AI Visibility Audit

Undaunted
Visibility Report

Competitive intelligence for AI-mediated buying decisions. Where Undaunted wins, where it loses, and a prioritized three-layer execution plan — built from 150 buyer queries across ChatGPT + Perplexity.

150 Buyer Queries
5 Personas
8 Buying Jobs
ChatGPT + Perplexity
March 8, 2026

TL;DR

8%
Visibility
12 of 150 queries
4.7%
Win Rate
7 wins of 150 queries
138
Invisible
queries where Undaunted absent
20
Recommendations
targeting 151 gap queries
Three things to know
A 70% win rate is wasted when 92% of buyers never encounter the brand
Undaunted wins 70% of visible high-intent queries (7/10) and beats Knightscope directly (2 wins, 0 losses, 4 ties). But overall visibility is 8% (12/150), and in the three early-funnel stages where consideration sets are formed, the rate drops to 2.3% (1/44 queries). The product is the strongest argument for buying Undaunted — but 143 queries return responses where Undaunted is never mentioned.
8% visibility · all 150 queries
Without a sitemap, every new page published is invisible to AI crawlers on day one
No sitemap.xml exists at getundaunted.com/sitemap.xml, no robots.txt explicitly allows AI crawlers or references a sitemap, and 11 of 12 existing pages carry zero freshness signals. These three gaps mean AI platforms have no structured mechanism to discover new content when published or assess whether existing content is current. All 12 L3 content recommendations depend on these L1 infrastructure fixes executing first.
Technical fix · 3 blockers
Knightscope holds 29.1% SOV by content volume — Undaunted's product can compete, but the content doesn't yet exist
Knightscope's 29.1% AI share of voice (67/230 mentions) dwarfs Undaunted's 5.22% (12/230), yet in direct matchups where both appear, Undaunted wins or ties in all 6 encounters (2 wins, 4 ties). The SOV gap is not product inferiority — it is that Knightscope has Comparison pages, deployment guides, case studies, and evaluation content that Undaunted lacks. Twelve new content recommendations targeting 143 gap queries would directly challenge Knightscope's default position in the queries where it currently wins by content volume alone.
Content void · 143 gap queries
Section 1
The Content Gap Behind a 70% Win Rate

Understanding why Undaunted wins 70% of visible high-intent queries but appears in just 8% of all queries — and what structural causes drive that gap.

Early Funnel — Where Undaunted is visible but not winning
Problem Identification
0%
Requirements Building
0%
Solution Exploration
6.2%
Late Funnel — Where Undaunted competes
Comparison
18.8%
Shortlisting
12%
Consensus Creation
8.3%
Validation
4.2%
Artifact Creation
0%

[Mechanism] Three compounding gaps create Undaunted's visibility deficit. First, no content exists for the three early-funnel buying stages — problem identification, solution exploration, and requirements building — where buyers form consideration sets, causing systematic exclusion before Shortlisting begins. Second, the site's technical infrastructure actively limits discovery: without a sitemap, robots.txt, or freshness signals on 11 of 12 pages, AI crawlers have no structured pathway to find or prioritize existing or new content.

Third, the pages that do exist use marketing taglines as H1 headings rather than descriptive topic signals, preventing accurate AI classification of what Undaunted does when pages are crawled. The consequence is that Undaunted's product wins when found but is structurally blocked from being found — a pattern that 20 targeted recommendations can reverse.

Layer 1
Unblock AI Crawlers
Eight technical fixes — including a missing sitemap, missing robots.txt, broken navigation link, and marketing-tagline H1 headings — resolve the infrastructure blockers that limit AI crawler discovery and classification of Undaunted's existing and future content.
5 fixes + 3 checks · Days to 2 weeks
Layer 2
No Existing-Page Optimizations
All 12 commercially relevant pages have insufficient content coverage to qualify for in-place remediation at this stage — no L2 actions apply, and effort is better directed toward creating the missing content inventory.
0 recommendations · 2–6 weeks
Layer 3
Build the Missing Content Footprint
Twelve new content recommendations target all 143 gap queries across every buying stage — from early-funnel problem awareness through competitive Comparison pages and C-suite business case justification assets.
12 recommendations · 1–3 months

[Synthesis] The missing sitemap is the most critical L1 dependency: without it, AI crawlers cannot discover new L3 content pages when published, effectively rendering the entire L3 content investment invisible until the infrastructure gap is resolved. The broken About Us navigation link compounds this by reducing entity-resolution confidence across the domain — AI platforms use company-identity pages to build citation authority, and a broken link on every site navigation reduces that signal for all existing and future pages. Both fixes require less than one week of engineering time but determine whether the L3 content investment generates returns in days or months.

Reference
How to Read This Report

Visibility

Whether Undaunted is mentioned at all in an AI response to a buyer query. Being visible does not mean being recommended — it just means Undaunted appeared somewhere in the answer.

Win Rate

Of the queries where Undaunted is visible, the percentage where it is the primary recommendation — the vendor the AI tells the buyer to evaluate first.

Share of Voice (SOV)

How often a vendor is mentioned by AI across all 150 buyer queries. Measures brand presence in AI-generated answers, not ad spend or traditional media.

Buying Jobs

The 8 non-linear tasks buyers perform during a purchase: Problem Identification, Solution Exploration, Requirements Building, Shortlisting, Comparison, Validation, Consensus Creation, and Artifact Creation.

NIO

Narrative Intelligence Opportunity — a cluster of related buyer queries where Undaunted has no content. Each NIO includes a blueprint of on-domain pages and off-domain actions to close the gap.

L1 / L2 / L3

The three execution layers. L1 = technical infrastructure fixes. L2 = optimization of existing pages. L3 = new content creation and off-domain authority building.

Citation

When an AI tool references a specific webpage as its source. AI systems build recommendations from cited pages — if your pages aren't cited, your content didn't influence the answer.

Invisible Query

A buyer query where Undaunted does not appear in the AI response at all. Distinct from a positioning gap, where Undaunted appears but is not the recommended vendor.
Section 2
Visibility Analysis

Where Undaunted appears and where it doesn't — across personas, buying jobs, and platforms.

[TL;DR] Undaunted is visible in 8% of buyer queries but wins only 5%.

Undaunted's 8% overall visibility (12/150) is driven by structural early-funnel absence — buyers cannot include a brand in their consideration set if it is invisible in 97.7% of the queries (43/44) where they first define their problem and explore solutions.

Platform Visibility

−4 percentage points
Regional Property Manager — widest persona swing
+8 percentage points
Consensus Creation — widest stage swing
DimensionCombinedPlatform Delta
All Queries8%Even
By Persona
Director of Construction / Project Executive15.6%ChatGPT +3 percentage points
Chief Operating Officer3.6%Even
Regional Property Manager19.2%Perplexity +4 percentage points
Director of Security3.2%Even
VP of Operations0%Even
By Buying Job
Artifact Creation0%Even
Comparison18.8%Even
Consensus Creation8.3%ChatGPT +8 percentage points
Problem Identification0%Even
Requirements Building0%Even
Shortlisting12%Even
Solution Exploration6.2%Perplexity +6 percentage points
Validation4.2%Even
Show per-platform breakdown (ChatGPT vs Perplexity raw %)
DimensionChatGPTPerplexity
All Queries6.7%6.7%
By Persona
Director of Construction / Project Executive15.6%12.5%
Chief Operating Officer3.6%3.6%
Regional Property Manager11.5%15.4%
Director of Security3.2%3.2%
VP of Operations0%0%
By Buying Job
Artifact Creation0%0%
Comparison18.8%18.8%
Consensus Creation8.3%0%
Problem Identification0%0%
Requirements Building0%0%
Shortlisting8%8%
Solution Exploration0%6.2%
Validation4.2%4.2%

Visibility by Buying Job

Artifact Creation0% (0/13)
Comparison18.8% (6/32)
Consensus Creation8.3% (1/12)
Problem Identification0% (0/13)
Requirements Building0% (0/15)
Shortlisting12% (3/25)
Solution Exploration6.2% (1/16)
Validation4.2% (1/24)
High-intent visibility
Shortlist + Compare + Validate
12.3% (10/81)
High-intent win rate70% (7/10)
Appearance → win conversion70% (7/10)

Visibility & Win Rate by Persona

Director of Construction / Project Executive15.6% vis · 80% win (4/5)
Chief Operating Officer3.6% vis · 0% win (0/1)
Regional Property Manager19.2% vis · 40% win (2/5)
Director of Security3.2% vis · 100% win (1/1)
VP of Operations0% vis · win
Decision-maker win rate
Chief Operating Officer + Director of Security + VP of Operations
50% (1/2 visible)
Evaluator win rate
Director of Construction / Project Executive + Regional Property Manager
60% (6/10 visible)
Role type gap10 percentage points

Visibility by Feature Focus

AI Detection0% vis (0/8) · 0% win (0)
All Terrain11.8% vis (2/17) · 100% win (2/2)
Analytics Reporting0% vis (0/7) · 0% win (0)
Autonomous Patrol7.5% vis (4/53) · 50% win (2/4)
Indoor Patrol20% vis (1/5) · 100% win (1/1)
Rapid Deployment7.1% vis (1/14) · 100% win (1/1)
Remote Monitoring11.1% vis (2/18) · 50% win (1/2)
System Integration0% vis (0/8) · 0% win (0)
Thermal Surveillance0% vis (0/9) · 0% win (0)
Two Way Audio22.2% vis (2/9) · 0% win (0/2)

Visibility by Pain Point

Construction Theft7.7% vis (1/13) · 0% win (0/1)
Coverage Gaps0% vis (0/9) · 0% win (0)
Guard Cost0% vis (0/19) · 0% win (0)
Guard Safety22.2% vis (2/9) · 0% win (0/2)
Guard Shortage11.1% vis (1/9) · 0% win (0/1)
Inconsistent Patrols0% vis (0/5) · 0% win (0)
Multi Site Scaling14.3% vis (1/7) · 100% win (1/1)
No Realtime Visibility0% vis (0/5) · 0% win (0)

[Data] Overall visibility: 8% (12/150 queries). High-intent visibility: 12.35% (10/81). Early-funnel invisibility: 97.7% (43/44 across Problem Identification, Solution Exploration, Requirements Building).

