Cognite

1.1 Identity and Business Model

Cognite is a B2B industrial AI and data platform company that enables asset-intensive industries—primarily oil and gas, energy, and process manufacturing—to liberate siloed operational data, contextualize it, and deploy AI-driven solutions at enterprise scale. The company's platform is anchored by two core products: Cognite Data Fusion (CDF), an industrial DataOps platform that creates and maintains an Industrial Knowledge Graph across heterogeneous OT and IT systems, and Cognite Atlas AI, a low-code industrial agent workbench launched in June 2024 that allows non-developers to build and deploy AI agents on top of the knowledge graph. In May 2026, the company added Cognite Flows as the "action layer" of its platform, designed to serve frontline workers with unified, AI-driven workflow experiences. Cognite's business model is software-as-a-service (SaaS): customers pay subscription fees for platform access, with professional services and implementation support layered on top. The company targets enterprise-scale industrial operators with complex multi-site environments, positioning CDF as the data foundation that unifies decades of operational data before AI agents and applications can deliver value. As of FY2025, approximately 40% of revenue came from customers outside the oil and gas sector, signaling meaningful diversification into manufacturing, chemicals, pharmaceuticals, and renewables. The company is headquartered at 40 E Rio Salado Pkwy, Suite 900, Tempe, AZ 85281, having recently relocated its global headquarters from Oslo, Norway to be proximate to Arizona's high-tech and semiconductor manufacturing ecosystem. Cognite maintains regional offices in Austin TX, Houston TX, Lysaker (Oslo) Norway, Tokyo Japan, and Bengaluru India. [CO001, CO002, CO004, CO010, CO015, CO016]

1.2 Founding, Leadership, and Governance

Cognite's origin story is intertwined with Norwegian conglomerate Aker. The official Cognite about-us page states that the company has been on its mission "since 2016," while its May 2021 press release explicitly states "Cognite was founded in 2017" and references "four years" since founding. CNBC's May 2026 Disruptor 50 profile states Cognite was "formed in 2016 in Oslo as a strategic partner with European oil company Aker BP." This dating conflict—2016 vs. 2017—is material and unresolved in public records. The founding thesis was to transform heavy-asset industries through industrial data liberation, with Aker BP as both incubator and anchor customer. John Markus Lervik, a Norwegian serial entrepreneur, co-founded the company alongside Geir Engdahl, a former Google software engineer, who serves as CTO AI and co-founder. Both remain active in the company. In 2022, Girish Rishi was named CEO after John Markus Lervik stepped aside to become Chief Strategy and Development Officer. Rishi, who also serves as Chairman of the Board, brings U.S. enterprise software experience as former CEO of Blue Yonder (a Panasonic-acquired supply chain software company) and prior executive leadership at Motorola and Tyco. This CEO transition is a notable governance event: the founder-to-professional-CEO handoff often signals a pivot toward commercialization and U.S. market expansion, consistent with the subsequent relocation of headquarters to Arizona. Cognite's current senior leadership team includes Jeff Coulter (CFO), Chirayu Shah (Chief Product Officer), Liat Berger (Chief Human Resources Officer), Paul Grenet (President, Global Field Operations), Sandy Joung (Chief Marketing Officer), and Håkon Bjerke (Chief Strategy Officer). Regional presidents cover Americas (Bill Hendricks), EMEA (Petteri Vainikka), Japan (Ryoichi Egawa), North and Southeast Asia (TS Park), and India (Guru Ananthanarayanan). Cognite employs professionals from more than 60 countries. Board oversight includes Aker ASA-affiliated directors and Jake Reynolds (TCV General Partner, joined board in 2021). Key-person dependence on Girish Rishi and Geir Engdahl is assessed as high given their strategic and technical leadership roles respectively. [CO002, CO003, CO004, CO005, CO006, CO007]

1.3 Funding, Capital, and Investors

Cognite's most recent publicly confirmed funding milestone is a $150 million round from growth equity firm TCV in May 2021, which valued the company at $1.6 billion and conferred unicorn status. TCV's Jake Reynolds joined the board as part of that transaction. Cognite's 2021 press release noted the company was growing rapidly with "over 500 employees" at that time and characterized the round as "one of the largest funding rounds for a SaaS company in Europe." Prior to the 2021 TCV round, Accel had made an earlier growth equity investment (amount undisclosed in public sources), and the 2021 press release describes TCV as complementing "earlier funding by global venture capital firm Accel." TCV subsequently led an additional investment round. TCV's press release page bears the title "TCV Leads $300M Investment in Cognite," and a June 2024 TechCrunch URL (since returned 404) referenced "cognite-raises-300m." CNBC's May 2026 Disruptor 50 listing reports "$300 million" in funding, which may refer to the 2024 round alone or total disclosed capital. The exact post-2024-round valuation and complete capital history are not publicly disclosed, as Cognite is a private company. The $1.6B valuation from May 2021 is the last publicly confirmed figure. Aker ASA, the investment vehicle of Norwegian billionaire Kjell Inge Roekke, is Cognite's majority shareholder and serves as its corporate parent. An independently run affiliate structure governs this relationship. In February 2022, Saudi Aramco purchased a 7.4% stake in Cognite from Aker BP (a distinct entity from Aker ASA), adding a strategically significant industrial customer as an equity holder. Saudi Aramco is also one of Cognite's largest customers, creating an alignment of financial interest and potential influence on product roadmap and customer access that warrants ongoing diligence. Board representation includes Aker ASA-affiliated directors, TCV's Jake Reynolds, and potentially Aramco representation (not confirmed in public sources). [CO011, CO012, CO013, CO014, CO017, CO018]

1.4 Scale, Metrics, and Geographic Footprint

Cognite's FY2025 was described as a "record-breaking year" in the company's January 2026 press release. Annual revenue exceeded $170 million for the first time, against a backdrop of 36% annual recurring revenue (ARR) bookings growth year-over-year. Net revenue retention (NRR) for newer customers reached 150%, indicating strong existing-customer expansion. Company headcount surpassed 800 employees globally by early 2026, representing 21% growth during FY2025. These metrics are preliminary and unaudited, as explicitly stated in Cognite's own January 2026 announcement. Cognite's Atlas AI product drove significant acceleration in FY2025: the number of Atlas AI customers grew nearly 700% year-over-year; Atlas AI was used by more than half of the customer base (59% by ARR); and more than 70% of 2025 new bookings included Atlas AI. Monthly active users (MAU) across the platform grew 26% in 2025. The Industrial Knowledge Graph underlying all Cognite products exceeded 80 trillion time series data points in context by end of 2025—a 48% increase—and doubled the number of data modeling instances. AI token usage grew nearly 500% year-over-year. Approximately 40% of FY2025 revenue came from customers outside oil and gas, representing meaningful diversification from the company's historical O&G concentration. Key named customers include bp, Saudi Aramco, TotalEnergies, Aker BP, Cosmo Energy (all three Japanese refineries), NOVA Chemicals, B. Braun (medical technology), and Idemitsu Kosan. The Forrester analyst blog from 2025 documents NOVA Chemicals deploying CDF in one plant with plans to expand to 11, TotalEnergies deploying across 39 assets over three years, and Cosmo Energy storing 250 billion data points across three refineries—providing evidence of enterprise-scale deployments. Gross margin, burn rate, and EBITDA remain undisclosed for a private company, and the absolute customer count is not published. [CO020, CO021, CO022, CO023, CO024, CO025]

1.5 Milestones and Adverse Events

Cognite's ten-year trajectory reflects three phases: a founding/incubation phase (2016/2017–2020) in which the company developed its platform under the Aker umbrella with Aker BP as anchor customer; a scale/unicorn phase (2021–2023) in which TCV investment, Microsoft partnership, and international customer expansion drove rapid growth; and a U.S.-first AI phase (2024–2026) characterized by the Atlas AI launch, CEO-led U.S. reorientation, and Flows platform expansion. The CEO transition in 2022—when co-founder John Markus Lervik stepped aside for Girish Rishi—is the most significant governance event to date. While framed positively in public communications, founder-to-CEO transitions carry execution risk and potential cultural disruption. No adverse events, regulatory actions, or significant litigation are documented in public sources reviewed during this research run. The UpGuard vendor risk report for Cognite is active as of May 2026 and monitors security posture across website, email, network, and brand dimensions; the detailed rating is paywalled, but the existence of the report and the "Start a free trial for full access" barrier suggests independent third-party security scrutiny that enterprise customers and investors should review. Cognite operates critical infrastructure software for energy companies, making security posture a material diligence item. No public incidents, data breaches, or material outages were identified in sources reviewed; absence of evidence is not evidence of absence, and StatusGator maintains an active monitoring page for Cognite service availability. CNBC Disruptor 50 listing incorrectly names Girish Rishi as a "founder"—he joined as CEO in 2022; this factual error in an otherwise relied-upon source is noted for diligence hygiene. [CO002, CO003, CO004, CO005, CO007, CO030]

