The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI that maps stakeholder relationships within target accounts and generates pre-meeting briefings with context and talking points. Includes org chart inference and relationship strength scoring; distinct from CRM data management which maintains records rather than generating intelligence.
Account intelligence — AI that maps stakeholder relationships within target accounts and generates pre-meeting briefings — has reached technical maturity without achieving broad adoption. The technology demonstrably works: production deployments show 35-50% sales cycle improvement, 40-50% pipeline growth, and revenue-per-seller gains. Forward-leaning organizations and Microsoft's internal teams report measurable wins. Yet organizational adoption remains stalled: enterprise account mapping accuracy hits only 50-80%, generic AI misclassifies high-intent signals at 40% error rates, and contact data decay runs 23-30% annually. The defining tension at this leading-edge stage is structural — not a capability gap, but a data quality, integration complexity, and organizational readiness gap that incremental features do not address. Specialist vendors have consolidated around Salesforce and Microsoft ecosystems; most mainstream enterprises still rely on manual relationship mapping or basic CRM lookups.
Production deployments in advanced organizations show consistent gains. Microsoft's internal 4,000-user Copilot for Sales deployment achieved 9.4% revenue-per-seller increase and 20% higher win rates. Analytic Partners saw research time drop from 3 hours to 15 minutes per account with 40% qualified pipeline growth. Three B2B SaaS organizations deploying AI-powered buying committee orchestration reported 35% sales cycle improvement, 48% pipeline conversion lift, and contract values 31% higher with pre-demo executive engagement. Introhive's Salesforce integration reached GA with 300+ contacts auto-captured per user and reported 36% sustained win-rate improvement and 30% cross-sell growth. A Forrester study validates 495% ROI in relationship-capital-intensive verticals. Gartner's 2026 Market Guide names relationship mapping and AI-powered briefing generation as key differentiation drivers among twelve category vendors (DemandFarm, Salesforce, HubSpot, Pipedrive, Kapta, Squivr, and others). The market grew from USD 1.41B in 2025 to projected USD 2.225B by 2036.
Yet broader enterprise adoption remains constrained by structural barriers that late-April 2026 research has crystallized. Relationship mapping accuracy in production systems tops at 80% for opportunity mapping; contact data decays at 23-30% annually. Generic AI misclassifies high-intent signals at 40% error rates. Danish Lead Co's assessment of enterprise deployments documents systematic failures: buying committee mapping and organizational structure mapping lack depth sufficient for multi-stakeholder deals. Microsoft 365 Copilot shows only 1.81% conversion from subscriber base; agentic AI completes only 25-33% of multi-step sales tasks reliably. Introhive's February 2026 public documentation of adoption barriers — CRM interoperability, simplistic relationship scoring, poor usability, email domain complexity, data quality degradation — concludes that implementation and user adoption struggles create roadblocks independent of feature velocity. RAND Corporation's April 2026 meta-analysis of 65 enterprise AI initiatives found 80.3% failed to deliver business value, with 33.8% abandoned pre-production and 28.4% reaching production but failing ROI—three failure patterns: uncleaned master data, absent organizational decision-making structure between business units and IT, and use-case drift. Stanford's 2026 AI Index documents governance becoming a board-level constraint: AI incidents rose 56% YoY (362 in 2025), with embedded governance and audit traceability identified as critical to enterprise AI deployment success. The adoption plateau persists among current users: 47% report no plans to expand AI integration. The tension is no longer technical but organizational and structural: production systems work in focused contexts but require data governance infrastructure, integration architecture, and change management capacity that most mainstream enterprises have not deployed or resourced.
— TechClass enterprise AI strategy: 70% of pilots fail to reach production ('pilot purgatory'). Hallucination spectrum by task: legal 17–33%, manufacturing 44%. Proposes governance as immune system and verification pipelines for trustworthy AI.
— McKinsey 2025: 88% use AI but only 39% see EBIT impact; 21% redesigned workflows. Distinguishes org chart (hierarchy) from work chart (workflow, decisions, accountability). Explains why stakeholder briefing maps must reveal decision flows, not just structure.
— Map My Relationships deployment for D365 relationship mapping: org charts, influence mapping, and multi-threaded deal navigation completing in under one week with faster closures and live relationship visualization replacing manual research.
— Demandbase relationship intelligence: auto-capture from emails/calendars, relationship scoring by frequency/recency/depth, buying committee identification; reports 130% win-rate increase on deals >$50K.
— HALIRO B2B account intelligence platform surfaces decision-makers, sponsors, buying committees via stakeholder mapping, intent signals, and account context; directly addresses invisible stakeholder problem.
— Sphere Partners analysis: generic LLMs achieve 50% on company-specific queries; RAG improves to 80%, RAG+persistent memory to 90%+. Documents core account intelligence challenge—organizational context never in public training data.
— Technical critique of silent failures in AI systems: null-result omission, hallucination rates 22–94%, label collapse causing invisible failures. Proposes epistemic layer and evidence tracking as missing infrastructure for account intelligence systems.
— Stanford HAI 2026 AI Index documents 362 AI incidents in 2025 (up 56% YoY) reflecting governance, model failures, and scale challenges. Identifies embedded governance and audit traceability as board-level risks for AI in transaction systems; directly applicable to account intelligence deployment constraints.