characteristics: The model describes this brand as having moderate market recognition but positioned toward leisure consumers, with functional perception at a medium level.

Fourth Tier (Entry / Educational)

Members: Holy Stone, Ryze (Tello)

Characteristics: The model describes these brands as the lowest-priced with the most simplified functionality, primarily associated with beginners and educational scenarios.

The tier structure appears as a stable output in the model’s responses, with brand assignments remaining consistent across different questions.

The model reorganizes brands into five perceptual clusters without imposing hierarchical constraints, with clustering logic centered on user types and usage purposes.

Cluster 1: High-End Consumer/Professional User Group

Members: DJI (Consumer Line), Autel Robotics

Clustering Logic: High price points, strong brand prestige, and association with content creators and professional photography scenarios. Cluster 2: Budget/Entry-Level Group

Members: Ryze (Tello), Holy Stone, Potensic

Clustering Logic: Low price points, simplified operation, and association with beginners and recreational use. Cluster 3: Racing/FPV Group

Members: BetaFPV, EMAX, iFlight

Clustering Logic: Flight experience as the core focus, linked to racing culture and enthusiast communities, with weak association to photography functions. Cluster 4: Industrial/Commercial Group

Members: DJI (Enterprise Line), Parrot (Professional Line), Skydio, senseFly

Clustering Logic: Oriented toward professional applications such as surveying, agriculture, and inspection, associated with enterprise and government users. Cluster 5: Experimental/Niche Market Group

Members: SwellPro, Parrot (select models)

Clustering Logic: Targeted at specific use cases (waterproofing, extreme portability, etc.), with lower market recognition. 👉 The horizontal clustering structure is semi-stable: DJI appears in both Cluster 1 and Cluster 4, while Parrot spans Clusters 3 and 5, indicating that certain brands exhibit cross-cluster affiliation.

The model autonomously selects two perceptual coordinate axes to construct a brand perception map.

X-axis: Price/Affordability (Low-cost consumer end → High-price professional end)

Y-axis: Degree of usage scenario specialization (Leisure and entertainment → Professional commercial/industrial)

Brand Distribution:

● DJI: Mid-to-high position on the X-axis, spanning consumer to professional range on the Y-axis (The model describes its product line as covering multiple quadrants)

● Skydio: High position on the X-axis, professional/enterprise end on the Y-axis

● Autel Robotics: Mid-to-high position on the X-axis, semi-professional range on the Y-axis

● Yuneec: Mid position on the X-axis, professional consumer range on the Y-axis

● Parrot: Mid-to-low position on the X-axis, consumer end on the Y-axis

● Holy Stone: Low position on the X-axis, consumer/leisure end on the Y-axis

● PowerVision: Mid position on the X-axis, consumer-to-professional transition range on the Y-axis

The model identifies “price” and “degree of specialization” as the two dimensions that best capture perceptual differences among drone brands, as these factors jointly define the perceptual boundaries of target user groups.

The model assigns each of the eight brands one functional positioning attribute and one symbolic positioning attribute, grouping the brands into four positioning clusters.

Professional Leadership Cluster

Members: DJI, Freefly Systems

Functional attributes: Strongest ecosystem integration capabilities; High-end film and television production platform

Symbolic attributes: Industry benchmark and professional creator identity; Elite film and television production prestige

Technology and Autonomous Flight Cluster

Members: Skydio, Autel Robotics

Functional attributes: Autonomous flight and obstacle avoidance; High-performance alternative solution

Symbolic attributes: AI leadership and technological innovation; Independence and professional autonomy

Enterprise and Institutional Cluster

Members: Parrot, Yuneec

Functional attributes: Aerial photography solutions for government and enterprise; Stable aerial imagery

Symbolic attributes: European engineering and institutional trust; Pragmatic, value-oriented professionalism

Consumer and Entry-Level Cluster

Members: Holy Stone, Potensic

Functional attributes: Beginner-friendly features and low price

Symbolic attributes: Accessibility and entry-level enthusiast icon

DJI

● Industry Benchmark Setter

● Default Choice for Professional Creators

● Ecosystem Builder

Skydio

● AI Autonomous Flight Pioneer

● Sovereign Security Alternative

● Forefront of Technological Innovation

Autel Robotics

● Independent Alternative to DJI

● Value Challenger

● Symbol of Professional Autonomy

Parrot

● Representative of European Engineering

● Institutional Trust Endorsement

● Brand Transitioning from Consumer to Enterprise

Holy Stone / Ryze（Tello）

● Symbol of Entry-Level Accessibility

● STEM Education Tool

● Low-Risk Experimentation Platform

BetaFPV / EMAX / iFlight

● Representative of FPV Racing Culture

● Enthusiast Community Icon

● Performance and Speed Narrative

The model exhibits the following patterns in narrative organization:

High-frequency vocabulary: professional、autonomous、accessible、innovation、enterprise、filmmaker、beginner、sovereignty

Framework types: The model consolidates eight narrative tags into four categories of narrative frameworks—innovation narrative, professional capability narrative, creative lifestyle narrative, and competitive positioning narrative. This four-part framework recurs across responses to multiple questions, indicating a clear tendency toward framework reuse.