By persona: VP of Operations = 0% (0/33), COO = 3.6% (1/28), Security Director = 3.2% (1/31), Construction Executive = 15.6% (5/32), Property Manager = 19.2% (5/26). By buying job: Comparison = 18.75% (6/32); Problem Identification = 0% (0/13); Requirements Building = 0% (0/15).

[Synthesis] The visibility pattern reflects a structural funnel failure: Undaunted is systematically excluded at the stages where buyers form consideration sets (97.7% early-funnel invisibility, 43/44), so the modest Comparison-stage visibility (18.75%, 6/32) arrives too late to matter for most buyers. The VP of Operations persona — a veto-holding decision-maker across 33 of 150 queries — has never encountered Undaunted in any of its queries (0/33), representing the most commercially acute single-persona gap in the audit.

Invisibility Gaps — 138 Queries Where Undaunted Doesn’t Appear

43 queries won by named competitors · 18 no clear winner · 77 no vendor mentioned

Sorted by competitive damage — competitor-winning queries first.

IDQueryPersonaStageWinner
⚑ Competitor Wins — 43 queries where a named competitor captures the buyer
und_022"Switching from a traditional guard contract to robotic security — is RaaS or buying hardware the better model?"Chief Operating OfficerSolution ExplorationAsylon Robotics
und_029"We have guards at 8 properties and it's getting expensive — how does centralized remote monitoring with robots work across sites?"Chief Operating OfficerSolution ExplorationRobotic Assistance Devices
und_045"We keep getting hit with overnight theft despite having cameras — best autonomous security robots for commercial properties?"VP of OperationsShortlistingKnightscope
und_046"Top robotic security patrol companies for construction sites that need all-terrain capability"Director of Construction / Project ExecutiveShortlistingAsylon Robotics
und_050"Security robot companies that can deploy within 24 hours on a new construction site without running wires"Director of Construction / Project ExecutiveShortlistingRobotic Assistance Devices
und_051"Best robotic security solutions for replacing overnight guard shifts at commercial properties"VP of OperationsShortlistingKnightscope
und_053"Which security robot vendors offer indoor patrol capability for building lobbies and parking garages?"Regional Property ManagerShortlistingCobalt Robotics
und_054"Our guard company can't tell us how many patrols actually happened — which robotic services provide real patrol analytics?"VP of OperationsShortlistingCobalt Robotics
und_055"Every new site means another security contract and more headcount — which robot vendors scale without proportional cost?"Chief Operating OfficerShortlistingAsylon Robotics
und_056"Best robotic patrol services for securing warehouse and distribution center perimeters"VP of OperationsShortlistingSMP Robotics
Show 33 more competitor wins + 95 uncontested queries

Remaining competitor wins: Knightscope ×17, SMP Robotics ×6, Asylon Robotics ×3, Cobalt Robotics ×3, Robotic Assistance Devices ×2, LiveView Technologies ×1, Prosegur Security ×1. 18 queries with no clear winner. 77 queries with no vendor mentioned. Full query-level data available in the analysis export.

Positioning Gaps — 5 Queries Where Undaunted Appears But Loses

Queries where Undaunted is mentioned but a competitor is positioned more favorably.

IDQueryPersonaBuying JobWinnerUndaunted Position
und_023"How do two-way audio systems on security robots actually deter trespassers in practice?"Regional Property ManagerSolution ExplorationNo Vendor MentionedBrief Mention
und_049"Looking to replace our guard service with something that includes live monitoring — which robotic security vendors have human operators?"Regional Property ManagerShortlistingKnightscopeMentioned In List
und_061"Autonomous security patrol services with two-way audio to verbally warn off intruders"Regional Property ManagerShortlistingSMP RoboticsMentioned In List
und_092"Choosing between Asylon and Undaunted for robotic dog security — what should I know?"Chief Operating OfficerComparisonAsylon RoboticsStrong 2nd
und_128"Case studies of construction companies that reduced theft by deploying security robots on job sites"Director of Construction / Project ExecutiveConsensus CreationRobotic Assistance DevicesMentioned In List
Section 3
Competitive Position

Who’s winning when Undaunted isn’t — and who controls the narrative at each buying stage.

[TL;DR] Undaunted wins 4.7% of queries (7/150), ranks #6 in SOV — H2H record: 7W–4L across 5 competitors.

The SOV rank of #6 (5.22% share, 12/230 mentions) understates Undaunted's competitive position — the product wins direct matchups against every tracked competitor except SMP Robotics, confirming that increasing content volume, not product improvement, is the path to competitive parity.

Share of Voice

CompanyMentionsShare
Knightscope6729.1%
Asylon Robotics3716.1%
SMP Robotics3515.2%
Robotic Assistance Devices3314.3%
Cobalt Robotics3213.9%
Undaunted125.2%
boston_dynamics41.7%
LiveView Technologies31.3%
avigilon20.9%
solink20.9%

Head-to-Head Records

When Undaunted and a competitor both appear in the same response, who gets the recommendation? One query with multiple competitors generates a matchup against each — so H2H totals will exceed the query count.

Win = primary recommendation (cross-platform majority). Loss = competitor was. Tie = neither or third party.

vs. Knightscope2W – 0L – 4T (6 mentioned together)
vs. Cobalt Robotics3W – 1L (4 mentioned together)
vs. Asylon Robotics1W – 1L – 2T (4 mentioned together)
vs. Robotic Assistance Devices1W – 1L – 2T (4 mentioned together)
vs. SMP Robotics0W – 1L (1 mentioned together)

Invisible Query Winners

For the 138 queries where Undaunted is completely absent:

Knightscope24 wins (17.4%)
Asylon Robotics5 wins (3.6%)
Cobalt Robotics4 wins (2.9%)
SMP Robotics4 wins (2.9%)
Robotic Assistance Devices4 wins (2.9%)
Prosegur Security1 win (0.7%)
LiveView Technologies1 win (0.7%)
Uncontested (no winner)95 queries (68.8%)

Surprise Competitors

Vendors appearing in responses not in Undaunted’s defined competitive set.

Daxbot — 4.3% SOVFlagged
Greenbotz — 1.7% SOVFlagged
Boston Dynamics — 1.7% SOVFlagged
Ghost Robotics — 1.3% SOVFlagged

[Synthesis] The SOV gap reflects content volume, not product quality — Knightscope's 29.1% share (67/230) is built on a larger content library, while Undaunted's H2H record shows it wins or ties against Knightscope in every direct matchup (2 wins, 0 losses, 4 ties in 6 shared queries). The 70% conditional win rate (7/10 visible high-intent queries) confirms AI platforms cite Undaunted favorably when content exists; the unconditional rate of 8.6% (7/81 high-intent queries) reveals how rarely that content is found.

Section 4
Citation & Content Landscape

What AI reads and trusts in this category.

[TL;DR] Undaunted had 5 unique pages cited across buyer queries, ranking #7 among all cited domains. 10 high-authority domains cite competitors but not Undaunted.

Five unique cited pages and domain rank #7 establish a foundation: existing content earns citations when AI platforms find it. Scaling to the breadth of citation coverage that buyers encounter requires both more on-domain pages and targeted off-domain authority building — Undaunted currently has neither at sufficient volume.

Top Cited Domains (citation instances)

smprobotics.com74
Knightscope.com51
en.wikipedia.org47
asylonrobotics.com44
radsecurity.com40
Show 15 more domains
cobaltai.com29
getundaunted.com23 (#7)
linkedin.com21
helpforce.ai20
reddit.com19
financialmodel.net18
oxmaint.com18
securenh.com17
greenbotz.co15
daxbot.com15
bostondynamics.com15
alibaba.com15
digitalsecurityguard.com11
bossecurity.com10
wikipedia.org10

Undaunted URL Citations by Page

www.getundaunted.com8
www.getundaunted.com/how-it-works6
www.getundaunted.com/industries/construction-sites6
www.getundaunted.com/pricing2
www.getundaunted.com/undaunted-announces-invest...1
Total Undaunted unique pages cited5
Undaunted domain rank#7

Competitor URL Citations

Note: Domain-level citation counts (above) tally instances per individual domain. Competitor-level counts (below) aggregate across all domains owned by a single vendor, which may include subdomains.

SMP Robotics62 URL citations
Knightscope53 URL citations
Asylon Robotics44 URL citations
Robotic Assistance Devices41 URL citations
Cobalt Robotics34 URL citations
LiveView Technologies4 URL citations
solink2 URL citations
boston_dynamics2 URL citations
ECAM1 URL citations
Prosegur Security1 URL citations

Third-Party Citation Gaps

Non-competitor domains citing other vendors but not Undaunted — off-domain authority opportunities.