1.6 Exhibits

2.1 Market Boundary and Definition

Cognite's primary market is the Industrial DataOps and Industrial AI platform segment: software platforms that ingest, contextualize, and operationalize operational technology (OT) and information technology (IT) data within asset-intensive industries to enable AI-driven workflow automation and decision support. This segment sits inside the broader Industrial Internet of Things (IIoT) umbrella but is meaningfully narrower: it excludes commodity sensor hardware, general-purpose cloud storage, SCADA/DCS control systems, and enterprise resource planning software. Status-quo substitutes include legacy historian databases (notably OSIsoft PI, now part of AVEVA/Schneider Electric), point analytics tools, OEM-bundled analytics in industrial automation suites (Siemens, GE Vernova, Honeywell, AspenTech), and spreadsheet-based engineering workflows. The category is also distinguished from general-purpose enterprise AI platforms (Palantir AIP, C3.ai) by its primary dedication to OT/IT convergence, real-time sensor data, and industrial knowledge graphs rather than business process AI. Cognite serves three primary end-market verticals: oil and gas (approximately 60% of FY2025 revenue based on implied reverse calculation of the "40% non-O&G" disclosure), process manufacturing/chemicals/pharmaceuticals, and power and renewables. These verticals share the defining characteristics of capital intensity, multi-site operations, heterogeneous legacy data infrastructure, and strong economic and safety incentives for operational optimization. Adjacencies that Cognite is actively exploiting include the hyperscaler data platform market (Databricks, Snowflake) and agentic AI tooling (Microsoft Copilot Studio, AWS Bedrock)—both competitive and complementary. The IOGP (International Association of Oil and Gas Producers) represents the primary industry body organizing operator standards and digital transformation initiatives globally. The term "Industrial DataOps" was adopted broadly post-2020 and was first validated as a distinct analyst category by IDC with the inaugural MarketScape assessment in early 2026. Prior to this, the relevant software category was variously called "industrial analytics," "operational intelligence," or "IIoT platforms"—making historical sizing comparisons methodologically unreliable. Cognite competes against both specialized pure-play vendors (Seeq, Aspentech Inmation/Mtell) and hyperscaler-native services (AWS IoT SiteWise, Azure Digital Twins, Google Cloud Manufacturing Data Engine) that offer commodity data aggregation at cloud pricing. [CM001, CM002, CM003, CM004, CM005, CM006]

2.2 Market Sizing: Multiple Lenses

No publicly accessible analyst report provides a directly bounded TAM for "Industrial DataOps platforms" or "Industrial AI platforms for asset-intensive industries" as Cognite defines its market. The IDC MarketScape Worldwide Industrial DataOps Platforms 2026 (Doc #US53013025) is the most authoritative reference but is sold at paywall pricing and only available in excerpt form via Cognite's own press release. Two independent sources on the broader IIoT market provide sizing anchors with materially different projections. Grand View Research (GVR) estimates the global IIoT market at $483.16 billion in 2024, projected to reach $1,693.44 billion by 2030 at a CAGR of 23.3% (2025–2030). Precedence Research estimates the same market at $514.39 billion in 2025, projecting $2,430.21 billion by 2035 at a CAGR of 16.8% (2026–2035). The two estimates conflict on both the 2024/2025 base ($483B vs $514B) and the CAGR (23.3% vs 16.8%), reflecting different market boundary definitions—GVR likely includes more industrial hardware/connectivity revenue while Precedence may weight software/services differently. Neither report provides a dedicated sub-segment for Industrial DataOps platforms. A bottom-up sizing lens using Cognite's disclosed metrics suggests a constrained addressable market: Cognite at $170M revenue is likely the second or third largest pure-play Industrial DataOps vendor globally (behind AVEVA/OSIsoft, which is now part of Schneider Electric at enterprise scale), implying a pure-play Industrial DataOps software market of perhaps $500M–$2B in annual contract value—orders of magnitude smaller than the IIoT TAM. The C3.ai SEC 10-K filing (fiscal 2025, filed July 2025) shows C3.ai at approximately $252M revenue, providing a public analog for enterprise industrial AI scale and confirming the market is real but not yet at multi-billion revenue concentration. Manufacturing accounts for the largest IIoT end-use segment (~28.7% in 2025 per Precedence Research), exceeding oil and gas alone—this is relevant as Cognite is actively expanding beyond its O&G base. [CM011, CM012, CM013, CM014, CM015, CM016]

2.3 Buyer, User, and Payer Segmentation

Cognite's customer base is skewed heavily toward large industrial enterprises—companies with significant multi-site OT infrastructure, capital budgets exceeding $1B, and existing investments in industrial automation. The archetypal buyer is an integrated oil and gas major, a large chemicals or process manufacturer, or a national energy company. Cognite's named customers include bp, Saudi Aramco, TotalEnergies, Aker BP, Cosmo Energy, TotalEnergies, NOVA Chemicals, B. Braun, and Idemitsu Kosan—all large enterprises with complex, multi-site operational environments where the economic case for data contextualization and AI is strongest. Budget ownership for Industrial DataOps sits at the intersection of IT and Operations: the Chief Digital Officer (CDO), VP of Digital, or VP of Engineering typically champions the investment, while procurement/finance controls the multi-year SaaS agreement mechanics. This means Cognite competes for both discretionary "digital transformation" budgets and core operational IT spend. Adoption triggers are consistently: (1) a digital transformation mandate from the C-suite driven by operational efficiency targets, (2) a new major capital project requiring modern data infrastructure, or (3) pressure from ESG reporting requirements demanding operational visibility. The land-and-expand pattern is central to Cognite's commercial model: NOVA Chemicals deployed CDF in one plant, then planned expansion to 11 plants within the year; TotalEnergies is deploying across 39 upstream assets over three years; Cosmo Energy deployed across all three Japanese refineries. This pattern implies that the initial sale is often a limited-scope pilot (one site, one use case), and the majority of revenue growth comes from expansion. Cognite's 150% NRR for newer customers validates this model commercially. The user personas span data scientists, process/plant engineers, IT/OT integration architects, and—with Atlas AI—non-technical frontline workers. The payer authority typically sits with a corporate-level budget rather than individual plant managers, which means enterprise procurement cycles are long (often 12–24 months for initial deals) but multi-year contracts are the norm. [CM021, CM022, CM023, CM024, CM025, CM026]

2.4 Growth Drivers and Adoption Constraints

The most significant near-term growth driver for Cognite's market is the convergence of the generative AI adoption wave with the decades-long backlog of industrial data integration work. Cognite's Atlas AI customers grew nearly 700% year-over-year in FY2025, demonstrating that AI agent capabilities have crossed a commercial threshold for industrial buyers who previously waited for the technology to mature. This AI demand driver is further amplified by ESG and decarbonization mandates: companies like TotalEnergies are deploying industrial data platforms explicitly to improve health-and-safety monitoring and reduce emissions tracking burden alongside production optimization. Structural tailwinds include industrial workforce aging and labor shortages: as experienced engineers retire, industrial operators face a "knowledge transfer" crisis that AI-powered knowledge graphs directly address. Cognite's Atlas AI positioning—enabling AI agents to conduct root-cause analysis that previously took weeks—maps directly to this constraint. Cloud migration in OT environments (historically lagged IT by 10–15 years) is now accelerating as vendors like AWS (IoT SiteWise), Azure (Digital Twins), and Siemens (Industrial Edge) normalize cloud-connected factory and field operations. This normalization lowers Cognite's infrastructure friction and expands the addressable deployment base. Adoption constraints are real and material. Forrester's analyst characterizes industrial AI deployments as inherently "slow, hard, and expensive" for enterprise-scale companies—noting that the shiny AI application is "a tiny little piece of the iceberg, resting atop a mountain of data operations groundwork." Switching costs are severe: a multi-year Industrial DataOps implementation integrates hundreds of OT/IT data sources, creating deep vendor lock-in once a knowledge graph is established. Budget cycles in oil and gas and process manufacturing are long—major IT/OT integration projects commonly span 18–36 months from approval to production. OT security and data sovereignty requirements add compliance overhead that extends sales cycles and demands dedicated customer success resources. Incumbent vendor competition from Siemens, GE Vernova, Honeywell, and AspenTech—each offering bundled analytics in their automation suites—also creates switching inertia at large accounts already using these vendors' hardware. [CM031, CM032, CM033, CM034, CM035, CM036]

2.5 Diligence Gaps and Contradictory Estimates

The most significant analytical gap in this chapter is the absence of a public, independently produced, unpaywalled estimate for the Industrial DataOps platform sub-segment. IDC's 2026 MarketScape (paywalled), Verdantix's two 2025 leadership reports (paywalled), and Gartner's relevant Market Guide (blocked/inaccessible at time of research) all confirm that the category exists and that Cognite is a leader—but none of these sources provide an extractable market size figure. Diligence buyers should acquire the full IDC and Verdantix reports before accepting management's implied market framing. The two available public estimates—GVR ($483B in 2024, CAGR 23.3%) and Precedence Research ($514B in 2025, CAGR 16.8%)—are methodologically inconsistent and cannot be reconciled into a single authoritative figure. Both are for the broad IIoT market, not Industrial DataOps specifically, and both acknowledge that the market boundary is still evolving as software and hardware components are counted differently across publications. This contradiction is preserved rather than resolved; neither estimate should be used as a proxy for Cognite's SAM without significant adjustment. Cognite's stated Moonshot goal ($100B in cumulative customer value by 2035) is aspirational and not a market size estimate—it reflects management's internal confidence but does not map to a verifiable TAM. The market penetration rate for Industrial DataOps in oil and gas (the dominant vertical) is not publicly quantifiable: there are approximately 50+ major oil and gas operators globally with the scale to be Cognite customers, but the fraction currently using any Industrial DataOps platform is unknown from public sources. [CM041, CM042, CM043, CM044, CM045]

