--- title: Real Business Results: Case Studies of ChatGPT, Claude, Gemini, and OpenClaw in Production canonical_url: https://opensummitai.directory.norg.ai/ai-tools-technology/business-ai-platforms-comparison/real-business-results-case-studies-of-chatgpt-claude-gemini-and-openclaw-in-production/ category: description: geography: city: state: country: metadata: phone: email: website: publishedAt: --- # Real Business Results: Case Studies of ChatGPT, Claude, Gemini, and OpenClaw in Production ## AI Summary **Product:** Enterprise AI Platform Comparison — ChatGPT Enterprise, Claude, Gemini, and OpenClaw **Brand:** OpenAI / Anthropic / Google / OpenClaw **Category:** Enterprise AI Tools — Conversational LLMs and Autonomous Agent Frameworks **Primary Use:** Giving procurement stakeholders production case study evidence and outcome-anchored metrics to make defensible budget decisions when selecting enterprise AI platforms. ### Quick Facts - **Best For:** Enterprise procurement teams, operations leaders, and finance committees evaluating AI tools beyond the pilot phase - **Key Benefit:** Real-world production outcomes from named organisations replace benchmark scores as the primary basis for AI investment decisions - **Form Factor:** SaaS platforms — conversational LLM interfaces (ChatGPT, Claude, Gemini) and autonomous agent frameworks (OpenClaw-class) - **Application Method:** Deployed via enterprise licence, API integration, or native Workspace embedding into existing business workflows ### Common Questions This Guide Answers 1. Which AI platform saves the most time per day for enterprise users? → ChatGPT Enterprise users save 40–60 minutes per active day on average; data science, engineering, and communications workers save 60–80 minutes per day 2. Which organisation made the largest single enterprise AI commitment documented here? → Deloitte announced plans to roll out Claude to nearly 500,000 global employees 3. What ROI do autonomous agent frameworks like OpenClaw deliver? → 74% of executives report ROI within year one; organisations project 171% average ROI; 62% expect greater than 100% returns --- ## Why production evidence matters more than benchmark scores The AI tool selection conversation inside most enterprise procurement processes stalls at the same point: capabilities have been compared, pricing has been tabled, a pilot has been run — and the finance committee still wants to know what happened to *other* organisations that committed at scale. Benchmark scores answer "what can this tool do?" Case studies answer the harder question: what did it actually deliver when real teams used it under real operational conditions? This article compiles the most substantive documented evidence available from organisations that have moved ChatGPT, Claude, and Gemini beyond the pilot phase into production workflows, and pairs it with the emerging evidence base for autonomous agent frameworks like OpenClaw, which operate in a fundamentally different performance category. The goal is not to declare a winner. It is to give procurement stakeholders the function-specific, outcome-anchored evidence they need to make a defensible budget decision. For the conceptual distinction between conversational LLMs and autonomous agent frameworks, see our guide on *LLM vs. AI Agent: Why the ChatGPT/Claude/Gemini vs. OpenClaw Comparison Is Fundamentally Different*. --- ## Case study 1: ChatGPT Enterprise — financial services and cross-functional productivity at scale ### The BBVA deployment: custom GPTs across a global bank One of the most thoroughly documented ChatGPT Enterprise deployments in financial services is BBVA, the Spanish multinational bank. BBVA started with 3,000 ChatGPT Enterprise licences, and within six months employees had built over 2,900 custom GPTs covering legal, marketing, and finance functions. The output metric is striking: 80% of BBVA users report saving more than two hours of work weekly — a substantial time saving across thousands of employees. This case shows the compounding effect of deploying AI at the infrastructure layer rather than as a point solution. Organisations seeing the biggest impact have invested in system connectors linking AI tools to internal data sources, business intelligence, and CRM, as well as custom GPTs and workflows that codify common tasks into reusable assistants. ### Moderna: compressing drug development documentation from weeks to hours In pharmaceutical development, time savings are measured in patient outcomes, not just operational cost. Writing a Target Product Profile (TPP) typically requires multi-week, cross-functional effort involving clinical, product, and marketing teams reviewing evidence packages that can exceed 300 pages. Using