Perly Consulting │ Beck Eco

The State of Play

A living index of AI adoption across industries — where established practice meets the bleeding edge
UPDATED DAILY

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.

The Daily Dispatch

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.

AI Maturity by Domain

Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail

DOMAIN
BLEEDING EDGEESTABLISHED

Real-time conversation guidance

GOOD PRACTICE

TRAJECTORY

Stalled

AI that provides live prompts, battle cards, and suggested responses to sales reps during calls and meetings. Includes contextual objection handling and competitive talk tracks; distinct from post-call coaching which provides feedback after rather than during conversations.

OVERVIEW

Real-time conversation guidance uses AI to surface talking points, objection responses, and competitive intelligence to sales reps during live calls and meetings. Unlike post-call analytics, it must operate at conversational speed—capturing buyer signals mid-dialogue and delivering contextual prompts before the moment passes.

The practice has reached good-practice maturity: GA tooling from Microsoft, Revenue.io, Salesken, Spiky.ai, and others is production-ready, and named deployments at organisations like Ernst & Young (100,000+ professionals) and Revenue.io's customer base (HPE, Nutanix, AWS) demonstrate proven ROI. The defining tension is no longer whether the technology works but whether organisations can absorb it. Deployment outcomes are strong where infrastructure and change management align—real-time guidance delivers 34%–58% win-rate improvements and 30-day ramp compression in proven deployments—yet 64% of revenue-role employees still underutilise their AI tools and 95% of enterprise AI sales pilots fail to deliver measurable returns within 6 months. Real-time guidance is proven and accessible, but rolling it out successfully remains an organisational challenge, not a technical one. The 2026 inflection point is clear: real-time guidance works, but scaling it depends on data infrastructure readiness, governance discipline, and architectural choices to contain hallucination risk in agentic systems.

CURRENT LANDSCAPE

The vendor ecosystem is mature and actively advancing agentic capabilities with critical reliability engineering now table stakes. Microsoft's Dynamics 365 Sales 2026 Wave 1 (April–September rollout) introduces Delta-first guidance in Sales Close Agent—AI suggestions on which specific actions move deals forward—representing a shift toward prescriptive, autonomous coaching. Revenue.io's Moments™ continues delivering real-time coaching notifications with phrase-triggered alerts; Liminal launched Coach AI (May 2026) grounding guidance in verified competitive intelligence; Revenue Grid has expanded its Deal Guidance and Meetings Assistance capabilities; Salesken remains the #1 ranked real-time coaching solution on the Salesloft marketplace. Emerging entrants like Spiky.ai (10,000+ users, 34%→58% win-rate improvement, -31 days ramp) and Curvo AI (live MEDDIC tracking, objection guidance) demonstrate production maturity across the ecosystem. GoNoGo's April 2026 market analysis identifies real-time call coaching as 28% of the voice AI market and the fastest-growing segment driven by SDR automation and ramp-time compression. The category continues converging with autonomous agents and forecasting as platforms absorb guidance into unified revenue-execution suites. A critical architectural threshold has crystallized: sub-200ms inference latency became the enabling constraint for in-call guidance in mid-2025, shifting the market from post-call review to real-time coaching as standard category capability. Late-June 2026 market data confirms ecosystem consolidation: Pavilion's 2026 coaching benchmark of 268 GTM teams identifies six major vendors commanding 38–7% market share each (Gong, Chorus, Clari, Mindtickle, Outreach, Second Nature), with real-time guidance now standard table-stakes across all platforms. Organizations committing to a single primary tool see 31% higher adoption than multi-tool strategies, but layering a specialist real-time coach atop a primary platform outperforms single-tool deployments by 9 adoption points.