2.6 Exhibits

3.1 Competitive Landscape Overview

Cognite operates in a multi-layer competitive landscape with no single direct twin. The primary competitive categories are: (1) legacy OT data historians and industrial software incumbents—AVEVA PI System (Schneider Electric), AspenTech (Emerson), and Siemens Industrial Edge—whose entrenched installed bases create high switching friction; (2) enterprise AI horizontal platforms—C3.ai, Palantir AIP—that compete for the same enterprise AI budget at large industrial operators; (3) cloud hyperscaler industrial services— AWS IoT SiteWise, Azure Digital Twins—that commoditize the data ingestion and connectivity layer; (4) adjacent Industrial IoT vendors—Seeq (analytics), PTC ThingWorx/Velotic (IIoT connectivity), Hexagon (asset lifecycle/3D spatial intelligence), and GE Vernova (power-focused)—that overlap on specific capabilities; and (5) status-quo and internal-build alternatives—OSI PI historian plus manual spreadsheet analysis plus siloed SCADA/DCS systems, or custom cloud data-lake architectures built on Snowflake or Databricks. No single competitor replicates CDF's full stack of data ingestion, knowledge graph contextualization, and low-code AI agent workbench for industrial operations. The competitive threat landscape is therefore asymmetric: incumbents hold distribution power, hyperscalers hold infrastructure leverage, horizontal AI platforms hold enterprise sales relationships, and Cognite holds domain depth and AI-native architecture. [CP001, CP006, CP009, CP014, CP021, CP025]

3.2 Incumbent and Direct Peer Profiles

AVEVA PI System—now owned by Schneider Electric following a ~$9.5 billion acquisition completed in January 2023—is the dominant industrial time-series data historian with a 30-plus-year installed base across oil and gas, power, and process industries. The PI System's competitive advantage is its pervasive deployment footprint: tens of thousands of historian tags at large industrial facilities create deep operational dependency and multi-year switching friction. AVEVA has launched AVEVA CONNECT, a cloud-native data management platform, as its next-generation offering designed to compete with modern cloud alternatives; however, it occupies a different architectural philosophy from CDF's knowledge graph approach. AspenTech, acquired by Emerson Electric in a transaction valued at approximately $11 billion in 2022, targets chemical engineering, refining, power generation, and asset performance management with its aspenONE simulation suite and Aspen Mtell predictive maintenance product. The Emerson/AspenTech combination bundles OT control hardware and software under one enterprise roof, creating a channel advantage over standalone software vendors. AspenTech's focus on process simulation rather than operational data contextualization limits direct overlap with CDF's core use cases. Seeq positions itself as an advanced analytics, machine learning, and AI platform for industrial time-series data, operating as a data-source agnostic layer that can run on top of PI System, CDF, or other historians. Unlike Cognite's full-stack approach, Seeq addresses the analytics and insight layer rather than data ingestion or knowledge graph construction, making it complementary to—but in some deployments competitive with—CDF's analytical capabilities. C3.ai (NYSE: AI) offers a horizontal enterprise AI application platform with marquee industrial customers including Shell, Baker Hughes, Koch Industries, and the U.S. Air Force, but its horizontal scope and reliance on customer-provided data infrastructure contrast with Cognite's vertically integrated approach. Palantir AIP targets large-enterprise and government/defense organizations with a high-touch delivery model and deal sizes that typically exceed standard industrial SaaS contracts. [CP002, CP003, CP004, CP006, CP007, CP008]

3.3 Capability, Pricing, and GTM Comparison

Cognite's key differentiators across the competitor landscape are its Industrial Knowledge Graph architecture—which contextualizes relationships between assets, time series, documents, and 3D models in a unified semantic layer—its Atlas AI low-code agent workbench, and its cloud-native open-API design. No incumbent historian (AVEVA PI, AspenTech) or horizontal AI platform (C3.ai, Palantir) replicates this combination. AWS IoT SiteWise and Azure Digital Twins address the infrastructure layer but lack Cognite's domain-specific data modeling and pre-built industrial contextualization. Pricing across this landscape is largely opaque for private-market participants. Cognite operates on a SaaS subscription model with per-asset or per-data-volume pricing components, implementation services, and professional-services layers, but exact list pricing is not publicly disclosed. AVEVA PI System is sold via perpetual plus maintenance and subscription models; AspenTech uses a token-based subscription model for its aspenONE engineering suite. C3.ai uses an application-specific subscription model; Palantir's engagements are large-scale custom contracts often in the $10 million-plus range. AWS IoT SiteWise and Azure Digital Twins use cloud consumption-based pricing (per-message, per-query, per-tag) that creates a lower entry point at small scale but potentially high costs at enterprise scale. Cognite's GTM strategy emphasizes executive-level enterprise sales, Microsoft co-sell partnerships, and strategic integrations with Databricks and Snowflake for analytics and data lake interoperability. Its key distribution moat versus incumbents is the Microsoft partner ecosystem: named Microsoft Energy and Resources Partner of the Year for three consecutive years through 2024, Cognite benefits from Microsoft's global enterprise account relationships. This partially offsets the OEM/bundled channel advantages held by AVEVA (inside Schneider Electric's broader industrial automation portfolio) and AspenTech (inside Emerson's process automation suite). [CP023, CP024, CP034, CP035, CP040]

3.4 Switching Costs, Lock-in, and Multi-homing

Cognite CDF creates multi-layered switching costs for customers who complete full deployments. The Industrial Knowledge Graph schema, once populated with thousands of asset models, contextual relationships, and time-series tag mappings across an enterprise's facilities, represents a significant proprietary data structure investment. Custom integrations with OT source systems (DCS, SCADA, historians), API applications built on CDF, and trained Atlas AI agents are all specific to CDF's data model and API surface, increasing the operational cost of migration. Multi-homing—running CDF alongside a competing platform—is technically feasible but operationally complex and is rarely observed beyond pilot/evaluation contexts. The migration path from a legacy AVEVA PI System deployment to Cognite CDF requires rebuilding OT source connectors, remapping the PI tag hierarchy to CDF's knowledge graph schema, and retraining operational users who rely on PI-based tools for day-to-day operations. This transition typically involves multi-month implementation efforts with system integrators, creating a meaningful barrier that benefits the incumbent (PI System) even when buyers prefer CDF's capabilities. Conversely, the status-quo alternative—the combination of PI historian, manual data exports, and spreadsheet-based analysis—has the lowest switching cost profile for initial replacement but the highest analyst productivity cost, and represents Cognite's most common greenfield opportunity. The internal-build alternative (custom Snowflake/Databricks data lake plus Python/ML pipelines) is a growing competitive threat among industrial IT teams with sufficient data engineering resources, offering zero license cost at the price of sustained internal development investment. [CP033, CP031, CP032, CP039]

3.5 Moat Durability, Commoditization Risk, and Adverse Evidence

Cognite's most durable competitive advantage is the combination of industrial domain expertise, the Industrial Knowledge Graph's semantic data architecture, and the Atlas AI agent workbench that sits atop this foundation. However, three structural threats warrant ongoing diligence: first, AVEVA's entrenched PI System installed base and Emerson/AspenTech's bundled OT+software channel represent distribution advantages that cannot be overcome quickly by product superiority alone—Cognite must demonstrate multi-million-dollar ROI per deployment to displace incumbent contracts; second, AWS and Azure are actively expanding industrial IoT service portfolios, potentially commoditizing the lower layers of CDF's value stack over a 3-to-5-year horizon as cloud-native connectivity and storage become undifferentiated; and third, Cognite's last publicly confirmed valuation of $1.6 billion (May 2021) has not been updated, while publicly traded competitors (C3.ai, Palantir) have full market-cap transparency, and well-funded enterprise divisions of Schneider Electric and Emerson have virtually unlimited investment capacity, meaning Cognite's competitive capitalization relative to peers is opaque. Adverse evidence from Gartner Peer Insights reviews of the industrial IoT platform market indicates that enterprise buyers cite implementation complexity, total cost of ownership, and professional services dependency as top concerns across the category—concerns that apply directly to Cognite's high-touch deployment model. The Gartner reviews also reflect a buyer landscape that is genuinely competitive across multiple vendors, with no single platform earning overwhelming user endorsement, underscoring execution risk for all vendors including Cognite. No specific documented competitive losses by Cognite to named competitors have been identified in public sources reviewed through May 2026, though Cognite does not disclose win/loss data publicly. [CP033, CP036, CP037, CP038, CP011, CP018]

3.6 Exhibits

4.1 Revenue Model and Pricing

Cognite operates a software-as-a-service (SaaS) subscription model anchored by two principal platforms: Cognite Data Fusion (CDF), the industrial DataOps engine, and Cognite Atlas AI, the low-code industrial agent workbench launched in June 2024. Both are sold to enterprise-scale industrial operators under multi-year subscription agreements, with professional services and Cognite Success Tracks layered on top. The third platform layer—Cognite Flows, launched in May 2026—extends the action layer of the platform to frontline workers and is expected to deepen per-seat or per-workflow monetization over time. Cognite has not published list pricing for any of its products as of the May 2026 run date. Industry convention for industrial DataOps and AI platforms at this scale suggests total contract value (TCV) for enterprise accounts typically ranges from several hundred thousand to multi-million dollars per year, driven by data volume, number of connected assets, seat count, and professional-services scope. The TotalEnergies 3-year partnership to deploy CDF across all upstream assets globally is consistent with a multi-million-dollar annual contract value (ACV); however, no specific dollar figure has been disclosed. Similarly, the B. Braun engagement and Celanese deployment (via NVIDIA NV-Tesseract) suggest mid-to-large enterprise ACV, but specific terms are not public. Revenue mix: In FY2025, approximately 60% of revenue derived from oil and gas customers and ~40% from outside oil and gas (manufacturing, chemicals, pharma, renewables, utilities). The Atlas AI product accounted for more than 70% of new bookings in 2025, suggesting a platform deepening dynamic where existing CDF customers layer on Atlas AI subscriptions, driving NRR above 100%. The company's "Moonshot" mission—to deliver $100 billion in measurable customer value by 2035—provides a publicly articulated North Star that frames both its GTM narrative and its premium pricing posture. The Norwegian statutory annual accounts filed with Brønnøysundregistrene (Cognite AS, org. 918274758) recorded FY2024 revenue of approximately NOK 1.67 billion (~$158M USD at the 2024 average NOK/USD rate of ~10.55). This represents the Norwegian legal entity only; consolidated group revenue including US, Japan, and India subsidiaries would differ. It is consistent with—and provides a FY2024 floor for—the reported >$170M FY2025 global revenue. [CI001, CI002, CI003, CI004, CI005, CI006]