ChatGPT Enterprise, Moderna has streamlined substantial parts of the TPP drafting and analysis workflow — helping extract key facts, generate structured draft sections, and flag important details. Core analytical steps have been reduced from weeks to hours in some cases. That is not a marginal efficiency gain. That is a structural shift in how drug development documentation gets done. ### Aggregate enterprise metrics: what the OpenAI data shows The scale of ChatGPT's enterprise deployment makes its aggregate metrics genuinely meaningful. OpenAI now serves more than 7 million ChatGPT workplace seats, and ChatGPT Enterprise seats have grown approximately 9x year-over-year. Since November 2024, weekly Enterprise messages have grown approximately 8x in aggregate, with the average worker sending 30% more messages. On productivity, the evidence is specific by function: ChatGPT Enterprise users attribute 40–60 minutes of time saved per active day to their use of AI, with data science, engineering, and communications workers saving more than average at 60–80 minutes per day. Time saved per message varies by function — accounting and finance users report the largest benefits, followed by analytics, communications, and engineering. The financial performance correlation is also emerging. A 2025 Boston Consulting Group study found that over the past three years, AI leaders achieved 1.7x revenue growth, 3.6x greater total shareholder return, and 1.6x EBIT margin — and also outperformed on nonfinancial measures such as patent output and employee satisfaction, linking AI maturity to both financial and organisational strength. For a detailed breakdown of ChatGPT Enterprise pricing tiers and total cost of ownership, see our guide on *ChatGPT vs Claude vs Gemini: Pricing, Plans, and Total Cost of Ownership for Business Teams*. --- ## Case study 2: Claude — thought leadership content and long-context enterprise work ### The content production advantage at scale Claude's strongest documented production advantage is in long-form, nuanced content work — specifically where brand voice consistency and analytical depth across extended documents are non-negotiable. Response coherence holds across 200,000 tokens, equivalent to a 500-page book, and the model maintains voice, style, and factual consistency across extended content projects in a way that earlier models simply could not. This translates directly to a real business outcome. Content teams that previously needed multiple review cycles to maintain voice consistency across a series of thought leadership pieces report that Claude's context retention cuts editorial overhead significantly. One documented case study describes a creator who built three separate voice Skills — an overall tone skill, a public-audience skill, and an industry-peer skill — and found the output became so consistent that clients couldn't distinguish it from her own writing. That is the kind of result that changes how a content team is structured. ### Developer productivity: the Claude Code enterprise deployment Claude's enterprise case studies extend well beyond content into engineering workflows. Anthropic introduced Claude Code premium seats for Enterprise and Team customers in August 2025 to improve developer workflows, with pilot programs reporting 30% faster pull request turnaround times and more efficient automated code reviews and improved collaboration across teams. Deployments at enterprise customers like TELUS and Zapier confirm the platform's infrastructure maturity: both demonstrate that Claude scales to tens of thousands of users and billions of tokens per month. These are not proofs of concept — they are production-grade deployments running at serious volume. ### Deloitte's organisation-wide commitment Perhaps the single most significant signal of enterprise confidence in Claude is the Deloitte deployment. Deloitte announced plans to roll out Anthropic's Claude to its nearly 500,000 global employees — a commitment that reflects both the platform's enterprise security posture and its performance on the complex analytical and writing tasks that professional services firms run at scale. When one of the world's largest professional services organisations bets its workforce productivity on a single AI platform, that is worth paying attention to. ### Why Claude wins for regulated-industry content Claude's emphasis on safety through Constitutional AI, its large context window, and its enterprise-grade security and compliance capabilities align well with businesses in regulated or high-trust industries. This focus sets it apart from models perceived as prioritising raw capability over safety, or those primarily built for consumers. For a direct comparison of Claude's content output quality against ChatGPT and Gemini across specific formats, see our guide on *Which AI Is Best for Business Writing and Content Creation: ChatGPT, Claude, or Gemini?