These domains cited competitors but did not cite Undaunted pages in the queries analyzed. This reflects citation patterns in AI responses, not overall platform presence.

en.wikipedia.org47 citations · Undaunted not cited
linkedin.com21 citations · Undaunted not cited
helpforce.ai20 citations · Undaunted not cited
reddit.com19 citations · Undaunted not cited
financialmodel.net18 citations · Undaunted not cited

[Synthesis] Five pages generating 23 citation instances at domain rank #7 reveals that Undaunted's existing content is found and cited when discovered — the constraint is inventory breadth, not page quality. Ten third-party domains outrank getundaunted.com by citation frequency, indicating Undaunted needs both more on-domain pages and more third-party coverage to close the citation gap; the NIO blueprints address both through on-domain content creation and targeted off-domain authority building.

Section 5
Prioritized Action Plan

Three layers of recommendations ranked by commercial impact and implementation speed.

[TL;DR] 20 recommendations targeting 151 queries where Undaunted is currently invisible. 5 L1 technical fixes + 3 verification checks, 0 content optimizations (L2), 12 new content initiatives (L3).

Twenty recommendations close the gap in sequence: 8 L1 infrastructure fixes unlock crawler discovery first, then 12 L3 new content recommendations build the full-funnel presence that turns Undaunted's 70% conditional win rate into wins across significantly more than 8% of buyer queries.

Reading the priority numbers: Recommendations are ranked 1–20 across all three layers by commercial impact × implementation speed. Within each layer, items appear in priority order. Gaps in the sequence (e.g., L1 shows 1, 2, then 12) mean higher-priority items belong to a different layer.

Layer 1 Technical Fixes

Configuration and infrastructure changes. Owner: Engineering / DevOps. Timeline: Days to weeks.

Priority Finding Impact Timeline
#1Main navigation links to broken /about-us page (404)Medium< 1 day

Issue: The main site navigation displays an 'About Us' label, and the path https://www.getundaunted.com/about-us returns a 404 error. The actual about page lives at /why-undaunted. If the navigation links to /about-us rather than /why-undaunted, this is a broken link on every page of the site.

Fix: Verify whether the navigation 'About Us' link points to /about-us or /why-undaunted using browser DevTools. If it points to /about-us, either update the link to /why-undaunted or implement a 301 redirect from /about-us to /why-undaunted.

#2No sitemap.xml foundMedium< 1 day

Issue: No sitemap.xml exists at https://www.getundaunted.com/sitemap.xml (returned 404). The site has only 12 discoverable commercial pages, and without a sitemap, AI crawlers must rely entirely on link-following from the homepage to discover content.

Fix: Generate an XML sitemap including all commercially relevant pages with accurate lastmod dates. For a Webflow-hosted site, enable the auto-generated sitemap in Site Settings > SEO. Submit the sitemap to Google Search Console and Bing Webmaster Tools.

#3No visible dates on 11 of 12 pagesMedium1-3 days

Issue: Only 1 of 12 pages (the investment announcement dated February 23, 2026) displays a visible publication or last-updated date. All 11 remaining pages — including all product and industry pages — lack any detectable freshness signal. Without a sitemap providing lastmod timestamps, there is no freshness signal available for these pages.

Fix: Add lastmod timestamps to the sitemap (primary fix — addresses all pages at once). For blog/press content, display visible publication dates. Consider adding a 'Last updated' date to the pricing page and how-it-works page, as these are most sensitive to staleness.

#14Homepage H1 is a marketing tagline, not a descriptive headingMedium< 1 day

Issue: The homepage primary heading appears to be a marketing tagline ('Coverage at Half the Cost') rather than a descriptive heading that communicates what Undaunted is and does. Similarly, the pricing page H1 is 'Zero Maintenance. Simple Setup.' — a benefit statement rather than a heading that signals this is a pricing page.

Fix: Update the homepage H1 to a descriptive heading that includes key terms: company name, product category, and primary value proposition. Example: 'Undaunted — Autonomous Robotic Security Patrols for Commercial Properties.' Update the pricing page H1 to 'Undaunted Pricing — Robotic Security Plans.' Keep marketing taglines as H2s or subheadings.

#15Schema markup status cannot be verified — manual check recommendedMedium1-3 days

Issue: Our analysis method returns rendered page content as markdown, which does not include JSON-LD structured data blocks. We cannot confirm whether the site implements schema markup (Organization, Product, LocalBusiness, FAQ, or Article schemas) on any of its 12 pages.

Fix: Test all commercially relevant pages using Google's Rich Results Test or Schema.org Validator. Implement at minimum: Organization schema on the homepage, Service schema on the how-it-works and industry pages, PriceSpecification on the pricing page, and Article schema on the investment announcement.

Verification Checks

Items requiring manual review before determining if action is needed.

Priority Finding Impact Timeline
#18Client-side rendering status should be verifiedLow< 1 day

Issue: The site appears to be built on Webflow (based on cookie consent patterns and page structure). While Webflow sites are typically server-rendered and AI-crawler-friendly, we cannot confirm CSR status from rendered output alone. All pages returned substantive text content, suggesting no major rendering issues.

Fix: Test the site with JavaScript disabled in Chrome DevTools to verify all critical content renders without JS. Pay particular attention to testimonials, statistics, and video sections. If using Webflow (likely), this is typically not an issue but worth confirming.

#19Meta descriptions and OG tags cannot be verified — manual check recommendedLow1-3 days

Issue: Our analysis returns rendered text content, not raw HTML. We cannot confirm whether pages include meta descriptions, Open Graph tags, or Twitter Card markup.

Fix: Verify meta descriptions and OG tags using a social preview tool or browser DevTools. Ensure every commercial page has a unique, descriptive meta description (under 160 characters) and complete OG tags.

#20No robots.txt file present — no explicit AI crawler policyLow< 1 day

Issue: No robots.txt file exists at https://www.getundaunted.com/robots.txt. All crawlers are implicitly allowed to access all pages, but the site has no explicit crawler access policy.

Fix: Create a robots.txt file that explicitly allows major AI crawlers (GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot, Googlebot) and references the sitemap location once created.

Click any row to expand full issue/fix detail.

Layer 2 — Content Optimization

[Note] No existing pages matched the optimization criteria for Layer 2 recommendations. This typically means gaps are better addressed through new content creation (Layer 3) rather than optimizing existing pages. Review the content inventory in Module 2 to verify page coverage.

Layer 3 Narrative Intelligence Opportunities

Net new content addressing visibility and positioning gaps. Owner: Content Strategy. Timeline: Months.

NIO #1: Early-Funnel Problem Identification Hub
Gap Type: Structural Gap — No content hub exists for buyers articulating security problems before they know robotic patrol is a solution. Undaunted is invisible in 0% of 13 problem-identification queries (0/13), while most responses surface general industry resources rather than any specific vendor.
Critical

Buyers across all five personas — construction executives, COOs, VP of Operations, security directors, and property managers — ask identical problem-aware questions at the start of their journey, and Undaunted appears in none of them (0/13 queries). These 13 queries span guard cost, guard shortage, coverage gaps, construction theft, and real-time visibility — the exact pain points Undaunted solves — yet the site has no content addressing buyers at this stage. Without problem-stage presence, Undaunted is excluded from buyer consideration sets before the shortlist is ever formed. This structural gap is the primary driver of the 97.7% early-funnel invisibility rate (43/44 queries across Problem Identification, Solution Exploration, and Requirements Building).

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_001, und_002, und_003, und_004, und_005, und_006, und_007, und_008, und_009, und_010, und_011, und_012, und_013
“How are construction companies handling equipment theft on job sites when they can't afford 24/7 security guards?”
“Spending $200K a year on security guards and still getting break-ins — what's a better approach?”
“I'm paying for hourly patrols but the guard sits in his car — how do I actually verify patrol coverage?”
Blueprint
  • On-Domain: Create a 'Commercial Property Security Challenges' hub page organizing buyer pain points (guard cost, guard shortage, coverage gaps, theft, patrol accountability) as named problems with statistics — structured for AI extractability with H2-level problem headers.
  • On-Domain: Publish 6–8 problem-specific articles: 'How Commercial Property Operators Replace Expensive Guard Programs,' 'Real-Time Security Alerts vs. After-the-Fact Incident Reports,' 'The Multi-Site Security Scaling Problem and How Companies Solve It' — each written from the buyer's operational perspective, not as product marketing.
  • On-Domain: Add a 'Guard Replacement Guide' pillar page that aggregates all problem content with a clear structural hierarchy (anchor: guard-replacement-guide) — this single URL can rank across dozens of long-tail problem queries.
  • On-Domain: Include first-person buyer language in H2 headings to match the conversational query patterns AI platforms are surfacing (e.g., 'When Your Guard Sits in the Car Instead of Patrolling').
  • Off-Domain: Contribute problem-framing content to industry publications: Security Magazine, Construction Dive, IREM (Institute of Real Estate Management) — establishing third-party citations that AI platforms treat as authority signals.
  • Off-Domain: Seek mentions in guard-replacement discussion threads on LinkedIn and property management forums; Perplexity and ChatGPT frequently surface forum content for problem-identification queries.
  • Off-Domain: Pursue syndication or guest bylines on industry analyst sites covering the physical security technology space (e.g., ASIS International, SecurityInfoWatch) to build third-party citation volume for problem-stage queries.
Platform Acuity

ChatGPT (medium): Problem-identification queries return general advisory responses that cite industry publications rather than vendor pages. ChatGPT favors authoritative third-party sources at this stage; building off-domain citations is as important as on-domain content for this NIO. Perplexity (high): Perplexity surfaces well-structured problem-framing content with clear headings and self-contained passages. An on-domain problem hub with H2-organized pain points and statistics has high retrieval probability for these conversational queries.