4.2 GTM Motion and Sales Efficiency

Cognite sells exclusively through a direct enterprise-sales motion. Paul Grenet carries the title of President and Chief Revenue Officer, indicating a combined field-operations and revenue-ownership structure. Regional presidents cover Americas (Bill Hendricks), EMEA (Petteri Vainikka), Japan (Ryoichi Egawa), North and Southeast Asia (TS Park), and India (Guru Ananthanarayanan), establishing a full global direct-sales footprint consistent with enterprise software at $100M+ ARR. Sales cycles in industrial AI and DataOps are long—typically 9–18 months for greenfield enterprise deployments—due to OT-integration complexity, procurement governance at energy majors, and the need for proof-of-concept phases. Cognite's strategy of offering a "Cognite Quick Start" on-ramp and embedding Atlas AI into 72% of customer ARR (as of Q1 2026) is designed to accelerate initial land and compress expansion timelines. The BusinessWire Flows announcement cited that a global pharmaceutical customer achieved 30X faster time-to-value and reduced UAT from 6–9 months to two months—these are public productivity benchmarks offered to support the pricing conversation. A key co-sell vector is the Microsoft Azure partnership (three-time Energy & Resources Partner of the Year, 2022–2024) which routes Cognite into Azure Marketplace commercial agreements and co-sell motions with Microsoft's enterprise field. The Snowflake and Databricks ecosystem partnerships (2026) extend channel reach into data-cloud-centric procurement workflows. The NVIDIA NV-Tesseract integration (March 2026) adds an AI-infrastructure co-sell dimension. Customer acquisition cost (CAC), sales-cycle length, magic number, and payback period are not publicly disclosed. Proxies suggest high CAC consistent with long-cycle enterprise sales compensated by high ACV and strong land-and-expand economics (150% NRR for newer customers). Customer count is not disclosed publicly. [CI009, CI010, CI011, CI012, CI013, CI014]

4.3 Cost Structure and Unit Economics

Cognite is a cloud-native software company deployed primarily on Microsoft Azure. Its cost structure is dominated by research and development (platform engineering, AI model development, knowledge graph infrastructure), global sales and marketing (direct enterprise field force across 6 geographies, marketing in a niche industrial vertical), and general and administrative costs associated with a 800+ person global organization. Cloud infrastructure costs (Azure compute, storage, and AI inference) are material given the scale of the Industrial Knowledge Graph (85+ trillion time series data points as of Q1 2026) but are largely passed through as platform costs to customers or embedded in subscription pricing. Gross margin is not publicly disclosed. Industrial SaaS platforms with significant data-infrastructure components and embedded professional-services elements typically carry gross margins in the 60–75% range, lower than pure horizontal SaaS (70–85%) due to OT-integration services and per-customer data contextualization work. Cognite's "Success Tracks" offering and the implementation-intensive onboarding for multi-site enterprise customers suggest a material services component that would compress blended gross margin below a pure-software benchmark. The Norwegian statutory accounts (FY2024) disclose that Cognite AS reported operating expenses (implicitly ~NOK 2.08 billion against revenues of ~NOK 1.67 billion) resulting in a significant net loss of approximately NOK -547M (~-$52M USD). This is a partial picture—the Norwegian entity likely bears centralized R&D costs while some revenue and costs sit in US and other subsidiaries—but it confirms the magnitude of cash consumption at the entity level. A fully consolidated group loss could be larger or smaller depending on intercompany transfer pricing and non-Norwegian subsidiary performance. Unit economics at the customer level—LTV, CAC, payback, and net dollar retention by cohort—are fully undisclosed. The 150% NRR for "newer" customers is the only public unit-economics proxy and is a strong positive signal, but the definition of "newer" (vintage, cohort composition) is not specified and the full-base NRR is not stated. [CI015, CI016, CI017, CI018, CI019, CI020]

4.4 Capital Adequacy and Financing

Cognite's most recent public funding milestone is the ~$300M round led by TCV in June 2024, the title of which appears on the TCV press page ("TCV Leads $300M Investment in Cognite"). Combined with the $150M TCV round from May 2021 and an undisclosed earlier Accel investment, total known capital raised is at least ~$450M. The 2024 round valuation was not publicly disclosed. The last confirmed valuation of $1.6B is from the 2021 round—approximately five years stale by the May 2026 run date. CNBC Disruptor 50 (May 2026) cites $300M in total funding, which may reflect the 2024 round alone or represent their calculation of total publicly confirmed capital. The Company Overview chapter (chapter 1 of this report) contains the full round-by-round chronology and investor map; this chapter references it for background and mints only the financing claims needed to assess forward capital adequacy. Cash on hand, monthly burn rate, and runway are entirely undisclosed. The Norwegian statutory accounts (FY2024) provide the only available proxy: Cognite AS reported total assets of approximately NOK 2.26 billion (~$214M USD) and a net loss of approximately NOK 547M (~$52M USD) for FY2024 alone. If we assume the Norwegian entity holds the majority of group cash (which is not confirmed), the 2024 year-end balance sheet implied a meaningful cash cushion. However, the 2024 round (~$300M) closed in June 2024—it is unclear how much of that capital has been deployed by the May 2026 assessment date. At the FY2024 burn rate (~$52M/year for the Norwegian entity alone), a group-level burn rate of $60–90M/year is plausible, implying the 2024 round would provide roughly 3–5 years of runway—a reasonable cushion but not an indefinite horizon. Aker ASA, as majority shareholder, provides an implicit backstop: Aker's Net Asset Value exceeded NOK 100 billion in Q1 2026, its strongest quarter on record. Saudi Aramco's 7.4% equity stake (held since February 2022, purchased from Aker BP) adds a strategic investor with deep pockets but creates a governance dependency: Aramco is simultaneously Cognite's largest customer and a material equity holder, a structure that has historically created both access advantages and product-roadmap influence risk. [CI021, CI022, CI023, CI024, CI025, CI026]

4.5 Public Traction and Private Information Gaps

Cognite's public financial disclosure profile is consistent with a late-stage private company actively managing investor narrative: it releases selected ARR growth, NRR, and headcount metrics via press releases and analyst validations, while withholding gross margin, absolute ARR, customer count, EBITDA, and balance sheet data. The following traction metrics are company-reported and preliminary/unaudited, as explicitly stated in Cognite's January 2026 announcement. Confirmed public metrics (FY2025): >$170M annual revenue; 36% ARR bookings growth YoY; 150% NRR for newer customers; >800 employees globally (+21% YoY); 40% of revenue from non-O&G customers; >70% of new 2025 bookings included Atlas AI; 59% of customers by ARR using Atlas AI. Q1 2026 acceleration signals: ARR bookings +57% YoY; Atlas AI penetration 72% by ARR; knowledge graph 85T+ data points (+30% YoY); AI token usage +900% YoY; MAU +54% YoY. Norwegian statutory data (FY2024, Cognite AS entity): NOK 1.67B (~$158M USD) revenue; NOK -547M (~-$52M USD) net loss; NOK 2.26B (~$214M USD) total assets; 367 Norwegian employees (as of May 2026 per registry). This is partial coverage of the global group. The central underwriting gaps are: (1) gross margin and EBITDA path to profitability; (2) absolute ARR base (growth rate without the denominator is insufficient for underwriting); (3) customer count and concentration (Saudi Aramco + bp likely > 30% of revenue based on strategic positioning— unconfirmed); (4) cash on hand and runway; (5) 2024 round valuation and resulting dilution. These gaps are systemic for a pre-IPO private company and can only be closed through formal due-diligence data room access. [CI028, CI029, CI030, CI031, CI032]

4.6 Financial Verdict and Diligence Blockers

Revenue quality is the clearest positive in Cognite's financial picture. The combination of 36% ARR bookings growth in FY2025, 150% NRR for newer customers, and Q1 2026 ARR growth accelerating to 57% YoY is a compelling compounding profile. The Atlas AI platform land-and-expand motion—where Atlas AI was in 72% of customer ARR by Q1 2026 and drove >70% of new bookings—suggests durable expansion economics within the installed base. The Forrester Total Economic Impact study (Q1 2026) citing 465% ROI for Cognite customers, though vendor-facilitated, provides a third-party framework for premium pricing defense. Capital adequacy is plausible but not confirmed. At a group burn rate estimated at $60–90M/year, the June 2024 ~$300M round would provide 3–5 years of runway from close, implying a runway horizon of mid-2027 to mid-2029. But this is an estimate; actual burn, cash on hand, and any debt facilities are undisclosed. The Aker ASA majority-owner backstop (NOK 100B+ NAV) reduces existential financing risk. The principal financial blockers are: (1) stale $1.6B valuation from 2021—any new investment must negotiate blind on current valuation; (2) absence of gross margin disclosure prevents underwriting the economics of the platform business; (3) Saudi Aramco's combined customer-and-investor role creates a conflict-of-interest structure that could influence product roadmap, data access terms, and eventual exit optionality; (4) the private-company opacity on absolute ARR and customer count prevents triangulating revenue quality against comparable SaaS benchmarks; and (5) no IPO, SPAC, or secondary liquidity pathway has been disclosed, leaving strategic acquisition by Aker, Saudi Aramco, or a major technology firm as the most plausible exit scenario. [CI033, CI034, CI035, CI036]