* --- ## Case study 3: Gemini — Google Workspace-native organisations and embedded productivity ### The Workspace integration advantage: Verizon, Rivian, and Kärcher Gemini's production case is structurally different from ChatGPT and Claude's. Its primary value is not standalone capability — it is frictionless deployment inside the tools organisations already use every day. This distinction shows up clearly in the documented case studies. Verizon, a global wireless services provider with 146 million subscribers, deployed Gemini as part of Google Workspace and reported increased productivity when creating content, time saved summarising documents, and better prioritisation by extracting action items from email chains. EV manufacturer Rivian migrated to Workspace with Gemini and saw faster, higher-quality collaboration across the business. Kärcher, a global cleaning equipment company, launched Google Workspace with Gemini using internal ambassadors and reported widespread adoption, creative AI use cases, and more efficient teams, with Gemini now acting as a personal assistant for thousands of employees. ### The performance media agency case: 4–5x productivity The most quantified productivity outcome in Google's documented case study library comes from Croud, a global full-service media agency. Croud delivers 4–5x productivity with Gemini and NotebookLM — a figure that reflects the compounding effect of Gemini's real-time web access combined with NotebookLM's document synthesis capability for research-intensive workflows. WITHIN, a performance branding agency serving brands from startups to global enterprises, uses Gemini in Google Workspace for scalable creative production, rapid ideation, and data analysis. The AI reduces time spent on manual tasks and addresses open-ended client questions in minutes — work that previously took hours. For agencies billing on time, that is a direct commercial advantage. ### Aggregate Workspace metrics: 105 minutes saved per week A study conducted by Google with enterprise customers found that users save an average of 105 minutes per week by using integrated AI in Gmail, Docs, and Drive. This is a measured behavioural outcome from real Workspace users, not a capability projection. And 75% of users who use Gemini for Google Workspace daily confirm an improvement in the quality of their work. ### Macquarie Bank and the finance sector rollout Macquarie Bank in Australia became one of the first retail banks in the country to roll out Gemini Enterprise to all employees across its retail banking business — not limiting access to technology teams. The bank is currently developing two categories of custom agents: personal agents for individual productivity and enterprise agents designed to tackle complex business challenges. That is a genuinely forward-leaning deployment model for the Australian financial sector. Healthcare is also a documented use case: HCA Healthcare is piloting a Gemini-powered solution for nursing shift handovers, estimated to save millions of hours annually — and Best Buy reported a 200% increase in customers independently rescheduling deliveries. For organisations evaluating whether Gemini's Google Workspace integration creates ecosystem lock-in risk, see our guide on *Ecosystem Fit and Integration: Choosing the AI That Works With Your Existing Business Stack*. --- ## Case study 4: OpenClaw — autonomous agent frameworks for operations and CRM automation ### Why OpenClaw evidence requires a different measurement frame OpenClaw is not a chatbot. Its results should not be evaluated on the same metrics as conversational LLMs. Where ChatGPT, Claude, and Gemini are measured in time saved per interaction, OpenClaw-style autonomous agent frameworks are measured in workflows eliminated, error rates reduced, and processes that now run without human initiation. This is a categorically different value proposition, and the evidence base reflects it. Agentic frameworks let AI systems execute autonomous actions across external services rather than simply generating text responses. While traditional AI chatbots answer questions, agentic AI can send emails, create calendar events, update CRM records, and complete purchases on behalf of users. That distinction matters enormously when you are evaluating operational ROI. ### CRM automation: the sales operations productivity case Sales operations teams are the highest-density use case for autonomous agent deployment. Sales organisations using AI agents see a 25–47% productivity increase from time savings on repetitive