NIO #2: Solution Education Hub — How Robotic Security Works
Gap Type: Structural Gap — No educational content series exists for buyers exploring robotic security as a solution category. Undaunted achieves 6.3% visibility across 16 solution-exploration queries (1/16), and that single appearance (und_023) is a positioning loss where a competitor wins the citation.
Critical

Buyers at the solution-exploration stage ask comparative and operational questions — 'how does robotic patrol actually work on a construction site?', 'what's the difference between RaaS and buying hardware?', 'can robots integrate with our existing cameras?' — and Undaunted appears in 1 of 16 such queries (6.3%), with that appearance resulting in a competitive loss. The absence here is structural: the site has no content that explains how Undaunted's service works at a category-education level, separate from product marketing copy. Competitors who have published category-education content (Knightscope's blog, Cobalt's resource center) are capturing buyers who would consider Undaunted favorably if they knew it existed. Without solution-stage content, Undaunted cannot be included in the buyer's mental shortlist before the Comparison stage.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_014, und_015, und_016, und_017, und_018, und_019, und_020, und_021, und_022, und_023, und_024, und_025, und_026, und_027, und_028, und_029
“How do autonomous security robots actually work on construction sites? Do they patrol on their own or need an operator?”
“We're paying $15K/month for guard service — how does remote monitoring with robots compare on cost and effectiveness?”
“Can you integrate robotic patrol systems with existing access control and alarm systems?”
Blueprint
  • On-Domain: Publish a 'How Robotic Security Patrol Works' pillar page covering: autonomous navigation, RaaS vs. hardware purchase, operator-in-the-loop monitoring, AI detection mechanics, and deployment process — with H2-level section headers matching buyer query language.
  • On-Domain: Create feature-specific explainers as sub-pages: 'Robotic Patrol vs. Guard Service: Real Cost Comparison,' 'How AI Detection Reduces False Alarms,' 'Indoor and Outdoor Patrol Explained' — each self-contained enough for AI citation.
  • On-Domain: Add a FAQ section to the how-it-works page with exact-match H3 questions pulled from these 16 queries, answered in 2–3 factual sentences each for maximum AI extractability.
  • On-Domain: Build a 'Security Technology Comparison' page contrasting robotic patrol, camera towers, virtual guard services, and human guards — this captures the 'build vs. buy' and 'robotic vs. cameras' Comparison queries in this cluster.
  • Off-Domain: Submit guest articles to security and operations publications explaining 'What Property Managers Get Wrong About Robotic Security' — thought leadership that builds citation authority for solution-exploration queries.
  • Off-Domain: Engage with YouTube and podcast content in the physical security space; Perplexity increasingly surfaces video and podcast transcripts for 'how does X work' queries.
Platform Acuity

ChatGPT (high): Solution-exploration queries return structured explanations. ChatGPT cites pages with clear entity signals (organization schema, company name in headings) and factual, Comparison-rich content. A well-structured how-it-works pillar page would be a strong citation candidate. Perplexity (high): Perplexity surfaces Comparison and explainer content for 'how does X work vs. Y' queries. Self-contained FAQ sections and Comparison tables are the highest-receptivity format for this buying stage.

NIO #3: Buyer Evaluation Framework — Requirements Building Content
Gap Type: Content Type Deficit — Undaunted has no evaluation framework content: no RFP templates, vendor assessment checklists, or buyer's guides. Visibility is 0% across 15 requirements-building queries (0/15), and no vendor is consistently winning these queries — the gap is content type, not competitive loss.
Critical

Requirements-building queries are written by veto-holding buyers assembling vendor shortlists — COOs demanding TCO data, VP of Operations specifying analytics requirements, security directors defining integration standards. Undaunted appears in none of these 15 queries (0/15), meaning it is systematically excluded from the RFP-specification stage before Shortlisting begins. These buyers are not asking about Undaunted specifically; they are building evaluation criteria, and the vendor whose content best shapes those criteria has a structural advantage in every subsequent evaluation stage. No AI platform surfaces Undaunted as the source of evaluation frameworks, checklists, or vendor assessment tools in this cluster.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_030, und_031, und_032, und_033, und_034, und_035, und_036, und_037, und_038, und_039, und_040, und_041, und_042, und_043, und_044
“Key requirements for evaluating robotic security patrol services for a commercial real estate portfolio”
“What should a COO ask about total cost of ownership when evaluating robotic security vs guard services?”
“Our guards don't produce any reports — what analytics and patrol verification should a robotic security vendor provide?”
Blueprint
  • On-Domain: Publish a 'Robotic Security Vendor Evaluation Guide' — a comprehensive buyer's guide with sections on: terrain capability requirements, deployment timeline expectations, analytics and reporting minimums, AI detection accuracy benchmarks, integration compatibility, and pricing models. Structured as a downloadable resource with a dedicated landing page.
  • On-Domain: Create a 'Questions to Ask a Robotic Security Vendor' page with 20–30 evaluation questions organized by decision-maker role (COO questions, operations questions, security director questions) — this directly matches the query patterns in this cluster.
  • On-Domain: Build an 'RFP Template for Robotic Security Patrol Services' page with downloadable template — this explicitly addresses und_138 (Artifact Creation) and creates a high-value, AI-citable resource that establishes Undaunted as the authority on vendor selection.
  • On-Domain: Add a 'Total Cost of Ownership Calculator' or structured TCO Comparison framework (robotic patrol vs. guard service over 3 years) — directly serving COO and VP personas making financial cases.
  • Off-Domain: Submit the buyer's guide to property management associations (BOMA, IREM, NMHC) for distribution to their member networks — creating third-party citations that AI platforms recognize as authority signals.
  • Off-Domain: Publish the RFP template on shared resources platforms (LinkedIn Articles, SlideShare, Scribd) to maximize retrieval probability by AI platforms that index structured documents.
Platform Acuity

ChatGPT (high): Requirements-building queries return structured lists and frameworks. ChatGPT cites pages that directly answer 'what questions should I ask' or 'what requirements matter' in a well-organized format. Pages with H2 headers matching question stems have high citation probability. Perplexity (high): Perplexity surfaces evaluation checklists and structured assessment frameworks for vendor-Comparison queries. Tables with feature requirements, self-contained bullet lists, and clearly labeled criteria sections are the highest-receptivity formats.

NIO #4: All-Terrain & Outdoor Detection Capability
Gap Type: Content Type Deficit — No dedicated outdoor terrain capability or thermal surveillance page exists on the Undaunted site. Undaunted is invisible in 0% of 16 Shortlisting, Comparison, Validation, consensus, and artifact queries focused on all-terrain and thermal detection capability (0/16), while SMP Robotics, Asylon, and Knightscope win these queries by virtue of having published capability content.
High

Outdoor terrain capability is Undaunted's physical differentiator — the robot operates on construction sites, gravel lots, hills, and uneven ground — but this capability is not documented in any content that AI platforms can extract and cite. Across 16 queries where buyers are actively Shortlisting, comparing, or validating vendors for outdoor terrain requirements (including rough terrain, adverse weather, and thermal night detection), Undaunted appears in none (0/16). SMP Robotics wins 6 of these 16 queries (37.5%) — on terrain and thermal capability specifically — despite Undaunted's product having comparable or superior outdoor performance. The fix requires a single dedicated 'Outdoor Capability' page with specific terrain specifications, weather ratings, thermal detection specs, and operational evidence in a format AI platforms can extract and attribute to Undaunted.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_046, und_047, und_048, und_068, und_073, und_076, und_078, und_091, und_096, und_097, und_108, und_113, und_114, und_122, und_135, und_138
“Top robotic security patrol companies for construction sites that need all-terrain capability”
“SMP Robotics vs Knightscope for large outdoor property patrol — which handles more terrain types?”
“How well do security robots actually hold up in bad weather — rain, snow, extreme heat?”
Blueprint
  • On-Domain: Create a dedicated 'Outdoor & All-Terrain Security Patrol' capability page with specific technical specs: terrain types supported (gravel, grass, slopes, unpaved), IP weather rating, operating temperature range, payload and camera specs for thermal detection — formatted as extractable data points with H3-level spec categories.
  • On-Domain: Publish a 'Thermal Detection & Night Surveillance' sub-page covering: thermal camera specifications, detection range, night patrol effectiveness vs. standard CCTV, and how thermal detection catches intruders that visible cameras miss.
  • On-Domain: Add a 'Construction Site Security' industry page with operational case evidence: patrol coverage of active construction sites, terrain traversal examples, how robots move between sites as project phases complete, and deployment timelines for new sites.
  • On-Domain: Include a competitive Comparison table: Undaunted vs. SMP Robotics vs. Asylon vs. Knightscope on outdoor terrain metrics — directly addressing the Comparison queries in this cluster.
  • Off-Domain: Publish field performance data or third-party test results in construction industry publications (ENR, Construction Dive) — AI platforms weight third-party performance evidence highly for capability claims.
  • Off-Domain: Seek inclusion in security industry capability roundups on SecurityInfoWatch and SSI (Security Sales & Integration) covering outdoor and terrain-capable patrol robots.
Platform Acuity

ChatGPT (medium): Comparison queries in this cluster return competitor-sourced capability claims. ChatGPT needs third-party corroboration of Undaunted's terrain specs — publishing performance evidence in industry publications is as important as on-domain documentation. Perplexity (high): Perplexity surfaces structured spec tables and Comparison data for terrain and thermal capability queries. A page with well-formatted technical specifications and a Comparison table would have high retrieval probability for Shortlisting and Comparison queries.