4.7 Exhibits

5.1 Product Module Map and Definitions

Cognite's commercial platform consists of three named product layers and a set of built-in applications, all sharing the same Industrial Knowledge Graph data foundation. Cognite Data Fusion (CDF) is the core industrial DataOps platform: it ingests, contextualizes, and stores data from disparate IT, OT, and engineering systems — time-series from SCADA and DCS, documents, events, 3D models, and P&IDs — and makes the result available through a versioned REST API, a Python SDK, and a JavaScript SDK. CDF is the only product in the stack that can operate independently; Atlas AI and Flows both require CDF as their data foundation. Cognite Atlas AI, launched in June 2024, extends CDF with a low-code industrial agent workbench. Digital teams and domain experts build AI agents using pre-configured templates without writing code. Agents are powered by Context Augmented Generation (CAG), which injects CDF's graph context into LLM prompts, and by autoLLM, a routing mechanism that selects the optimal Large Language Model, Small Language Model, or Custom Language Model for each agent task. As of fiscal year 2025, 59% of customers by ARR had operationalized Atlas AI, and Atlas AI customer count grew nearly 700% year-over-year. Cognite Flows, announced at Cognite's IMPACT conference and launched May 12, 2026, is the action layer of the platform. It provides a single-screen, persona-based adaptive workspace for frontline operators and engineers, embedding AI agents and surfacing data from all enterprise applications simultaneously. Flows also redefines developer experience: using AI-native coding tools such as Claude or Cursor, developers can build production-grade industrial applications in days instead of months. More than 30% of Cognite's customer base was early-enabled with Flows at launch, with 50+ customers actively using it. Built-in applications bundled under Flows include Industrial Canvas (visual workspace for P&IDs and time-series), InField (mobile field access), Maintain (maintenance work order management), and Charts (time-series analytics). [CE001, CE002, CE003, CE004, CE005, CE006]

5.2 Customer Workflows and Use Cases

Cognite's platform targets asset-heavy industries — oil and gas, energy, manufacturing, pharmaceuticals, and chemicals — where operational data is voluminous, heterogeneous, and historically siloed. The platform's primary value proposition is converting raw sensor, equipment, and engineering data into AI-ready knowledge that frontline workers, process engineers, and digital teams can act on in real time. Across oil and gas, Cognite's most established vertical, customers use CDF to eliminate the time spent hunting for operational data. Cognite's oil and gas product page cites 90% less time spent making sense of data through unique IT/OT/ET contextualization, 10–25x faster time to deploying solutions into live production, and 0.5–1.5% increase in overall production throughput. TotalEnergies expanded its Cognite deployment in September 2024 to cover all upstream assets worldwide, citing improved operational excellence through AI-ready data. A Forrester Total Economic Impact study commissioned by Cognite cited 465% ROI and USD 29.4 million in total benefits. In manufacturing, B. Braun (medical technology) used Cognite Flows to achieve unified, real-time visibility into asset health across all sites within four weeks of deployment. Idemitsu Kosan (energy/chemicals, Japan) deployed Flows to digitize accumulated expert knowledge and reduce operational risk. A global top-10 pharmaceutical company achieved 30x faster time-to-value using Flows versus traditional methods: what previously required a 20-plus-person team and 3–5 months was delivered in four days, with user acceptance testing compressed from 6–9 months to two months. In chemicals, Celanese (specialty chemicals, Texas) partnered with Cognite and NVIDIA to deploy NV-Tesseract NIM time-series forecasting integrated with CDF's knowledge graph, targeting elimination of prediction bias jumps in reaction water level monitoring. The NVIDIA partnership, announced March 16, 2026, operationalizes foundational time-series AI for heavy industry through the Cognite platform. [CE013, CE014, CE015, CE016, CE017]

5.3 Architecture and Integration Model

CDF's core architectural innovation is the Industrial Knowledge Graph — a unified semantic model that links time-series data, events, documents, visual data streams, 3D models, and engineering drawings within a single contextual structure. This graph stores over 80 trillion time-series data points in context as of fiscal year 2025, a 48% year-over-year increase. In the same period, customers doubled data modeling instances, and AI token usage grew nearly 500%, signaling rapid AI adoption on top of the graph. Cognite's developer surface is deliberately API-first. The REST API uses versioned endpoints with a published backward compatibility policy and end-of-life schedule for deprecated API versions. The Python SDK (v8, the current major release) introduced full async support via the new AsyncCogniteClient, enabling non-blocking concurrent operations and native async/await patterns for all API operations. The SDK is tightly integrated with pandas. A JavaScript SDK is also provided. Cognite additionally ships an MCP (Model Context Protocol) server integration, allowing developers to connect their development environments and AI coding tools directly to Cognite's documentation and API. The Python SDK is distributed through PyPI as the `cognite-sdk` package. On the infrastructure side, CDF is cloud-agnostic: it deploys on AWS, Azure, and Google Cloud Platform with globally distributed clusters to meet data sovereignty and latency requirements. Identity and access management relies on OIDC, with support for Microsoft Entra ID, Okta, and other identity providers. Cognite also white-labels CDF to other software providers. Integration depth is broad: Cognite has announced integrations with Snowflake (Energy Solutions collaboration, January 2026), Databricks (for enterprise AI in industrial operations), Microsoft Azure OpenAI Service, and NVIDIA NIM microservices. Customers connect source systems — SCADA, DCS, ERP, OSIsoft PI, and others — via CDF's IT/OT ingestion connectors. The Flows architecture builds on CDF's knowledge graph rather than adding a separate data layer, ensuring that insights and new applications are always grounded in real-time operational context. [CE018, CE019, CE020, CE021, CE022, CE023]

5.4 Trust, Security, and Compliance

Cognite's security program is architected for industries where breaches extend beyond data loss to production shutdowns, environmental incidents, and safety events. The platform undergoes testing and third-party audits across infrastructure, applications, and operations as part of a formal Secure Development Lifecycle (SDL). The company describes security as a "functional requirement of the product, not a perimeter fence." As of May 2026, Cognite holds the following certifications and attestations: ISO 27001 (information security management), ISO 9001 (quality management), ISO 27018 (cloud privacy for personal data), SOC 2 Type 2 (Security criterion), and SOC 3 (public summary security report). CDF additionally supports customer compliance with a set of industrial control and critical infrastructure standards: NIST CSF, IEC 62443 (sub-standards 2-4, 3-2, 3-3, and 4-1), CMMC, FIPs, NERC CIP v.5, GxP, and CSA STAR Level 1. EU AI Act compliance is declared and globally available clusters enable data sovereignty for regulated deployments. The Trust Center at trust.cognite.com provides continuous, real-time visibility into security posture, system health, and compliance documentation. SOC 2 Type 2 and SOC 3 audit reports are available to prospective customers. UpGuard continuously monitors Cognite's externally visible security posture using open-source and proprietary threat intelligence feeds, assessing website security, email security, phishing and malware, brand reputation, and network security; however, UpGuard has not publicly disclosed a specific numeric score or identified material vulnerabilities as of May 2026. The statusgator.com service monitors Cognite's platform status for outages, including detecting disruptions before they become public. No major publicly documented outage affecting customer operations has been identified in the research period. Diligence gaps remain: SOC 2 Type 2 full reports are shared only under NDA, not publicly posted; IEC 62443 compliance is stated as customer-supportable but an independent third-party IEC 62443 audit certificate has not been identified; EU AI Act classification for high-risk industrial autonomy use cases may require additional documentation; and CMMC certification level is undisclosed. [CE027, CE028, CE029, CE030, CE031, CE032]

5.5 Roadmap, Releases, and Ecosystem

Cognite follows a quarterly product release cadence for CDF, Atlas AI, and Flows capabilities. The most significant recent milestones span June 2024 through May 2026 and include two major new product launches and multiple partner integrations. The June 2024 Atlas AI launch marked the platform's formal entry into industrial agentic AI, adding a low-code agent builder and a pre-built agent library on top of CDF's existing DataOps capabilities. The January 2026 Snowflake Energy Solutions collaboration enabled joint customers to unify IT/OT data in the Snowflake AI Data Cloud while retaining Cognite's knowledge graph enrichment layer. In March 2026, Cognite announced and demonstrated the NVIDIA NV-Tesseract NIM integration at NVIDIA GTC, with Celanese as the anchor production deployment. In March 2026, IDC recognized Cognite as a Leader in the IDC MarketScape for Worldwide Industrial DataOps Platforms 2026, citing CDF's support for all five key industrial DataOps capabilities, its white-label ecosystem, and its downstream AI capabilities through Atlas AI. Cognite Flows launched on May 12, 2026, and the company appeared on the CNBC Disruptor 50 list on May 19, 2026. The partner ecosystem spans hyperscalers (Microsoft Azure, AWS, GCP), specialized data platforms (Snowflake, Databricks), AI infrastructure (NVIDIA), and a growing ISV/SI channel including Radix, L&T Technology Services, and RoviSys as Flows launch partners. The developer community is accessible through hub.cognite.com, with active discussions on data type configuration, API authentication, and time-series query patterns. Stack Overflow questions tagged 'cognite' reflect active practitioner engagement, including challenges with CDF data type management and SDK authentication flows. The Python SDK on PyPI (cognite-sdk) is the primary distribution channel for developer adoption. The GitHub repository for cognite-sdk-python reflects ongoing active development, with the v8 release introducing async support. [CE034, CE035, CE036, CE037, CE038, CE039]