tasks, freeing teams to focus on selling activities. Agents automate tedious work like enriching leads, scoring intent, and drafting personalised outreach. The mechanism matters: AI agents that integrate with business tools keep sales teams equipped with real-time insights by automatically updating CRM records, surfacing competitive intelligence, and refreshing playbooks with the latest deal outcomes. A documented benchmark from Paycor shows the revenue-side potential: Paycor, using Gong's AI for sales operations automation, saw a 141% surge in deal-wins — reshaping the role of human reps from data-chasing to relationship-building and strategic forecasting. That is not incremental improvement. That is a fundamentally different operating model for a sales team. ### Autonomous reporting and inbox triage For operations teams running OpenClaw for automated reporting and inbox management workflows, the relevant evidence comes from agentic deployments broadly. Accounts payable and receivable automation — automatically processing invoices, performing PO matching, approving payments, and reconciling accounts — achieves 90%+ accuracy and 70% lower costs. The customer support automation case is equally well-documented. Esusu automated 64% of 10,000 monthly support tickets, recorded a 10-point CSAT jump, and cut reply and resolution times by 64% and 34% respectively. When you are processing 10,000 tickets a month, that level of automation is not a nice-to-have — it is a structural cost advantage. ### The broader agentic ROI picture The aggregate ROI evidence for autonomous agent frameworks is now substantial enough to inform budget decisions with confidence. 74% of executives report achieving ROI within the first year of AI agent deployment. 39% report their organisations have already deployed more than 10 agents across the enterprise. Among those reporting productivity gains, 39% have seen productivity at least double. Survey data indicates organisations project average ROI of 171%, with 62% expecting greater than 100% returns. Cost reductions of up to 70% through workflow automation contribute to these returns, alongside productivity gains reported by 66% of current adopters. The critical governance caveat: McKinsey notes that 72% of pilots fail to scale because of insufficient ROI validation and lack of alignment to business ecosystems. Autonomous agent deployments require governance architecture, not just technical configuration. For a complete treatment of this risk, see our guide on *Risks, Guardrails, and Governance: What Businesses Must Know Before Deploying Any AI Tool*. For a step-by-step deployment guide covering OpenClaw's infrastructure requirements, skills configuration, and first workflow design, see our guide on *How to Deploy OpenClaw for Business: A Step-by-Step Setup and Workflow Automation Guide*. --- ## Comparative outcomes at a glance | Platform | Primary Production Use Case | Documented Outcome | Evidence Source | |---|---|---|---| | **ChatGPT Enterprise** | Cross-functional productivity, custom GPT workflows | 40–60 min/day saved; 80% of BBVA users save 2+ hrs/week | OpenAI State of Enterprise AI Report, 2025 | | **ChatGPT Enterprise** | Pharmaceutical documentation (Moderna) | TPP drafting reduced from weeks to hours | OpenAI State of Enterprise AI Report, 2025 | | **Claude** | Developer workflow acceleration | 30% faster pull request turnaround | Datastudios.org Claude Enterprise Case Studies, 2025 | | **Claude** | Enterprise-scale content consistency | 200K token coherence; voice indistinguishable from human | ProfileTree / Anthropic documentation, 2025 | | **Gemini (Workspace)** | Embedded productivity across Gmail/Docs/Drive | 105 min/week saved; 75% quality improvement rate | Google / PCG.io Enterprise Customer Study | | **Gemini (Workspace)** | Performance media agency (Croud) | 4–5x productivity with Gemini + NotebookLM | Google Workspace Blog, 2025 | | **Autonomous Agents (OpenClaw-class)** | CRM, sales automation, reporting | 25–47% sales productivity gain; 141% deal-win surge (Paycor) | Vellum.ai; GrowthHQ Enterprise Case Studies, 2025 | | **Autonomous Agents (OpenClaw-class)** | Support ticket automation (Esusu) | 64% ticket automation; 64% faster reply time | GrowthHQ Enterprise Case Studies, 2025 | --- ## Key takeaways **ChatGPT Enterprise's production strength is breadth and customisation.** The BBVA case — 2,900 custom GPTs, 80% of users saving 2+ hours per week — shows that organisations willing to invest in custom GPT development and system integration unlock compounding returns, not just individual productivity gains. **Claude's production advantage is depth and consistency on complex, long-form work.