NIO #5: Rapid Deployment & Construction Site Mobility
Gap Type: Content Type Deficit — No deployment process or construction site mobility content exists on the Undaunted site. Undaunted is invisible in 0% of 10 queries spanning Shortlisting, Comparison, Validation, consensus, and artifact stages focused on deployment speed and multi-site portability (0/10), while Asylon, RAD, and SMP Robotics win by default.
High

Rapid deployment is a primary differentiator for Undaunted in the construction market — the ability to deploy on a new site without running wires and to redeploy between sites as projects finish — but no content explains this capability to buyers. Across 10 queries where buyers are specifically asking about deployment timelines, site readiness, and multi-site robot portability, Undaunted appears in none (0/10). Construction executives and COOs in this cluster are actively Shortlisting and asking AI platforms 'how fast can you get robots on-site?' — a question Undaunted's product answers better than most competitors. Without documented deployment process content, Undaunted cannot be cited in these commercially critical queries.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_050, und_057, und_065, und_069, und_087, und_102, und_134, und_141, und_146, und_150
“Security robot companies that can deploy within 24 hours on a new construction site without running wires”
“Which robotic security vendors can move their robots between construction sites as projects finish?”
“What's involved in deploying a robotic security patrol on a new job site — how fast can you get it running?”
Blueprint
  • On-Domain: Create a 'Rapid Deployment' capability page documenting: site activation timeline (days from contract to first patrol), what site readiness requires (power source, connectivity, or lack thereof), the redeployment process for moving between construction phases, and a case example of a multi-site deployment sequence.
  • On-Domain: Publish a 'Construction Site Security Deployment Guide' — a practical operational document covering pre-deployment checklist, site assessment criteria, typical go-live timeline, and what buyers should expect in the first 30 days.
  • On-Domain: Build a 'Multi-Site Deployment' page specifically addressing how Undaunted services a portfolio of active construction sites simultaneously, with scalability data (how many sites can be managed under one service agreement).
  • On-Domain: Add a 'Deployment FAQ' section to the existing how-it-works page with questions matching und_035 and und_024 directly: 'How long does it take to get a robot patrolling a new site?' and 'What site prep is required?'
  • Off-Domain: Publish deployment case evidence in construction industry outlets (AGC, Construction Executive) documenting real deployment timelines and multi-site operations — third-party evidence that AI platforms cite alongside the on-domain content.
  • Off-Domain: Engage on LinkedIn with construction executives sharing site security challenges; Perplexity surfaces LinkedIn content for operational how-to queries in this cluster.
Platform Acuity

ChatGPT (medium): Deployment speed queries return responses that cite competitor press releases and news articles with deployment timeline data. ChatGPT needs third-party corroboration — press coverage of Undaunted deployments would significantly increase citation probability. Perplexity (high): Perplexity retrieves structured process content and timelines for operational how-to queries. A deployment guide with a clearly structured timeline (Day 1: site assessment, Day 3: configuration, Day 7: live patrol) would have strong retrieval probability.

NIO #6: Remote Monitoring & Multi-Site Operations Hub
Gap Type: Content Type Deficit — No centralized remote monitoring operations content exists on the Undaunted site. Undaunted is invisible or losing in 10 of 10 Shortlisting, Comparison, Validation, consensus, and artifact queries about remote monitoring and multi-site patrol coordination (0/10 wins), while Knightscope, Cobalt Robotics, and RAD win by default with their monitoring center documentation.
High

Remote monitoring — the ability to replace on-site guards with a centralized monitoring operation across multiple properties — is the economic core of the robotic security value proposition, and Undaunted has no content explaining how its monitoring capability works. COOs in this cluster are asking AI platforms 'how does centralized monitoring work across 8 properties?' (und_029) and 'which vendor has the best remote monitoring operators?' (und_100), and Undaunted is absent from every response. The 10 queries in this cluster span the full buying journey (Shortlisting through artifact creation), meaning the gap persists from initial consideration to vendor-selection documentation. Cobalt Robotics wins 3 of 10 queries (30%) in this cluster through superior remote monitoring content; Undaunted's existing page structure has no equivalent asset.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_049, und_055, und_080, und_084, und_100, und_117, und_123, und_132, und_137, und_147
“We have guards at 8 properties and it's getting expensive — how does centralized remote monitoring with robots work across sites?”
“Comparing remote monitoring center quality across top robotic security vendors — who has the best operators?”
“Looking to replace our guard service with something that includes live monitoring — which robotic security vendors have human operators?”
Blueprint
  • On-Domain: Create a 'Remote Monitoring Operations' page documenting: how the monitoring center works, operator training and certification, alert response SLA (the time from detection to operator action), escalation protocol, and how multi-site coverage is managed under one service agreement.
  • On-Domain: Publish 'Multi-Site Remote Security Management' content with a scalability narrative: how adding a new property to the service works, what the per-site economics look like at scale (2 properties vs. 10 properties), and how a central operations team replaces distributed guard headcount.
  • On-Domain: Add a 'Monitoring Center Quality Scorecard' page — a Comparison framework for evaluating remote monitoring quality across vendors, with Undaunted's own standards documented as the baseline. This directly addresses und_100 and creates a content asset that shapes evaluation criteria.
  • On-Domain: Build a case narrative around guard-to-monitoring transition: 'How a Property Manager with 8 Sites Replaced Their Guard Program in 90 Days' — framed as an operational story, not a sales piece.
  • Off-Domain: Seek inclusion in industry reports comparing remote monitoring SOC (Security Operations Center) quality across robotic security vendors — Perplexity and ChatGPT cite these Comparison reports heavily in monitoring quality queries.
  • Off-Domain: Publish monitoring response time benchmarks and operator training standards in ASIS International publications to establish third-party credibility for monitoring quality claims.
Platform Acuity

ChatGPT (high): Monitoring quality queries return responses citing vendor pages with documented SLAs and operator training standards. ChatGPT responds well to authoritative operational documentation — a monitoring operations page with specific response time data and process description would be directly citable. Perplexity (medium): Perplexity surfaces Comparison content for monitoring quality queries. Including a structured Comparison table (Undaunted vs. Cobalt vs. Knightscope on monitoring metrics) improves retrieval probability, but Perplexity currently relies on competitor pages that have more structured content.

NIO #7: Patrol Analytics & Accountability Dashboard
Gap Type: Content Type Deficit — Undaunted has no patrol analytics or reporting content, and no dedicated analytics product page. Invisible in 0% of 5 queries (0/5) where buyers explicitly demand patrol verification, route analytics, and reporting dashboards — while Cobalt Robotics and Knightscope win 4 of these 5 queries (80%) through documented analytics capabilities.
High

The patrol accountability gap is the most common trigger for switching from human guards — buyers whose guards skip patrols with no verifiable record are actively searching for robotic security vendors who provide analytics dashboards. Undaunted is invisible in all 5 queries in this cluster (0/5), which span Shortlisting, Comparison, Validation, and consensus stages — meaning the gap persists through the full buying journey. Cobalt Robotics is the primary winner (3 of 5 queries, 60%) because its analytics documentation is the most complete in the category; Knightscope wins 1. Publishing a dedicated analytics capability page would address the second-most acute accountability concern buyers raise and positions Undaunted directly against Cobalt's strongest content asset.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_054, und_063, und_093, und_121, und_131
“Our guard company can't tell us how many patrols actually happened — which robotic services provide real patrol analytics?”
“Security robot vendors that provide detailed patrol verification so I know routes are actually being covered”
“Can we get insurance premium reductions by switching to robotic security with verified patrol logs?”
Blueprint
  • On-Domain: Create a dedicated 'Patrol Analytics & Reporting' page documenting: what data is captured per patrol (GPS route, timestamps, incident flags, coverage heatmaps), report formats available (real-time dashboard, weekly summary, incident log), how reports are delivered, and how patrol verification data is formatted for insurance or audit purposes.
  • On-Domain: Include sample report screenshots or structured data examples (anonymized) on the analytics page — AI platforms prioritize pages with specific, verifiable data over pages with only written descriptions.
  • On-Domain: Add an 'Insurance & Liability' section to the analytics page documenting how verified patrol logs reduce insurance risk and support claims evidence — directly addressing und_121 and und_131, and introducing a CFO-level commercial argument.
  • On-Domain: Link the analytics page from the pricing page, how-it-works page, and any Comparison pages — making it accessible from multiple crawl paths to increase discovery probability.
  • Off-Domain: Seek coverage in risk management publications (RIMS, PropertyCasualty360) on how robotic patrol data supports insurance claims and premium reduction — creating third-party citations for the insurance angle.
  • Off-Domain: Submit case evidence of patrol log data supporting insurance claims or operational accountability to property management industry publications (IREM Journal, Multifamily Executive).
Platform Acuity

ChatGPT (high): Analytics queries return structured responses citing specific vendor features (Cobalt's dashboard, Knightscope's reporting portal). ChatGPT cites pages with specific data format descriptions and named capabilities — a dedicated analytics page with named report types would be directly citable. Perplexity (high): Perplexity surfaces structured feature Comparison content for analytics queries. A Comparison table of analytics capabilities across vendors, anchored on the Undaunted analytics page, would have the highest retrieval probability in this cluster.