5.6 Product and Technical Risks

Cognite's product and technology risks fall into four categories: data dependency and contextualization complexity; AI reliability in mission-critical environments; vendor and platform lock-in; and regulatory/compliance depth. The Industrial Knowledge Graph, while a strategic differentiator, also represents the primary lock-in mechanism. All downstream AI capabilities — Atlas AI agents, Flows workflows, and built-in applications — require CDF as the data foundation. Migrating away from Cognite requires re-contextualizing potentially hundreds of millions of data relationships across customer IT/OT estates. The upfront investment in data contextualization also creates an implementation risk: without specialized OT/IT integration expertise, customers face long time-to-value cycles despite Cognite's claims of 7 weeks from deployment to measurable impact. Context Augmented Generation improves AI accuracy in industrial settings, but LLM hallucination remains an inherent risk for mission-critical decisions — equipment shutdowns, process parameter adjustments, or safety recommendations. Cognite acknowledges this risk through its CAG architecture and emphasis on traceable, sourced outputs, but no independent benchmark quantifying Atlas AI hallucination rates versus competing approaches has been identified. Developer community signals on hub.cognite.com and Stack Overflow reveal recurring integration challenges with CDF data type management (string versus numeric classification), API authentication setup, and time-series query optimization — all potential friction points for new customers. The SOC 2 Type 2 audit scope is available but only under NDA; the IEC 62443 compliance claim is not supported by a publicly referenced independent certification; and the EU AI Act may impose additional documentation burdens on Cognite customers deploying high-risk industrial AI (e.g., autonomous process control or safety-critical recommendations). The platform's OIDC reliance on third-party identity providers (Microsoft Entra ID, Okta) creates a dependency risk: an IdP outage or misconfiguration could lock operators out of mission-critical workflows. [CE040, CE041, CE042, CE043, CE044, CE045]

5.7 Exhibits

6.1 Customer Base Overview and Segmentation

Cognite's industrial data and AI platform serves asset-intensive operators across energy, manufacturing, chemicals, pharmaceuticals, and metals and mining. As of FY2025 the company reported revenues exceeding $170M with 36% ARR bookings growth year-over-year, implying a customer base that spans multiple verticals and geographies. Management disclosed that approximately 40% of 2025 revenue derives from outside oil and gas — a meaningful diversification from the company's origins as a Norway-based O&G data platform. Named anchor accounts include integrated majors such as bp and Saudi Aramco, national oil companies such as Cosmo Energy, diversified energy companies such as TotalEnergies and Aker BP, chemicals producers (NOVA Chemicals, Celanese), a Japanese steel company (JFE Steel), a German medical device manufacturer (B. Braun), and a Japanese refiner (Idemitsu). The platform's go-to-market emphasizes enterprise digital transformation programs; most deployments cited involve dozens to hundreds of assets and thousands of users. Monthly active users grew 26% year-over-year, and 59% of customers by ARR are using the Atlas AI layer as of the run date, indicating broad product adoption depth beyond the foundational data integration layer.

6.2 Named Customer Deployments and Measured Outcomes

Ten named customers with documented production deployments provide the strongest evidence of real-world adoption. Aker BP, Cognite's longest-tenured oil and gas reference, reports up to 50% reduction in visual inspection time and 30-80% reduction in maintenance execution time across six Norwegian North Sea assets. The company's CDO, Dr. Paula Doyle, publicly credited the platform as enabling "the next generation of operational excellence." This outcome is independently corroborated by Aker BP's own investor communications. Idemitsu onboarded four refinery sites in under six months, scaled to over 3,000 users, uploaded more than 10,000 P&IDs, and management estimates 50,000 hours of annual labor savings. TotalEnergies expanded its deployment to 39 upstream assets across three years with over 200 key production systems integrated, supported by a press release from TotalEnergies itself. NOVA Chemicals, Canada's largest polyethylene producer, reports reducing data access time from hours to minutes across an eleven-plant target scope. Celanese's "Digital Plants of the Future" program references Cognite as the enabling platform. B. Braun attributes a four-week improvement in a key production workflow to Cognite Flows. Cosmo Energy integrated three Japanese refineries with 250 billion data points and five million engineering documents. JFE Steel is also a named Flows customer. The proof set spans five verticals and four geographies, providing meaningful breadth for a platform positioned as industry-agnostic.

6.3 Customer Retention, Expansion, and Commercial Durability

Cognite disclosed a net revenue retention rate of 150% for its newer cohorts of customers, indicating that expansion revenue substantially exceeds any churn. The company reported 700% growth in Atlas AI customers since that product's launch, and 30%+ of the entire customer base has been enabled with Cognite Flows as of May 2026. These figures are company-reported and unaudited, but they are corroborated directionally by Forrester analyst coverage and by the pattern of named customer deployments: TotalEnergies' multi-year expansion from a pilot to 39 assets, Aker BP's sustained multi-asset deployment, and Idemitsu's platform-wide rollout all reflect durable enterprise commitments. The knowledge graph now holds 80 trillion time series data points — a 48% year-over-year growth — implying continuous customer data activity rather than static deployments. Independent evidence of gross revenue retention or churn rates is not publicly available. No regulatory filing or third-party audit of NRR or churn metrics was found. Concentration risk exists: Saudi Aramco is simultaneously Cognite's majority shareholder (74% stake acquired in 2022) and one of its largest anchor customers, creating an ownership-customer dependency that is structurally unusual and potentially concerning to competing oil companies evaluating the platform.

6.4 Competitive Dynamics and Go-to-Market Evidence

Cognite's customer wins suggest a pattern of large enterprise land-and-expand: initial pilots at one or two assets followed by multi-year expansions to dozens of facilities. The company's ecosystem positioning with Snowflake and Databricks enables it to integrate into customer data architectures rather than displacing them, lowering procurement friction. The IDC MarketScape positioned Cognite as a Leader in worldwide industrial DataOps platforms — an independently produced analyst assessment that carries third-party weight. The CNBC Disruptor 50 inclusion (May 2026) signals broad technology-media recognition. On the go-to-market side, Cognite relies primarily on direct sales into the C-suite of industrial operators; channel partners (Snowflake, Databricks, AWS) expand reach without becoming primary distribution channels. Independent review platforms (Gartner Peer Insights, Slashdot, PeerSpot, SourceForge) show sparse to no user reviews relative to the claimed enterprise customer count — a meaningful gap in independent social proof. The platform's strength lies in reference-able production deployments at large named accounts rather than bottom-up review-driven adoption. Competitive differentiation vs. AVEVA, OSIsoft (AVEVA PI), AspenTech, and Palantir rests on industrial knowledge graph depth and the Atlas AI agent layer.

6.5 Concentration Risk and Evidence Quality Assessment

Two concentration risks are material. First, Saudi Aramco's 74% equity ownership creates a structural conflict of interest: any competing integrated major (bp is also a named customer) or national oil company evaluating Cognite must weigh whether sensitive operational data shared with Cognite flows to a majority-owned entity of a competitor. This risk is documented by Reuters' 2022 coverage of the acquisition and has not been publicly resolved by Cognite. Second, O&G still represents approximately 60% of revenue, exposing Cognite to energy-transition uncertainty and capital discipline cycles in the sector. Evidence quality is strong for named customers in O&G (Aker BP, TotalEnergies, Idemitsu) where Cognite provides case study pages with specific metrics, and for Idemitsu and TotalEnergies where third-party or customer-owned corroboration exists. Evidence quality for newer vertical entrants (B. Braun, Celanese, JFE Steel) is thinner: outcomes are asserted by Cognite with minimal independent verification. Net revenue retention and customer count data are company-disclosed only and have not been verified by an independent auditor or regulatory filing.

6.6 Exhibits

7.1 Risk Overview

Cognite operates at the intersection of industrial operations, enterprise software, and generative AI — a combination that concentrates multiple risk dimensions simultaneously. As of May 2026, the company reports $170M+ in FY2025 ARR with 36% YoY growth, yet remains privately held with no audited public financials, creating a fundamental opacity around burn rate, gross margin, and capital runway. The risk heatmap (FR001) plots severity and likelihood for each primary risk category. The highest residual-severity risks are oil-and-gas sector concentration and competitive displacement by well-capitalised incumbents and hyperscalers. The risk transmission map (FR002) illustrates how upstream shocks — an oil price collapse, Aker governance change, or AWS outage — cascade to revenue, customer renewal, and ultimately to financing capacity and valuation. Three risks stand out as thesis-breaking: (a) loss of Aker BP or TotalEnergies as anchor customers without replacement revenue, (b) a material security incident affecting industrial control data, and (c) a sustained oil-price decline below ~$60/bbl that causes major O&G operators to defer or cancel digital transformation programmes. The chapter draws on regulatory text (GDPR, EU DSA, EU AI Act provisions), third-party security assessments, and trade-press risk coverage. Evidence is absent on undisclosed litigation, specific customer contract indemnification terms, and precise cash burn — each of these gaps is explicitly documented in the evidence-gap register. [CR023, CR024, CR025, CR033]

7.2 Regulatory and Legal Risks

Cognite processes industrial operational data under GDPR as a data processor for its EU-based customers, including major operators in Norway, the Netherlands, Germany, and the UK. The Norwegian Data Protection Authority (Datatilsynet) is the lead supervisory authority for Cognite's Norwegian entity, which remains the legal base for much of its EU operations despite the 2025 global-HQ relocation to Tempe, Arizona. The EU AI Act (Regulation 2024/1689, effective August 2024) classifies AI systems used in critical infrastructure — including oil refineries, chemical plants, and energy grids — as high-risk systems, requiring conformity assessment, risk management documentation, and human oversight provisions before deployment. Cognite's Atlas AI agents, which make autonomous recommendations on production operations and maintenance scheduling, may fall into this category depending on the level of autonomy granted by operator customers. The specific applicability depends on customer configuration and the degree to which Atlas AI outputs are acted upon without human review. This is an active regulatory interpretation risk with no definitive precedent as of May 2026. NIS2 (Directive 2022/2555), effective October 2024, imposes cybersecurity obligations on operators of essential services and their critical technology suppliers. Cognite's role as an IT/OT integration layer for energy and critical manufacturing customers potentially triggers third-party ICT supplier obligations under Article 21, requiring customers to conduct security audits of their Cognite deployments and contractually bind Cognite to incident notification timelines. No publicly disclosed regulatory enforcement actions, lawsuits, or litigation involving Cognite were found in searches conducted through May 2026. The absence of disclosure does not confirm clean legal standing; private companies do not always disclose pending disputes. This constitutes a material evidence gap documented below. The regulatory/legal risk register (TR001) ranks each exposure by severity and maps it to a diligence action. [CR001, CR002, CR003, CR004, CR005, CR006]