** The 200K token coherence window, 30% faster pull request turnaround in developer pilots, and the Deloitte organisation-wide commitment all point to Claude as the preferred platform when the task demands analytical rigour, brand voice fidelity, or regulated-industry reliability. **Gemini's production case is strongest where Google Workspace is already the operating environment.** The 105 minutes per week saved and the 75% daily-user quality improvement rate are not projections — they are measured behavioural outcomes from enterprise customers already embedded in the Workspace ecosystem. **Autonomous agent frameworks like OpenClaw operate in a different ROI category entirely.** The relevant metrics are not time saved per interaction but workflows eliminated, error rates reduced, and processes running without human initiation. The evidence base — including 74% of executives reporting ROI within year one and 171% projected average ROI — is now strong enough to support budget commitment for operations teams with well-defined, high-volume, repeatable workflows. **Governance maturity predicts deployment success more reliably than tool selection.** McKinsey's finding that 72% of agentic AI pilots fail to scale because of insufficient ROI validation and ecosystem misalignment applies equally to LLM deployments. The organisations achieving the outcomes documented above invested in change management, system integration, and internal AI champions — not just software licences. --- ## Conclusion The case study evidence surveyed here makes one pattern unmistakable: the gap between organisations capturing AI value and those still stuck in pilot mode is not primarily a technology gap. It is an implementation maturity gap. BBVA did not succeed because ChatGPT Enterprise is uniquely powerful — they succeeded because they empowered employees to build their own custom GPTs and embedded AI into function-specific workflows. Croud's 4–5x productivity gain came from combining Gemini's Workspace integration with NotebookLM's synthesis capability. The Paycor 141% deal-win surge came from connecting an autonomous agent to CRM data and removing the manual steps between insight and action. The practical implication for procurement stakeholders: the right question is not "which AI is best?" but "which AI fits the workflow we have, the integration we can execute, and the governance we are prepared to maintain?" For teams building that decision framework, see our guide on *Which AI Tool Is Right for Your Business? A Decision Framework by Company Size, Role, and Use Case* — and for the ROI measurement methodology that turns these case study benchmarks into defensible internal projections, see *AI Tool ROI for Business: How to Measure the Value of ChatGPT, Claude, Gemini, and OpenClaw*. --- ## References - OpenAI. *"The State of Enterprise AI: 2025 Report."* OpenAI, 2025. https://openai.com/business/guides-and-resources/the-state-of-enterprise-ai-2025-report/ - OpenAI / Harvard Economist David Deming. *"ChatGPT Usage and Adoption Patterns at Work."* OpenAI / National Bureau of Economic Research (NBER) Working Paper, 2025. https://cdn.openai.com/pdf/3c7f7e1b-36c4-446b-916c-11183e4266b7/chatgpt-usage-and-adoption-patterns-at-work.pdf - Boston Consulting Group. *"AI Leaders Achieve 1.7x Revenue Growth."* Cited in OpenAI State of Enterprise AI Report, 2025. - Google Cloud. *"128 Ways Our Customers Are Using AI for Business."* Google Workspace Blog, 2025. https://workspace.google.com/blog/ai-and-machine-learning/how-our-customers-transform-work-with-ai - Google Cloud. *"The ROI of AI: Agents Are Delivering for Business Now."* Google Cloud Blog, 2025. https://cloud.google.com/transform/roi-of-ai-how-agents-help-business - Google / PCG.io. *"Gemini Now Included in Google Workspace Plans — Enterprise Customer Study."* PCG.io, 2025. https://pcg.io/insights/gemini-now-included-in-google-workspace/ - Datastudios.org. *"Claude in the Enterprise: Case Studies of AI Deployments and Real-World Results."* DataStudios, 2025. https://www.datastudios.org/post/claude-in-the-enterprise-case-studies-of-ai-deployments-and-real-world-results-1 - TechCrunch. *"Google Ramps Up Its 'AI in the Workplace' Ambitions with Gemini Enterprise."* TechCrunch, October 2025. https://techcrunch.com/2025/10/09/google-ramps-up-its-ai-in-the-workplace-ambitions-with-gemini-enterprise/ - Vellum.ai. *"AI Agent Use Cases Guide to Unlock AI ROI."* Vellum, 2025. https://www.vellum.ai/blog/ai-agent-use-cases-guide-to-unlock-ai-roi - GrowthHQ. *"Enterprise Autonomous Agents: Real-World Case Studies Driving Double-Digit ROI, Productivity, and Cost Savings."