NIO #8: AI Detection Accuracy & Active Deterrence
Gap Type: Content Type Deficit — Undaunted has no AI detection accuracy content and no two-way audio deterrence documentation. Invisible or losing in all 10 queries spanning AI detection and verbal deterrence (0/10 wins), while competitors (Asylon, Cobalt Robotics, SMP Robotics) win by default with published detection specification content.
Medium

Buyers are asking two related questions in this cluster: 'How accurate is the AI detection, and will it reduce false alarms?' (AI-Powered Threat Detection queries) and 'Does the robot actually deter trespassers, or just record them?' (Two-Way Audio Deterrence queries). Undaunted's product answers both — with AI-powered threat classification and live two-way audio for verbal deterrence — but neither capability is documented in AI-citable form on the site. Security directors (who evaluate false alarm rates) and property managers (who care about active deterrence vs. passive recording) are the primary buyers. This cluster is medium priority because the query volume is smaller (10 queries) and the competitive wins in this space are less concentrated than in the autonomous patrol or Comparison clusters.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_052, und_061, und_086, und_099, und_110, und_111, und_118, und_124, und_129, und_142
“Autonomous security platforms with AI-powered threat detection that filters out false alarms”
“Which robotic security vendor has the best two-way audio and verbal deterrence capability?”
“Do autonomous security robots actually deter criminals or just record them for evidence?”
Blueprint
  • On-Domain: Create an 'AI Threat Detection' capability page with: detection categories (human vs. animal vs. vehicle vs. object), false alarm rate data, how the AI model is trained and updated, detection range specs, and how alerts are verified before escalation to the monitoring center.
  • On-Domain: Create a 'Two-Way Audio & Active Deterrence' page documenting: how verbal deterrence works in practice, case examples where audio warnings stopped trespassers before the monitoring center needed to call police, audio system specifications, and how this compares to passive CCTV recording.
  • On-Domain: Add 'Deterrence vs. Detection' Comparison content that explicitly addresses und_118 — buyers debating whether robots 'actually deter' vs. 'just record' — with operational evidence of deterrence outcomes.
  • On-Domain: Embed specific detection performance data (if available) in the product pages as extractable data points — AI platforms cite specific numbers (e.g., '94% detection accuracy') far more readily than qualitative descriptions.
  • Off-Domain: Pursue third-party Validation of detection accuracy metrics — an independent test or security industry evaluation creates the authoritative citation that ChatGPT needs to surface Undaunted for detection accuracy queries.
  • Off-Domain: Submit content to security trade publications covering AI detection technology in robotic patrol (SecurityInfoWatch, ASIS) — establishing Undaunted as a credible voice on AI detection accuracy.
Platform Acuity

ChatGPT (medium): AI detection queries return responses citing competitor specs and independent security research. ChatGPT weights third-party Validation of detection claims higher than vendor-self-reported data — off-domain authority building is essential for this NIO. Perplexity (medium): Perplexity surfaces spec-level comparisons for detection accuracy queries. Specific numbers (false alarm rate, detection accuracy percentage, audio range) structured in an extractable format improve citation probability, but overall receptivity is moderate given the technical nature of the content.

NIO #9: Indoor Patrol & System Integration
Gap Type: Content Type Deficit — Undaunted has no indoor patrol content and no system integration documentation. Invisible in 0% of 6 queries spanning indoor patrol capability and security system integration (0/6), while Cobalt Robotics (indoor specialist) and Knightscope win these queries with dedicated indoor and integration content.
Medium

Property managers and COOs managing mixed indoor/outdoor properties (lobbies, parking garages, and outdoor perimeters) ask AI platforms whether security robots can cover both environments — and Undaunted appears in none of the 6 queries in this cluster (0/6). The indoor patrol queries (und_053, und_060) are particularly commercial because they come from property managers who need a single vendor for full-property coverage rather than separate solutions for indoor and outdoor patrol. System integration queries (und_058, und_064, und_089, und_145) are medium commercial weight but relevant to security directors building a technology stack. This cluster is medium priority because the query volume is small (6 queries) and the coverage_status for Indoor Patrol Capability is 'missing' — Undaunted may need to assess whether indoor patrol is a product capability before creating content.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_053, und_058, und_060, und_064, und_089, und_145
“Which security robot vendors offer indoor patrol capability for building lobbies and parking garages?”
“Which autonomous security companies offer both patrol robots and sensor networks as a package?”
“Which security robot companies have the best integration with existing camera and alarm systems — RAD vs Cobalt vs Knightscope?”
Blueprint
  • On-Domain: If indoor patrol is a supported capability: create a dedicated 'Indoor & Outdoor Patrol' page documenting indoor environments supported (lobbies, parking structures, corridors), robot dimensions and maneuverability specs for indoor use, and how the service transitions between interior and exterior patrol zones.
  • On-Domain: Create a 'System Integration' page listing compatible platforms: access control systems (brands/protocols), alarm system integrations (APIs, ONVIF compatibility), and how the Undaunted platform connects to existing camera infrastructure — with specific named integrations buyers can verify.
  • On-Domain: Build an 'Integration Checklist' downloadable resource for security directors adding robotic patrol to an existing camera and alarm stack — directly addressing und_145 and und_041 (from NIO 003).
  • On-Domain: If indoor patrol is not currently supported, publish a transparent 'Outdoor-Optimized Patrol: When to Choose Undaunted vs. an Indoor Specialist' page — honest positioning that builds trust and prevents buyers from disqualifying Undaunted unfairly.
  • Off-Domain: Pursue inclusion in security technology integration directories (PSA Security Network, security system integrator resources) to establish Undaunted's integration capabilities with third-party Validation.
  • Off-Domain: Contribute to security integrator forums and publications (Security Dealer & Integrator, SDM) on robotic patrol integration best practices.
Platform Acuity

ChatGPT (medium): Indoor patrol queries return responses dominated by Cobalt Robotics, which has strong indoor capability documentation. ChatGPT cites specific capability evidence — a page documenting indoor operating specs with specific environment types listed would compete effectively. Perplexity (medium): Integration queries surface structured compatibility lists and technical documentation. A system integration page with named compatible platforms and protocols would have stronger Perplexity retrieval probability than narrative-only descriptions.

NIO #10: Competitor Comparison Pages — Direct Matchup Content
Gap Type: Content Type Deficit — Undaunted has no competitor Comparison pages. Invisible or losing in 13 of 13 Comparison-stage autonomous patrol queries (0/13 wins), while Knightscope is cited in 9 of 13 (69.2%) of these queries as the frame of reference. Buyers in active vendor Comparison never encounter Undaunted because no 'Undaunted vs. X' content exists.
Critical

Comparison is the highest-visibility buying stage in this audit — 18.75% overall (6/32), with an 83.3% win rate (5/6) when Undaunted appears. These 13 queries represent the stage where buyers are writing names on a shortlist and comparing them side by side, and Undaunted appears in none of them (0/13). Knightscope wins 9 of these 13 queries (69.2%) not because buyers prefer Knightscope, but because Knightscope has the most extensive Comparison content in the category. The H2H record confirms the product is competitive: Undaunted beats Knightscope 2-0-4 when both appear in the same query. The fix is Comparison pages — 'Undaunted vs. Knightscope,' 'Undaunted vs. Asylon,' 'Undaunted vs. Cobalt Robotics' — that give AI platforms structured content to include Undaunted in these buyer-decisive queries.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_070, und_072, und_075, und_079, und_081, und_082, und_083, und_088, und_092, und_094, und_095, und_098, und_101
“Knightscope vs Cobalt Robotics for commercial property security — which is better for outdoor patrol?”
“Asylon Robotics pricing vs Knightscope pricing — what does robotic security actually cost per month?”
“Choosing between Asylon and Undaunted for robotic dog security — what should I know?”
Blueprint
  • On-Domain: Create 'Undaunted vs. Knightscope' Comparison page: structured side-by-side on terrain capability, deployment speed, pricing model (RaaS vs. lease), monitoring center quality, analytics reporting, and use case fit (construction sites vs. corporate campuses) — with H3 headers for each dimension.
  • On-Domain: Create 'Undaunted vs. Asylon Robotics' Comparison page focusing on the quadruped patrol differentiator, terrain performance head-to-head, pricing transparency, and deployment flexibility for construction applications.
  • On-Domain: Create 'Undaunted vs. Cobalt Robotics' Comparison page emphasizing outdoor vs. indoor specialization — Cobalt's indoor focus vs. Undaunted's outdoor/terrain capability, with explicit guidance on which use cases favor each vendor.
  • On-Domain: Build a general 'Robotic Security Vendor Comparison' hub page that aggregates all Comparison content and links to the specific matchup pages — this catches Comparison queries that name multiple competitors or ask 'what are my options?'
  • Off-Domain: Pursue inclusion in security analyst Comparison reports covering robotic patrol vendors — analyst coverage of Undaunted in a multi-vendor Comparison creates authoritative third-party citations that AI platforms weight highly for Comparison queries.
  • Off-Domain: Engage with security procurement communities (ASIS, Securing Industry, SDM) where vendor Comparison discussions occur — building brand recognition before buyers ask Comparison questions of AI platforms.
Platform Acuity

ChatGPT (high): Comparison queries directly return competitor Comparison content. ChatGPT cites vendor Comparison pages by name when they exist — 'Undaunted vs. Knightscope' would be directly surfaced for the most commercially valuable queries in the audit. Entity trust (broken_about_us_link fix) is important here. Perplexity (high): Perplexity is the highest-receptivity platform for structured Comparison tables. A Comparison page with a side-by-side table (features as rows, vendors as columns) would have the strongest retrieval probability across all platforms for these queries.