7.3 Operational and Security Risks

Cognite's platform is built on multi-cloud infrastructure spanning AWS, Microsoft Azure, and Google Cloud Platform. This architecture reduces single-cloud outage exposure but creates dependency on all three major cloud providers simultaneously. The May 7, 2026 AWS outage that disrupted over 150 cloud services illustrates the systemic risk to SaaS platforms when hyperscaler infrastructure degrades; Cognite's specific SLA commitments and redundancy failover procedures are not publicly documented. Cognite holds ISO 27001:2022, SOC 2 Type II, ISO 27018, and ISO 9001 certifications, all publicly referenced on its Trust Center. The platform also supports NERC CIP v.5 for energy-sector customers and IEC 62443 industrial cybersecurity standards. UpGuard continuously monitors Cognite's external security posture; the full risk rating report is behind a paywall, preventing independent third-party verification in this research cycle. The absence of any publicly disclosed security breach is noted but cannot be taken as confirmation of a clean record for non-public incidents. Industrial OT data presents a distinctive attack surface: Cognite ingests data from SCADA systems, DCS platforms, and historian databases that contain real-time process conditions of operating facilities. A breach affecting live industrial process data would carry safety-critical consequences beyond typical enterprise software incidents. Cognite's IEC 62443 certifications partially address this exposure, but residual risk remains in customer-side integration layers that are outside Cognite's direct control. Enterprise CDF deployments typically require 6–18 months of implementation work, creating extended periods of integration risk during which data pipelines are only partially validated. Implementation failures — data-model mismatches, historian connector failures, or CDF performance degradation under load — have been cited anecdotally in trade press coverage of industrial DataOps platforms as the most common cause of delayed value realisation. [CR009, CR010, CR011, CR012, CR013, CR014]

7.4 Partner and Dependency Risks

Aker ASA is Cognite's majority shareholder and the de facto strategic backstop. Aker's Q1 2026 Net Asset Value exceeded NOK 110 billion, confirming continued financial capacity to support Cognite. However, majority ownership by an industrial conglomerate creates governance risks: strategic direction, capital allocation, and exit decisions are substantially influenced by Aker's portfolio priorities, which may diverge from pure enterprise-software growth strategies. A change in Aker's own financial position, leadership, or strategic priorities could materially affect Cognite's access to capital and strategic independence. Saudi Aramco acquired a 7.4% stake in Cognite in February 2022. This creates a valuable commercial relationship with the world's largest oil company and an anchor customer on the Arabian Peninsula, but also introduces potential governance complexity: Aramco's strategic priorities could influence Cognite's product roadmap, particularly for O&G-specific features, in ways that slow diversification into other verticals. Additionally, Cognite's operations in Saudi Arabia create export control and geopolitical risk exposure given evolving US and EU sanctions landscapes. Cognite's NVIDIA partnership for NV-Tesseract integration creates a direct technology dependency on GPU availability and NVIDIA's NIM microservice roadmap. In 2025–2026, GPU supply constraints have been well-documented across the industry; any interruption to NVIDIA's commercial API availability would degrade Atlas AI's time-series forecasting capability. The Snowflake and Databricks partnerships create bidirectional data flow dependencies where a pricing dispute or competitive conflict could affect the co-sell pipeline. Microsoft Azure and AWS are both platform partners and potential horizontal competitors as they expand their industrial AI offerings. The dependency map (FR003) visualises these critical upstream dependencies and their paths to customer value delivery. [CR016, CR017, CR018, CR019, CR020, CR021]

7.5 Financial, People, and Execution Risks

Cognite's financial risk profile is dominated by sector concentration. Oil and gas customers drove approximately 60% of revenue as recently as FY2024, though the company reports 40% of FY2025 revenue now comes from outside O&G. A sustained oil price decline below approximately $60/bbl — the historically observed threshold for major capex programme deferrals — would immediately reduce the digital-transformation budgets of Cognite's core customer base. The company offers no publicly disclosed view on customer industry distribution or contract duration, making independent concentration analysis difficult. Cognite's last publicly disclosed valuation of $1.6B dates to its 2021 unicorn milestone, with no subsequent equity round at a new price. The $300M TCV-led investment in 2024 was structured as new capital but at an undisclosed valuation. Any future financing round or strategic transaction will reset valuation expectations; a flat or down round would constrain employee retention through option dilution and signal competitive difficulty to enterprise buyers evaluating long-term vendor viability. The 2025 relocation of global headquarters from Oslo to Tempe, Arizona is a significant people risk. Cognite's engineering heritage is Norwegian, and its founding technical team (including co-founder Geir Engdahl, CTO) is Oslo-based. Relocation creates a geographic divide between corporate strategy (US) and technical execution (Norway), potentially increasing coordination overhead and attrition among senior Norwegian engineers who did not relocate. Headcount grew 21% in 2025 to over 800 employees; fast growth in new geographies risks diluting institutional knowledge and culture. Rapid product cadence is both a competitive strength and an execution risk. Cognite launched Atlas AI (2024) and Cognite Flows (May 2026) in rapid succession; each major launch requires substantial field-enablement, customer onboarding capacity, and support infrastructure. If go-to-market execution lags product innovation, deal closure rates and time-to-value metrics will deteriorate. [CR023, CR024, CR025, CR026, CR027, CR028]

7.6 Mitigations, Monitoring, and Thesis-Break Triggers

Several structural mitigations reduce the probability of the most severe risk scenarios. Cognite's 150% NRR for newer customers and 59% Atlas AI adoption by ARR demonstrate platform stickiness that makes unplanned churn less likely in the near term. ISO 27001, SOC 2 Type II, and IEC 62443 certifications provide institutional compliance scaffolding that addresses the most common procurement security barriers. The multi-cloud architecture (AWS/Azure/GCP) reduces single-provider outage risk. Aker ASA's NOK 110B NAV confirms balance-sheet depth sufficient to fund Cognite through extended growth phases. Quantitative monitoring indicators can provide early warning on key risks. Oil-price futures below $65/bbl for more than 90 days should trigger a re-examination of O&G customer renewal probability. A quarterly NRR print below 120% would signal account contraction preceding churn. Gartner Peer Insights scores declining more than 0.5 points over two quarters would flag a customer satisfaction deterioration. Executive departures (CTO, CPO, CFO) within 12 months of each other would trigger a people-risk escalation. Thesis-break triggers are events that would fundamentally challenge the investment thesis: loss of Aker BP or TotalEnergies as a customer without visible pipeline replacement, a material OT security breach attributed to the Cognite platform, an oil price sustained below $50/bbl for 12+ months, or discovery of undisclosed litigation with material financial exposure. A confirmed acquisition of Cognite by a large incumbent (AVEVA/Schneider, SAP, Microsoft) at below the 2021 $1.6B valuation mark would represent a negative carry outcome. The mitigation and kill-criteria table (TR005) maps each primary risk to its monitorable trigger and recommended action implication. [CR039, CR040, CR041, CR042]

7.7 Exhibits

8.1 Investment Thesis and Recommendation

Cognite operates at the intersection of two durable trends—industrial asset digitization and enterprise AI agent adoption—with a differentiated Industrial Knowledge Graph that no single competitor replicates end-to-end. The investment thesis rests on four pillars: (1) a proven land-and-expand model evidenced by 150% NRR for newer customers and Atlas AI reaching 72% of customer ARR by Q1 2026; (2) a platform architecture with genuine switching costs built from multi-year data contextualization and embedded AI workflow dependencies; (3) ARR bookings growth accelerating from 36% in FY2025 to 57% in Q1 2026, suggesting the Atlas AI upsell cycle is working; and (4) a blue-chip customer base including Saudi Aramco, bp, TotalEnergies, Aker BP, and B. Braun, several of whom are multi-year anchor customers. The anti-thesis is driven primarily by valuation opacity and financial transparency gaps. The last confirmed valuation of $1.6 billion was set in May 2021—before Atlas AI existed, before the June 2024 ~$300M round, and at roughly half the current revenue scale. Investors cannot underwrite a position without knowing the 2024 round's post-money mark, the gross margin structure (estimated at 60–75% but not disclosed), absolute ARR (denominator for the growth rate headline), or the full-base NRR beyond the "newer customers" cohort. Saudi Aramco's 7.4% equity stake and Aker ASA's majority ownership create exit overhang and potential strategic conflict risks that are difficult to price without governance disclosure. C3.ai's public market trajectory—multiple compression from 30x+ EV/Revenue in 2021 to 8–12x by 2025—provides a cautionary precedent for the industrial AI SaaS category. Recommendation: TRACK. Cognite earns a strong qualitative score on product, customer proof, and growth trajectory, but cannot be underwritten to a specific entry price until gross margin, absolute ARR, full-base NRR, the 2024 round valuation, and customer concentration data are disclosed. At an entry price at or below the 2021 $1.6B mark, downside protection is moderate; at 2–3× the 2021 mark (consistent with the base-to-bull scenario range), the risk/ return profile is less favorable without confirmation of margin sustainability. [CV001, CV002, CV005, CV009, CV010, CV011]