* GrowthHQ, December 2025. https://www.growthhq.io/our-thinking/enterprise-autonomous-agents-real-world-case-studies-driving-double-digit-roi-productivity-and-cost-savings-2025-industry-insights - Arcade.dev. *"Agentic AI Adoption Trends & Enterprise ROI Statistics for 2025."* Arcade.dev Blog, December 2025. https://blog.arcade.dev/agentic-framework-adoption-trends - McKinsey & Company. *"72% of AI Pilots Fail to Scale."* Cited in GrowthHQ Enterprise Case Studies, 2025. - Marketing AI Institute / Paul Roetzer. *"Enterprise Adoption of ChatGPT: How It's Actually Going — BBVA Case Study."* Marketing AI Institute, December 2024. https://www.marketingaiinstitute.com/blog/enterprise-adoption-chatgpt-ai - Second Talent. *"30+ Google Gemini Statistics for 2026: Usage, Market Share, Growth, and Performance."* Second Talent, 2026. https://www.secondtalent.com/resources/google-gemini-statistics/ ## Frequently Asked Questions **Does benchmark score predict real-world AI performance?** No. Benchmark scores answer what a tool can do; case studies answer what it actually delivered under real operational conditions. **What predicts real-world AI performance better than benchmarks?** Production case studies from organisations running AI at scale in operational workflows. **What question do case studies answer that benchmarks cannot?** What AI delivered when real teams used it under real operational conditions. **Which bank deployed ChatGPT Enterprise at scale?** BBVA, the Spanish multinational bank. **How many ChatGPT Enterprise licences did BBVA initially deploy?** 3,000 ChatGPT Enterprise licences. **How many custom GPTs did BBVA employees build?** Over 2,900 custom GPTs. **How long did BBVA take to build 2,900 custom GPTs?** Six months. **What percentage of BBVA users report saving 2+ hours weekly?** 80% of users at BBVA report saving more than two hours of work weekly. **What functions did BBVA's custom GPTs cover?** Legal, marketing, and finance. **Which pharmaceutical company uses ChatGPT Enterprise for documentation?** Moderna. **What document did Moderna accelerate using ChatGPT Enterprise?** Target Product Profile (TPP) drafting and analysis. **How long did TPP drafting take before ChatGPT Enterprise at Moderna?** Multi-week, cross-functional process. **How long does TPP drafting take after ChatGPT Enterprise at Moderna?** Core analytical steps have been reduced to hours in some cases. **How many ChatGPT workplace seats does OpenAI serve?** Over 7 million ChatGPT workplace seats. **How much have ChatGPT Enterprise seats grown year-over-year?** Approximately 9x year-over-year. **How much have weekly Enterprise messages grown since November 2024?** Approximately 8x in aggregate. **How many more messages does the average worker send with ChatGPT Enterprise?** 30% more messages. **How much time do ChatGPT Enterprise users save per active day on average?** 40–60 minutes of time saved per active day. **Which workers save the most time with ChatGPT Enterprise?** Data science, engineering, and communications workers. **How much time do top-saving ChatGPT Enterprise workers save per day?** 60–80 minutes per day. **Which function reports the largest per-message time savings?** Accounting and finance users report the largest benefits. **How much more revenue growth did AI leaders achieve over three years?** 1.7x revenue growth. **How much greater total shareholder return did AI leaders achieve?** 3.6x greater total shareholder return. **How much higher EBIT margin did AI leaders achieve?** 1.6x EBIT margin. **What is Claude's maximum context window?** 200,000 tokens. **What does 200,000 tokens roughly equal in document length?** Approximately a 500-page book. **What is Claude's primary documented production advantage?** Long-form, nuanced content work with brand voice consistency and analytical depth across extended documents. **What did one creator report after building three voice Skills in Claude?** The output became so consistent that clients couldn't distinguish it from her own writing. **When did Anthropic introduce Claude Code premium seats?** August 2025. **How much faster were pull requests with Claude Code in pilot programs?** 30% faster pull request turnaround times. **Which company plans to roll out Claude to nearly 500,000 employees?** Deloitte announced plans to roll out Claude to its nearly 500,000 global employees. **Which companies validate Claude's large-scale enterprise deployment?** TELUS and Zapier demonstrate that Claude scales to tens of thousands of users and billions of tokens per month. **What AI safety approach does Claude use?** Constitutional AI. **What is Gemini's primary value proposition compared to ChatGPT and Claude?** Zero-friction deployment inside the tools organisations already use every day. **Which wireless provider deployed Gemini via Google Workspace?