NIO #11: Autonomous Patrol Shortlisting Presence & Vendor Validation
Gap Type: Structural Gap — Undaunted has no content that establishes its autonomous patrol service in Shortlisting or vendor-Validation contexts. Invisible in all 17 Shortlisting and Validation queries for autonomous patrol (0/17 wins), while Knightscope, Asylon, and RAD are the default choices AI platforms recommend when buyers assemble vendor shortlists or research competitor limitations.
Critical

Shortlisting is a high-intent buying stage — buyers here are asking 'which vendors should I put on my evaluation list?' — and Undaunted appears in none of the 6 Shortlisting queries in this cluster (0/6), despite autonomous patrol being its core product. The 11 Validation queries in this cluster represent a different but equally critical gap: buyers researching competitor limitations ('what's wrong with Knightscope?' 'what do people complain about with Cobalt?') never encounter Undaunted as an alternative, because there is no 'why Undaunted instead' content. The 70% win rate when visible (7/10 high-intent queries overall) confirms the product would win these evaluations — the structural gap is that Undaunted's name never enters the conversation when buyers build their evaluation lists.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_045, und_051, und_056, und_059, und_066, und_067, und_103, und_104, und_105, und_106, und_107, und_109, und_115, und_116, und_119, und_120, und_125
“We keep getting hit with overnight theft despite having cameras — best autonomous security robots for commercial properties?”
“Best robotic security solutions for replacing overnight guard shifts at commercial properties”
“Common complaints about Cobalt Robotics from property management teams”
Blueprint
  • On-Domain: Rewrite core product landing pages with buyer-outcome H1s that match Shortlisting query patterns: 'Best Autonomous Security Patrol for Commercial Properties' and 'Robotic Security for Construction Sites & Outdoor Perimeters' — replacing marketing taglines with descriptive headings that signal category relevance to AI platforms.
  • On-Domain: Create 'Why Buyers Switch to Undaunted' pages for each major competitor (Knightscope, Cobalt, Asylon) — addressing the Validation queries where buyers research competitor complaints. Frame as operational guidance: 'If you're experiencing [Knightscope limitation], here's what Undaunted does differently.'
  • On-Domain: Publish outcome-focused case content: 'How a VP of Operations at [Industry] Company Replaced Their Guard Program with Undaunted' — stories that answer the Shortlisting question 'who uses this and for what?' with real-world evidence.
  • On-Domain: Ensure the 'About Us' / why-undaunted page is fully functional and resolves the broken_about_us_link (L1 fix), as entity-resolution content is a prerequisite for appearing in vendor-Shortlisting responses.
  • Off-Domain: List Undaunted in physical security category directories (G2, Capterra, Gartner Peer Insights for security software) to establish third-party category placement that AI platforms use to build 'best X vendors' lists.
  • Off-Domain: Pursue third-party reviews and coverage in security industry analyst reports that are cited when AI platforms construct shortlists — ASIS, Frost & Sullivan, or IDC physical security coverage.
Platform Acuity

ChatGPT (medium): Shortlisting queries return 'best X for Y' lists that cite competitor landing pages and third-party directory listings. ChatGPT relies heavily on entity signals and third-party citations to include companies in shortlists — schema markup, directory listings, and review platform presence are critical inputs. Perplexity (high): Perplexity retrieves landing page content for Shortlisting queries, returning vendor descriptions with key capability claims. Pages with outcome-focused H1 headings matching buyer query language ('best autonomous security patrol for commercial properties') have high retrieval probability.

NIO #12: Guard Replacement Business Case & Decision-Stage Assets
Gap Type: Content Type Deficit — Undaunted has no ROI calculators, TCO models, case studies, or business case templates for guard replacement decisions. Invisible in all 12 consensus-creation and artifact-creation queries (0/12), while competitors with published ROI frameworks (Knightscope, RAD) win these deal-accelerating queries by default.
Critical

The consensus-creation and artifact-creation buying stages represent buyers who have selected a vendor category and now need internal justification materials to secure approval from a skeptical board or C-suite. COO-persona queries dominate this cluster — 'how do I justify replacing our guard service to the board?' 'what's the typical payback period?' 'build me a TCO model' — and Undaunted appears in none of the 12 queries (0/12). This is a pure content-type gap: no vendor dominates these queries because the content type is rare, giving Undaunted a clear opportunity to become the category authority on guard-to-robot transition business cases. Publishing ROI calculators, TCO frameworks, and board-presentation templates would capture buyers at the moment they are converting a vendor preference into a purchase approval — the highest-leverage point in the buying cycle.

Show query cluster, blueprint & platform acuity
Query Cluster
IDs: und_126, und_127, und_128, und_130, und_133, und_136, und_139, und_140, und_143, und_144, und_148, und_149
“ROI of switching from human guards to robotic security patrol — what's the typical payback period?”
“How to justify replacing our guard service with robotic security to a board that's skeptical about robots”
“Build a TCO model comparing robotic security patrol service vs 24/7 human guards over 3 years for 5 properties”
Blueprint
  • On-Domain: Publish a 'Robotic Security ROI Calculator' — an interactive or structured tool where buyers input their current guard costs, number of sites, and coverage hours to generate a 3-year savings estimate. This is the highest-impact single content asset for this buying stage.
  • On-Domain: Create a 'Guard-to-Robot Transition Guide' pillar page with: typical payback period ranges, what hidden costs to include in the TCO calculation (guard turnover, workers' comp, training, coverage gaps), and a structured 5-property 3-year Comparison table — directly addressing und_140.
  • On-Domain: Publish 2–3 customer case studies in the 'Before / During / After' format: what security problem the customer had, what the transition looked like, and what measurable outcomes resulted (cost saved per month, theft incidents reduced, guard headcount eliminated) — addressing und_126–und_128.
  • On-Domain: Create a 'Board Presentation Template' downloadable resource — a slide deck framework for an internal champion presenting robotic security to a skeptical C-suite, covering the financial case, risk profile, and technology readiness argument.
  • Off-Domain: Publish guard replacement ROI research in CFO and COO publications (CFO Magazine, Chief Executive Group) to establish Undaunted as the authority on the financial case for guard replacement — generating third-party citations that AI platforms surface when COOs ask AI about payback periods.
  • Off-Domain: Pursue press coverage of specific customer ROI outcomes (with permission) in security and real estate trade publications to create verifiable third-party evidence that ChatGPT cites in ROI-focused queries.
Platform Acuity

ChatGPT (high): ROI and business case queries return responses citing published research and vendor case studies. ChatGPT cites specific customer outcomes and published financial benchmarks — a case study page with specific savings data would be directly cited for queries like und_130 ('how much do companies save per site annually?'). Perplexity (high): Perplexity retrieves structured financial Comparison content and templates. An ROI calculator page with structured input/output framing, and a TCO Comparison table with labeled columns, would have the highest retrieval probability across both platforms for this buying stage.

Unified Priority Ranking

All recommendations across all three layers, ranked by commercial impact × implementation speed.

  • 1

    Main navigation links to broken /about-us page (404)

    The main site navigation displays an 'About Us' label, and the path https://www.getundaunted.com/about-us returns a 404 error. The actual about page lives at /why-undaunted. If the navigation links to /about-us rather than /why-undaunted, this is a broken link on every page of the site.

    Technical Fix · Engineering · Site-wide — the link appears in main navigation on all pages
  • 2

    No sitemap.xml found

    No sitemap.xml exists at https://www.getundaunted.com/sitemap.xml (returned 404). The site has only 12 discoverable commercial pages, and without a sitemap, AI crawlers must rely entirely on link-following from the homepage to discover content.

    Technical Fix · Engineering · All 12 site pages — affects discoverability and freshness signaling for the entire site
  • 3

    No visible dates on 11 of 12 pages

    Only 1 of 12 pages (the investment announcement dated February 23, 2026) displays a visible publication or last-updated date. All 11 remaining pages — including all product and industry pages — lack any detectable freshness signal. Without a sitemap providing lastmod timestamps, there is no freshness signal available for these pages.

    Technical Fix · Engineering · 11 of 12 pages — all pages except /undaunted-announces-investment
  • 4

    Autonomous Patrol Shortlisting Presence & Vendor Validation

    Undaunted has no content that establishes its autonomous patrol service in Shortlisting or vendor-Validation contexts. Invisible in all 17 Shortlisting and Validation queries for autonomous patrol (0/17 wins), while Knightscope, Asylon, and RAD are the default choices AI platforms recommend when buyers assemble vendor shortlists or research competitor limitations.

    New Content · Content · 17 queries affecting personas: VP of Operations, Chief Operating Officer, Regional Property Manager, Director of Security, Director of Construction / Project Executive
  • 5

    Buyer Evaluation Framework — Requirements Building Content

    Undaunted has no evaluation framework content: no RFP templates, vendor assessment checklists, or buyer's guides. Visibility is 0% across 15 requirements-building queries (0/15), and no vendor is consistently winning these queries — the gap is content type, not competitive loss.

    New Content · Content · 15 queries affecting personas: VP of Operations, Director of Construction / Project Executive, Director of Security, Chief Operating Officer, Regional Property Manager
  • 6

    Competitor Comparison Pages — Direct Matchup Content

    Undaunted has no competitor Comparison pages. Invisible or losing in 13 of 13 Comparison-stage autonomous patrol queries (0/13 wins), while Knightscope is cited in 9 of 13 (69.2%) of these queries as the frame of reference. Buyers in active vendor Comparison never encounter Undaunted because no 'Undaunted vs. X' content exists.

    New Content · Content · 13 queries affecting personas: VP of Operations, Director of Security, Director of Construction / Project Executive, Chief Operating Officer, Regional Property Manager
  • 7

    Early-Funnel Problem Identification Hub

    No content hub exists for buyers articulating security problems before they know robotic patrol is a solution. Undaunted is invisible in 0% of 13 problem-identification queries (0/13), while most responses surface general industry resources rather than any specific vendor.