8.2 Financing, Valuation, and Capital Structure

Cognite's financing history consists of two large disclosed rounds. In May 2021, TCV led a $150M investment that valued the company at $1.6 billion, with TCV's Jake Reynolds joining the board. This represented the largest SaaS funding round in Europe at that time per the company's own press release. In June 2024, TCV led a second round described as approximately $300M; the post-money valuation for this round has never been publicly disclosed. Combined with an undisclosed earlier investment by Accel (Cognite investor confirmed via Accel's portfolio page), total known capital raised is at least $450M. CNBC's May 2026 Disruptor 50 profile cites $300M in funding, likely reflecting the 2024 round alone rather than cumulative capital. The ownership structure is materially opaque. Aker ASA holds a majority stake in Cognite; the exact percentage is undisclosed as of May 2026. Saudi Aramco acquired a 7.4% stake from Aker BP in February 2022, creating a strategic customer-shareholder that simultaneously provides revenue and equity upside but also concentration and influence risk. TCV holds board representation through Jake Reynolds. The combination of Aker ASA majority control, Aramco strategic stake, and TCV minority position creates a complex governance structure that could complicate a standard acquisition exit and may require multiple acquirer consents. From the Norwegian Registry (Brønnøysund Register Centre), Cognite AS (org. nr. 918274758) filed accounts showing FY2024 revenue of approximately NOK 1.67 billion (~$158M USD at the 2024 average exchange rate) and a net loss of approximately NOK 547M (~$52M USD). These figures cover the Norwegian legal entity only; global consolidated group revenue exceeded $170M in FY2025 per Cognite's January 2026 press release. The implied group-level burn rate is estimated at $60–90M/year, suggesting the June 2024 ~$300M round provides runway through mid-2027 to mid-2029 absent material burn acceleration. This estimate is highly uncertain given the absence of consolidated cash flow disclosure. The stale valuation mark is the single most operationally significant gap for any investor. At NTM revenue of approximately $200–220M (implied by current growth trajectory), a Cognite entry at the 2021 $1.6B mark implies an ~7–8x NTM Revenue multiple—toward the lower end of the applicable comp range for a high-growth industrial AI platform—providing some downside discipline. Any entry above $2.5B implies 11–12.5x NTM Revenue and demands confirmation of gross margin sustainability and full-base NRR to justify the premium. [CV001, CV002, CV003, CV004, CV006, CV007]

8.3 Comparable Company Analysis

No single public company is a clean analog for Cognite. The most relevant comparables cluster in two tiers: (1) high-growth enterprise AI and data platforms (Palantir, C3.ai) that share Cognite's AI-native positioning and sales profile but differ substantially in market mix and customer profile; and (2) industrial software incumbents (PTC) and acquired platforms (AVEVA, AspenTech) that share Cognite's vertical focus but represent different growth and margin profiles. C3.ai (NYSE: AI) is the closest public comparable in terms of product positioning—enterprise AI applications for industrial operations—but carries important caveats. C3.ai reported approximately $310M in FY2025 revenue, growing at roughly 25% year-over-year. Cognite's ARR bookings growth (36–57% YoY) is meaningfully faster, supporting a premium to C3.ai's current EV/NTM Revenue multiple of approximately 8–12x. Crucially, C3.ai's public market history provides a cautionary data point: from a peak EV/Revenue multiple above 30x in December 2020, C3.ai declined to 8–12x by 2025, illustrating how sentiment-driven industrial AI multiples can compress sharply when growth disappoints. C3.ai's 10-K annual report filed with the SEC for FY2025 confirms a net loss of approximately $214M—a larger absolute loss than Cognite's estimated group-level burn—at roughly twice Cognite's revenue scale. Palantir Technologies (PLTR) trades at an elevated EV/NTM Revenue multiple above 30x as of mid-2026, but is not a valid direct comparable: Palantir derives over 40% of revenue from US government and defense contracts, carries a different margin structure, and benefits from US defense AI spending tailwinds that are not available to Cognite. Cognite's comparable multiple should apply a material discount to Palantir's current mark. PTC Inc. (NASDAQ: PTC) represents the scaled industrial software benchmark at approximately $2.3B in FY2024 revenue and an EV/NTM Revenue multiple of approximately 8–10x. PTC's slower growth profile (8–12% organic) implies Cognite deserves a premium for its higher growth, partially offset by Cognite's private-company opacity and oil-and-gas concentration. Acquisition precedents provide bracketing reference points. AVEVA (acquired by Schneider Electric for approximately £9.5B, completing in January 2023) implies an acquisition multiple of approximately 5–7x trailing revenue. AspenTech (acquired by Emerson in a transaction valued at approximately $11B in 2022) implies 15–18x trailing revenue at closing—a premium justified by AspenTech's high gross margins and sticky process-simulation installed base. These precedent multiples bracket the applicable range: a Cognite acquisition exit at 8–14x NTM Revenue is defensible; below 6x would represent below-market pricing given growth trajectory. Verdantix, an independent industrial AI analyst firm, named Cognite a Leader in its Green Quadrant for Industrial AI Analytics (paywalled report, 2025), confirming third-party positioning at the top of the competitive landscape. PitchBook data on Cognite as of May 2026 is access-restricted; no updated private-round valuation or secondary mark is publicly accessible through standard research channels, which itself is an adverse transparency signal. S&P Global Market Intelligence paywalls industrial software valuation data, preventing independent derivation of a median NTM multiple from primary data. [CV017, CV018, CV019, CV020, CV021, CV022]

8.4 Scenario Analysis—Bull, Base, and Bear Cases

Three scenarios are modeled on next-twelve-months (NTM) revenue applied against a scenario-appropriate EV/NTM Revenue multiple. The NTM revenue base is estimated at $200–220M, derived from the >$170M FY2025 print and a conservative continuation of the current 36–57% ARR growth trajectory with a mean-reversion assumption to 30–40% NTM growth. Gross margin is not publicly disclosed and is held at the industrial SaaS benchmark range of 60–75% across all scenarios; changes to this input materially shift the thesis. Bear case ($1.0–1.4B): Multiple compression to 5–7x NTM Revenue, justified by (a) sustained oil-and-gas market downturn reducing expansion at Aramco and bp; (b) failed Atlas AI commercialization beyond the installed base (penetration falls below 50% of ARR); (c) ARR growth deceleration to below 20% YoY; or (d) broad industrial AI multiple compression following a C3.ai-style trajectory. At 5–7x × $200M NTM revenue, Cognite's implied enterprise value is $1.0–1.4B, below the 2021 $1.6B mark and representing a down-round scenario for investors who entered at or above the 2021 price. Base case ($1.6–2.6B): Multiple range of 8–12x NTM Revenue applied to $200–220M NTM revenue. This scenario assumes: continued 30–40% ARR growth; Atlas AI penetration sustaining above 60% of ARR; blended gross margin of 65–70%; and no material customer concentration event at Aramco or bp. The $1.6B floor corresponds to the 2021 valuation mark at current scale (a "no-premium" exit), while $2.6B requires 12x multiple validation through gross margin and retention disclosure. Bull case ($3.0–4.4B): Multiple premium of 14–20x NTM Revenue applied to $210–220M NTM revenue. This scenario requires: (a) Atlas AI establishing a durable platform moat with NRR sustaining above 130% across the full customer base; (b) successful non-O&G vertical expansion driving the 40% non-O&G revenue share to 55%+ within 24 months; (c) gross margin expanding above 70% as subscription mix improves; and (d) IPO or strategic M&A exit at AI platform comps rather than industrial software comps. Palantir's multiple is not applicable as a ceiling without US government contract equivalent. The Rule of 40 score cannot be calculated from public data. At 36–57% ARR growth and an estimated FCF margin of -20 to -30% (derived from the ~$52M FY2024 net loss at the Norwegian entity and assumed group-level burn of $60–90M), the estimated Rule of 40 score spans roughly 6–37—a wide range consistent with a company whose growth is confirmed but whose profitability trajectory is unknown. This uncertainty is the central valuation blocker. [CV009, CV010, CV011, CV013, CV014, CV015]

8.5 Exit Readiness, Final Diligence Asks, and Kill Triggers

Cognite's most plausible exit paths in 2026–2028 are: (1) strategic acquisition by a large industrial automation or enterprise software incumbent seeking an AI-native industrial DataOps capability; or (2) an IPO when the company achieves profitability milestones or market conditions support a growth-stage industrial AI listing. The CNBC Disruptor 50 recognition in May 2026 is consistent with building public profile ahead of a potential liquidity event. Strategic acquirer candidates include Schneider Electric (which acquired AVEVA as a DataOps precedent), Emerson Electric (acquired AspenTech), Siemens, Honeywell, and ABB— all of which compete in industrial automation and have strategic rationale for acquiring Cognite's Industrial Knowledge Graph and Atlas AI agent workbench. A hyperscaler (Microsoft, given the Energy and Resources Partner relationship; or AWS) could also be a strategic buyer. The AVEVA ($9.5B) and AspenTech ($11B) precedent M&A transactions bracket the relevant enterprise value range for strategic exits at Cognite's scale. Aker ASA's majority ownership creates a potential acquirer deterrent: a strategic buyer would need Aker ASA consent to acquire a controlling stake, and Aker ASA may prefer to retain exposure to industrial data assets aligned with its broader energy investment thesis. Saudi Aramco's 7.4% equity stake adds another veto dimension. These governance complexities should be surfaced in diligence and resolved before any investment thesis relies on a near-term M&A exit. The five most critical diligence asks before investment: (1) gross margin by revenue stream (subscription vs. professional services), without which SaaS benchmarking is impossible; (2) full-base NRR and gross dollar retention, to confirm whether the 150% "newer customer" NRR masks churn in older cohorts; (3) absolute ARR and customer count by tier, to verify concentration and ACV distribution; (4) the June 2024 post-money valuation and current cap-table structure including preference stack and liquidation preferences; and (5) cash on hand and 12-month cash runway, to assess whether the next financing event will be dilutive. [CV036, CV037, CV038, CV039, CV042]

8.6 Exhibits