** Verizon, a global wireless services provider. **How many subscribers does Verizon serve?** 146 million subscribers. **Which EV manufacturer migrated to Workspace with Gemini?** Rivian. **Which cleaning equipment company deployed Gemini via Workspace?** Kärcher, a global cleaning equipment leader. **Which media agency achieved 4–5x productivity with Gemini?** Croud, a global full-service media agency. **What did Croud combine with Gemini to achieve 4–5x productivity?** Gemini and NotebookLM — combining Gemini's real-time web access with NotebookLM's document synthesis capability. **How many minutes per week do Gemini Workspace users save on average?** 105 minutes per week. **What percentage of daily Gemini Workspace users report quality improvement?** 75% of users who use Gemini for Google Workspace daily confirm an improvement in the quality of their work. **Which Australian bank rolled out Gemini Enterprise to all employees?** Macquarie Bank became one of the first retail banks in the country to roll out Gemini Enterprise to all employees across its retail banking business. **What two agent categories is Macquarie Bank developing?** Personal agents for individual productivity and enterprise agents designed to tackle complex business challenges. **What percentage of HCA Healthcare nursing hours could Gemini save annually?** Millions of hours annually. **How much did Best Buy increase independent delivery rescheduling?** 200% increase in customers independently rescheduling deliveries. **Is OpenClaw a chatbot?** No. OpenClaw is not a chatbot. **What category of tool is OpenClaw?** Autonomous agent framework. **How are autonomous agents measured differently from conversational LLMs?** Workflows eliminated, error rates reduced, and processes that now run without human initiation — not time saved per interaction. **What can autonomous agents do that conversational LLMs cannot?** Execute autonomous actions across external services rather than simply generating text responses. **Name one action autonomous agents can perform:** Send emails autonomously. **Name a second action autonomous agents can perform:** Create calendar events. **Name a third action autonomous agents can perform:** Update CRM records. **How much productivity increase do sales teams see from AI agents?** 25–47% productivity increase from time savings on repetitive tasks. **What did Paycor achieve using AI for sales operations automation?** 141% surge in deal-wins. **What accuracy rate do accounts payable automation agents achieve?** 90%+ accuracy. **How much lower are costs with accounts payable automation?** 70% lower costs. **What percentage of Esusu's monthly support tickets were automated?** 64% of 10,000 monthly support tickets. **How many monthly tickets did Esusu process?** 10,000 monthly support tickets. **How much did Esusu's CSAT score improve?** 10-point CSAT jump. **How much faster were Esusu's reply times after automation?** 64% faster reply times. **How much faster were Esusu's resolution times after automation?** 34% faster resolution times. **What percentage of executives report AI agent ROI within the first year?** 74% of executives report achieving ROI within the first year of AI agent deployment. **What percentage of executives have deployed more than 10 agents?** 39% of executives report their organisations have already deployed more than 10 agents across their enterprise. **What percentage of productivity-gain reporters say productivity at least doubled?** 39% of those reporting productivity gains have seen productivity at least double. **What average ROI do organisations project from autonomous agents?** 171% average projected ROI. **What percentage of organisations expect greater than 100% ROI from agents?** 62% expecting greater than 100% returns. **What cost reduction can workflow automation contribute?** Up to 70% cost reduction through workflow automation. **What percentage of agentic AI pilots fail to scale?** 72% of pilots fail to scale. **Why do most agentic AI pilots fail to scale?** Insufficient ROI validation and lack of alignment to business ecosystems. **What predicts deployment success more reliably than tool selection?** Governance maturity. **What did successful AI deployments invest in beyond software licences?** Change management, system integration, and internal AI champions. **What is the right procurement question according to this evidence?