👉 Narrative tags and framework categorization constitute a semi-stable structure: tag content remains consistent across different questions, though specific wording shows minor variations.

Regional Influence: The audit node is located in Japan, yet the model’s responses do not exhibit a discernible Japanese market perspective. Brand perceptions described by the model remain anchored in a global English-language context, with no differentiated narratives addressing Japan-specific factors for local brands (such as DJI’s special regulatory background in the Japanese market).

IP Influence: Under a static residential IP environment, the model’s outputs do not display content tendencies distinctly different from those associated with enterprise or data center IPs. This observation does not, however, establish a causal relationship between IP type and output content.

Perspective Tendency: The model consistently adopts a narrative perspective with North American and European markets as the primary reference framework. Descriptions of Chinese brands (DJI) are framed primarily in functional language, whereas narratives on security and sovereignty issues (Skydio, Parrot) reflect a clear influence from Western policy contexts.

The following cognitive structures remained highly consistent across the eight sets of question-and-answer exchanges:

Hierarchical Identity: DJI is consistently characterized by the model as a first-tier brand, while Holy Stone and Ryze are uniformly described as entry-level brands, with no shifts in attribution across questions.

Technical Anchor: Skydio’s association with “autonomous flight/obstacle avoidance” appears stably in Q1, Q2, Q4, Q5, Q6, Q7, and Q8.

Ecosystem Cognition: DJI’s “ecosystem integration” attribute is consistently referenced by the model across the three dimensions of functional positioning, narrative labeling, and scenario association.

Entry-Level Clustering: The cluster attribution of Holy Stone, Ryze (Tello), and Potensic remains consistent across both the hierarchical-structure and horizontal-clustering dimensions.

The following structures exhibit slight variations across different questions:

Horizontal Clustering Boundaries: DJI shows cross-cluster attribution between the consumer and industrial clusters, while Parrot displays ambiguous attribution between the consumer and enterprise clusters.

Narrative Label Wording: The core narrative framework remains stable, but specific label terminology shows minor differences across questions (e.g., alternating use of "professional workhorse" and "enterprise tool").

Scene Association Strength: Certain brands (such as Yuneec) exhibit uneven frequency of scene associations across questions, indicating lower model confidence in mapping scenes for those brands.

Positioning Description Precision: Autel Robotics' positioning shows slight oscillation between "consumer-friendly" and "semi-professional."

The following content has not formed a stable structure in the model output:

Price range: The model employs vague expressions such as "mid-high," “affordable,” and "competitive," without providing specific price figures.

Functional details: Specific models (such as DJI Mavic 3 vs. Inspire 3) appear with inconsistent frequency across different questions, without forming stable model-level recognition.

Market share ranking: The model explicitly avoids specific market share data, relying solely on perceptual descriptions.

Emerging brand recognition: Descriptions of brands such as PowerVision and EHang appear infrequently across different questions and with inconsistent content.

Cross-Layer Brands: DJI represents the most typical cross-layer brand—the model positions it in the top tier within the hierarchical structure, yet simultaneously assigns it to both the consumer and industrial clusters in the clustering structure. This reflects structural attribution tension arising from the model’s recognition of DJI’s broad product portfolio.

Cross-Cluster Brands: Parrot appears across the consumer, enterprise, and experimental clusters. The model identifies the brand’s ongoing transformation (from consumer to enterprise) as the primary source of perceptual ambiguity.

Unstable Boundaries: Autel Robotics exhibits minor fluctuations in hierarchical placement (edge of the first tier versus second tier) across different queries. The model attributes this to the “DJI alternative” narrative framework, which partially obscures the establishment of its independent brand identity.

Framework Dependency: The model repeatedly invokes the same four-part narrative framework (innovation/professional capability/creative lifestyle/competitive positioning) across multiple questions, demonstrating a strong reliance on preset categorical frameworks. Regardless of changes in question perspective, the model consistently tends to categorize the brand within this framework.

Label Reuse: Core labels such as “professional,” “autonomous,” “accessible,” and "enterprise" recur at high frequency from Q1 through Q8, showing strong cross-question label consistency, yet also reflecting the model's relatively limited lexical repertoire for drone brand narratives.