    New Content · Content · 13 queries affecting personas: Director of Construction / Project Executive, VP of Operations, Director of Security, Regional Property Manager, Chief Operating Officer
  • 8

    Guard Replacement Business Case & Decision-Stage Assets

    Undaunted has no ROI calculators, TCO models, case studies, or business case templates for guard replacement decisions. Invisible in all 12 consensus-creation and artifact-creation queries (0/12), while competitors with published ROI frameworks (Knightscope, RAD) win these deal-accelerating queries by default.

    New Content · Content · 12 queries affecting personas: Chief Operating Officer, Director of Construction / Project Executive, VP of Operations, Director of Security, Regional Property Manager
  • 9

    Solution Education Hub — How Robotic Security Works

    No educational content series exists for buyers exploring robotic security as a solution category. Undaunted achieves 6.3% visibility across 16 solution-exploration queries (1/16), and that single appearance (und_023) is a positioning loss where a competitor wins the citation.

    New Content · Content · 16 queries affecting personas: Director of Security, Director of Construction / Project Executive, VP of Operations, Chief Operating Officer, Regional Property Manager
  • 10

    All-Terrain & Outdoor Detection Capability

    No dedicated outdoor terrain capability or thermal surveillance page exists on the Undaunted site. Undaunted is invisible in 0% of 16 Shortlisting, Comparison, Validation, consensus, and artifact queries focused on all-terrain and thermal detection capability (0/16), while SMP Robotics, Asylon, and Knightscope win these queries by virtue of having published capability content.

    New Content · Content · 16 queries affecting personas: Director of Construction / Project Executive, Director of Security, VP of Operations
  • 11

    Patrol Analytics & Accountability Dashboard

    Undaunted has no patrol analytics or reporting content, and no dedicated analytics product page. Invisible in 0% of 5 queries (0/5) where buyers explicitly demand patrol verification, route analytics, and reporting dashboards — while Cobalt Robotics and Knightscope win 4 of these 5 queries (80%) through documented analytics capabilities.

    New Content · Content · 5 queries affecting personas: VP of Operations, Regional Property Manager
  • 12

    Rapid Deployment & Construction Site Mobility

    No deployment process or construction site mobility content exists on the Undaunted site. Undaunted is invisible in 0% of 10 queries spanning Shortlisting, Comparison, Validation, consensus, and artifact stages focused on deployment speed and multi-site portability (0/10), while Asylon, RAD, and SMP Robotics win by default.

    New Content · Content · 10 queries affecting personas: Director of Construction / Project Executive, Chief Operating Officer, VP of Operations, Regional Property Manager
  • 13

    Remote Monitoring & Multi-Site Operations Hub

    No centralized remote monitoring operations content exists on the Undaunted site. Undaunted is invisible or losing in 10 of 10 Shortlisting, Comparison, Validation, consensus, and artifact queries about remote monitoring and multi-site patrol coordination (0/10 wins), while Knightscope, Cobalt Robotics, and RAD win by default with their monitoring center documentation.

    New Content · Content · 10 queries affecting personas: Regional Property Manager, Chief Operating Officer, VP of Operations, Director of Security
  • 14

    Homepage H1 is a marketing tagline, not a descriptive heading

    The homepage primary heading appears to be a marketing tagline ('Coverage at Half the Cost') rather than a descriptive heading that communicates what Undaunted is and does. Similarly, the pricing page H1 is 'Zero Maintenance. Simple Setup.' — a benefit statement rather than a heading that signals this is a pricing page.

    Technical Fix · Marketing · Homepage and pricing page — the two highest-value commercial pages
  • 15

    Schema markup status cannot be verified — manual check recommended

    Our analysis method returns rendered page content as markdown, which does not include JSON-LD structured data blocks. We cannot confirm whether the site implements schema markup (Organization, Product, LocalBusiness, FAQ, or Article schemas) on any of its 12 pages.

    Technical Fix · Engineering · All 12 commercially relevant pages
  • 16

    AI Detection Accuracy & Active Deterrence

    Undaunted has no AI detection accuracy content and no two-way audio deterrence documentation. Invisible or losing in all 10 queries spanning AI detection and verbal deterrence (0/10 wins), while competitors (Asylon, Cobalt Robotics, SMP Robotics) win by default with published detection specification content.

    New Content · Content · 10 queries affecting personas: Director of Security, Regional Property Manager, Chief Operating Officer, VP of Operations
  • 17

    Indoor Patrol & System Integration

    Undaunted has no indoor patrol content and no system integration documentation. Invisible in 0% of 6 queries spanning indoor patrol capability and security system integration (0/6), while Cobalt Robotics (indoor specialist) and Knightscope win these queries with dedicated indoor and integration content.

    New Content · Content · 6 queries affecting personas: Regional Property Manager, Director of Security, Chief Operating Officer, VP of Operations
  • 18

    Client-side rendering status should be verified

    The site appears to be built on Webflow (based on cookie consent patterns and page structure). While Webflow sites are typically server-rendered and AI-crawler-friendly, we cannot confirm CSR status from rendered output alone. All pages returned substantive text content, suggesting no major rendering issues.

    Technical Fix · Engineering · All pages with dynamic or interactive content
  • 19

    Meta descriptions and OG tags cannot be verified — manual check recommended

    Our analysis returns rendered text content, not raw HTML. We cannot confirm whether pages include meta descriptions, Open Graph tags, or Twitter Card markup.

    Technical Fix · Content · All 12 commercially relevant pages
  • 20

    No robots.txt file present — no explicit AI crawler policy

    No robots.txt file exists at https://www.getundaunted.com/robots.txt. All crawlers are implicitly allowed to access all pages, but the site has no explicit crawler access policy.

    Technical Fix · Engineering · Site-wide crawler access policy

Workstream Mapping

All three workstreams can start this week.

Engineering / DevOps

Layer 1 — Technical Fixes
Timeline: Days to 2 weeks
  • Main navigation links to broken /about-us page (404)
  • No sitemap.xml found
  • No visible dates on 11 of 12 pages
  • Homepage H1 is a marketing tagline, not a descriptive…

Content Team

Layer 2 — Content Optimization
Timeline: 2–6 weeks

Content Strategy

Layer 3 — NIOs + Off-Domain
Timeline: 1–3 months
  • Create a 'Commercial Property Security Challenges' hub page…
  • Publish a 'How Robotic Security Patrol Works' pillar page…
  • Publish a 'Robotic Security Vendor Evaluation Guide' — a…
  • Create a dedicated 'Outdoor & All-Terrain Security Patrol'…
  • Create a 'Rapid Deployment' capability page documenting:…

[Synthesis] The L1 fixes are prerequisites, not just improvements: publishing L3 content before resolving the missing sitemap means new pages may take months to be discovered by AI crawlers rather than days. Once L1 infrastructure is in place, the 12 L3 recommendations address all 143 invisible queries systematically — beginning with the three early-funnel content hubs (NIOs 001–003) that establish Undaunted in buyer consideration sets, and culminating with the business case and Comparison content (NIOs 010–012) that convert consideration into purchase decisions.

Methodology
Audit Methodology

Query Construction

150 queries constructed from persona × buying job × feature focus × pain point matrix
Every query carries four metadata fields assigned at creation time
High-intent jobs (Shortlisting + Comparison + Validation): 54% of queries (81 of 150)
Note: 150 queries across full buying journey.

Personas

VP of Operations — VP of Operations · Decision Maker
Director of Construction / Project Executive — Director of Construction / Project Executive · Evaluator
Chief Operating Officer — Chief Operating Officer · Decision Maker
Director of Security — Director of Security · Decision Maker
Regional Property Manager — Regional Property Manager · Evaluator

Buying Jobs Framework

8 non-linear buying jobs: Artifact Creation → Comparison → Consensus Creation → Problem Identification → Requirements Building → Shortlisting → Solution Exploration → Validation
High-intent jobs (Shortlisting + Comparison + Validation): 54% of queries (81 of 150)

Competitive Set

Primary: Knightscope, Cobalt Robotics, Asylon Robotics, Robotic Assistance Devices, SMP Robotics
Secondary: LiveView Technologies, Prosegur Security, ECAM
Surprise: Daxbot — flagged for review

Platforms & Scoring

Platforms: ChatGPT + Perplexity
Visibility: Binary — does the client appear in the response?
Win rate: Of visible queries, is the client the primary recommendation?

Cross-Platform Counting (Union Method)

When a query is run on multiple platforms, union logic is applied: a query counts as “visible” if the client appears on any platform, not each platform separately.
Winner resolution: When platforms disagree on the winner, majority vote is used. Vendor names are preferred over meta-values (e.g. “no clear winner”). True ties resolve to “no clear winner.”
Share of Voice: Each entity is counted once per query across platforms (union dedup), preventing double-counting when both platforms mention the same company.
This approach ensures headline metrics reflect real buyer-query outcomes rather than inflated per-platform counts.

Terminology

Mentions: Query-level visibility count. A company receives one mention per query where it appears in any platform response (union-deduped). This is the numerator for Share of Voice.
Unique Pages Cited: Count of distinct client page URLs cited across all platform responses, after URL normalization (stripping tracking parameters). The footer total in the Citation section uses this measure.
Citation Instances (Top Cited Domains): Raw count of citation occurrences per domain across all responses. A single domain can accumulate multiple citation instances from different queries and platforms. The Top Cited Domains table uses this measure.