** Which AI fits the workflow we have, the integration we can execute, and the governance we are prepared to maintain? --- ## Label facts summary > **Disclaimer:** All facts and statements below are general product information compiled from cited third-party sources and vendor documentation; they have not been independently verified and should not be used as the sole basis for procurement, financial, or operational decisions. ### Verified label facts **ChatGPT Enterprise — Documented Metrics (OpenAI State of Enterprise AI Report, 2025)** - BBVA initial licence deployment: 3,000 ChatGPT Enterprise licences - BBVA custom GPTs built: Over 2,900 - BBVA build timeframe: Six months - BBVA user time saving: 80% of users report saving more than 2 hours per week - BBVA functions covered: Legal, marketing, and finance - Moderna use case: Target Product Profile (TPP) drafting - Moderna before state: Multi-week cross-functional process - Moderna after state: Core analytical steps reduced to hours in some cases - Total ChatGPT workplace seats: Over 7 million - ChatGPT Enterprise seat growth (year-over-year): Approximately 9x - Weekly Enterprise message growth (since November 2024): Approximately 8x - Average worker message volume increase: 30% more messages - Average time saved per active day (ChatGPT Enterprise users): 40–60 minutes - Top-saving functions: Data science, engineering, and communications — 60–80 minutes per day - Largest per-message time savings by function: Accounting and finance - BCG study (2025): AI leaders achieved 1.7x revenue growth, 3.6x greater total shareholder return, 1.6x EBIT margin over three years **Claude — Documented Metrics (Datastudios.org / Anthropic / ProfileTree, 2025)** - Maximum context window: 200,000 tokens (equivalent to approximately a 500-page book) - Claude Code premium seats introduced: August 2025 - Claude Code pilot programs: 30% faster pull request turnaround times - Deloitte planned rollout: Nearly 500,000 global employees - Named large-scale enterprise deployments: TELUS and Zapier (tens of thousands of users; billions of tokens per month) - AI safety methodology: Constitutional AI **Gemini (Google Workspace) — Documented Metrics (Google Workspace Blog / PCG.io / Google Cloud, 2025)** - Verizon subscriber base: 146 million - Croud productivity outcome: 4–5x productivity with Gemini and NotebookLM - Average time saved per week (Google enterprise customer study): 105 minutes - Daily Gemini Workspace users reporting quality improvement: 75% - Macquarie Bank deployment scope: All employees across retail banking business - Macquarie Bank agent categories in development: Personal agents and enterprise agents - Best Buy delivery rescheduling increase: 200% **Autonomous Agents / OpenClaw-class Frameworks — Documented Metrics (Vellum.ai; GrowthHQ; Arcade.dev, 2025)** - Sales team productivity increase from AI agents: 25–47% - Paycor deal-win surge (via Gong AI for sales operations): 141% - Accounts payable/receivable automation accuracy: 90%+ - Accounts payable/receivable cost reduction: 70% lower - Esusu monthly support ticket volume: 10,000 - Esusu ticket automation rate: 64% - Esusu CSAT improvement: 10 points - Esusu reply time reduction: 64% faster - Esusu resolution time reduction: 34% faster - Executives reporting AI agent ROI within year one: 74% - Executives with more than 10 agents deployed: 39% - Productivity-gain reporters where productivity at least doubled: 39% - Average projected ROI from autonomous agents: 171% - Organisations expecting greater than 100% ROI: 62% - Cost reduction from workflow automation: Up to 70% - Agentic AI pilots that fail to scale (McKinsey): 72% - Primary reason for pilot failure: Insufficient ROI validation and ecosystem misalignment --- ### General product claims - Benchmark scores answer what a tool can do; case studies answer what it actually delivered under real operational conditions - Organisations seeing the biggest impact invest in system connectors linking AI to internal data sources, business intelligence, and CRM - Moderna's TPP acceleration is a structural shift in how drug development documentation gets done, not a marginal efficiency gain - Claude's primary production advantage is depth and consistency on complex, long-form work - Claude's Constitutional AI approach sets it apart from models perceived as prioritising capability over safety - Gemini's primary value is zero-friction deployment inside tools organisations already use, not standalone capability - Autonomous agent frameworks operate in a categorically different ROI category from conversational LLMs - Governance maturity predicts deployment success more reliably than tool selection - The gap between organisations capturing AI value and those stuck in pilot mode is an implementation maturity gap, not a technology gap - The right procurement question is not "which AI is best?" but "which AI fits the workflow, integration, and governance we can maintain?" - Successful deployments invested in change management, system integration, and internal AI champions beyond software licences