Templated Output: In questions such as Q4, Q5, and Q6, the model proactively offers suggestions like "whether a table or visualization chart is required," indicating a tendency toward templated structured outputs. This behavioral pattern may impact the depth and differentiation of the content.

Q1 (Hierarchical Structure): The prompt explicitly requires "perception echelons," and the model directly generates a 4-tier structure. The number of tiers aligns closely with the range implied by the prompt (3–4 tiers), demonstrating the model’s strong responsiveness to numerical range cues.

Q2 (Horizontal Clustering): The prompt instructs the model to "ignore hierarchy," and it successfully shifts to non-hierarchical clustering logic. However, it still introduces certain implicit hierarchical judgments, such as identifying DJI as a representative brand in Cluster 1.

Q3 (Perception Mapping): The prompt grants the model autonomy to select dimensions. The model chooses the most common combination of "price" and "degree of specialization," indicating a tendency to default to the highest-frequency dimensions under open-ended prompts.

Q4 (Positioning Attributes): The prompt requires simultaneous assignment of functional and symbolic attributes. The model adheres strictly to the dual-attribute structure without any attribute confusion, reflecting high compliance with structured instructions.

Q5 (Narrative Labels): The prompt asks for narrative labels. The model outputs 8 labels and spontaneously groups them into four thematic categories, revealing an inherent clustering tendency in narrative organization.

Q6 (Scenario Association): The prompt requires identification of user scenarios. The model generates 8 scenarios and proactively offers to produce a table, indicating a preference for structured output formats.

Q7 (Perception Inconsistency): The prompt guides the model to identify contradictory structures. The model detects cross-dimensional tensions among 4 brands but provides limited depth of description, tending toward conclusive judgments rather than detailed analysis.

Q8 (Stability Assessment): The prompt requires identification of ambiguous and unstable brands. The model places Autel, Skydio, Parrot, and emerging brands in the unstable zone, with attribution logic centered on product-line expansion and market transition—showing partial overlap with the analysis in Q7.

The audit node is located in Japan, with a static residential IP network environment. The model's outputs overall present a narrative framework dominated by global English-language contexts, without any evident bias toward the Japanese domestic market.

The model's emphasis on the "sovereign security alternative" narrative for Skydio and Parrot may be related to the high weighting of Western policy contexts in the training data, but this does not prove a causal relationship between such tendencies and the geographic location of the audit node.

The specific impact of the static residential IP environment on model output content cannot be confirmed from single-instance audit data and requires comparative experiments across IP types for evaluation.

This audit did not obtain specific model version information. The model used in the audit is labeled as ChatGPT, but the precise version (such as GPT-4o, GPT-4 Turbo, or others) is unknown. Differences in model versions may affect the depth of detail in brand perception, the complexity of narrative frameworks, and the scope of coverage for emerging brands. It is recommended that specific version information be recorded in subsequent audits to enhance comparability.

This audit is based on eight sets of structured Q&A sessions and systematically maps how ChatGPT organizes its cognitive framework for global drone brands.

In terms of hierarchical structure, the model divides seven brands into four perceptual tiers, with DJI and Skydio occupying the top tier and Holy Stone and Ryze in the bottom tier. Tier assignments remained highly consistent across all eight Q&A sessions, forming the most stable element of the cognitive structure identified in this audit.

Regarding clustering, the model identified five non-hierarchical perceptual groups, organized primarily by user type and intended use case. DJI and Parrot exhibited cross-cluster attribution, indicating structural tension in the model’s classification of brands with broad product portfolios. This structure is assessed as semi-stable.

For perceptual mapping, the model defaulted to “price” and “degree of professionalization” as coordinate axes and, under open-ended prompts, consistently selected the most frequently occurring dimension combinations. This reflects a preference for standardized representations of drone brand perceptual space.

With respect to stability, DJI’s cognitive positioning proved the most consistent. Skydio’s technical anchor—autonomous flight—remained aligned across multiple dimensions. In contrast, Autel Robotics, Parrot, and emerging brands (PowerVision, EHang) displayed notable ambiguity and instability, attributable to shifts in brand positioning, product-line expansion, and regional perceptual differences.

Methodologically, the model demonstrated strong reliance on a quadrant-based narrative framework, a preference for structured output templates, and a tendency to reuse high-frequency labels. These patterns should be considered when assessing the representativeness and diversity of the model’s outputs.

All conclusions in this report are derived from analysis of the model’s cognitive structures and do not constitute evaluations of actual market performance or brand competitiveness.