An emerging architectural debate shapes 2026 product design: break-based coaching (prompts at natural conversation transitions, avoiding mid-dialogue disruption) vs. continuous real-time guidance (detecting objections, competitor mentions, and signals in-moment to prevent errors before they compound). Continuous architectures show stronger outcomes for new reps building muscle memory but risk cognitive overload; break-based models suit experienced reps but may arrive too late. Integration depth determines revenue attribution—native CRM hooks enable deal tracking, while loose integrations prevent outcome measurement. Reliability engineering for real-time guidance systems has emerged as the critical differentiator: Seekr's May 2026 analysis documents hallucination rates rising sharply in agentic multi-step workflows (the mode real-time guidance operates in), contradicting vendor marketing claims of improvement; OpenAI o3 hallucinates on 33% of PersonQA prompts. Multi-model verification architectures demonstrate production-grade solutions: 480M-output empirical study (Jan-Apr 2026) shows 8.3% single-model hallucination rates drop to 3.2% with verification (61% reduction), with sector-specific improvements in legal (+63%), financial (+62%), and healthcare (+59%). RAG (retrieval-augmented generation) reduces hallucination rates from typical 20-30% down to 2.5-7.5%, while tool-grounding reduces errors 65-80%—architectural choices that directly affect guidance accuracy and guardrail deployment burden. The 2026 inflection is unmistakable: vendors now standardize on verification and grounding; the constraint is no longer whether guidance can be accurate, but whether enterprises have the data infrastructure and governance discipline to deploy it reliably.

Deployment evidence remains strong at enterprise scale. Microsoft's MSX global sales team continues to demonstrate scale (6x user growth in 2 weeks, 89% of prompts triggered by contextual guidance, 400k+ annual interactions). ServiceNow deployment via Pifini (8,000 sellers, 50% onboarding time reduction) signals continued enterprise adoption momentum. StratosGrid case study shows real-time call coaching compressing new-rep ramp from 6 months to 30 days, with 4.2× faster ramp-up and 78% first-month quota attainment. Careertrainer.ai reports 35% of sales teams now using AI for real-time customer interaction guidance, with 10–20% conversion uplift and 26% reduction in ramp time; Mevak analysis (May 2026) confirms new hires with AI coaching ramp 30-40% faster, replacing traditional manager ride-alongs. NOVOSLO case data shows real-time in-call coaching lifts win rates by 19% (45.5% to 64.1% qualified conversion). Latest evidence from June 2026 reinforces deployment momentum: LTMindtree's industrial distributor deployment (900+ sellers, 330,000+ customers) shows 30-50% reduction in information search time via AI-powered Sales Excellence agent delivering contextual guidance during customer interactions, redirecting 170,000 annual hours to revenue engagement with 20-30% faster onboarding. Outreach customer Renaissance documents 26% win-rate increase, 45% deal size lift, and 11-day cycle reduction with real-time content cards during live calls. Allego's Q2 2026 research of 318 revenue enablement leaders confirms 77% report AI revenue gains, 84% improved coaching quality, and 83% reduced onboarding time. The global sales enablement market reached $4.2–6.6 billion in 2025 with 17–22% CAGR; just-in-time enablement with real-time battle cards remains the defining 2026 trend among winning teams. However, a critical reality check persists: Gartner data shows only 39% of chief sales officers report quota improvement from AI sales initiatives—meaning 61% report no improvement despite tool maturity and proven case studies. This signals that deployment outcomes, not product capability, remain the limiting factor.

Adoption barriers remain structural despite vendor maturity and deployment proof points. The most critical emerging barrier is organizational ROI accountability: MIT's GenAI Divide (June 2026) documents 95% of enterprise AI projects failed to show measurable financial returns within six months; Gartner predicts 40% of agentic AI projects will be cancelled by end 2027 due to "escalating costs, unclear business value, insufficient risk controls." Gong research (April 2026) shows 58% of enterprises stalling AI projects due to trust deficits in data security and explainability—a persistent blocking concern. Infrastructure readiness emerges as the binding constraint: Dun & Bradstreet's May 2026 survey of 10,000 organizations reveals 97% have active AI initiatives but only 5% say data is adequately ready—a 92-point gap that directly constrains real-time guidance viability. Coevera's June 2026 analysis confirms: 63% of organizations uncertain about their data practices; only 5% say AI-ready data exists. Practitioner audits document that real-time guidance still suffers from execution friction: Pifini identifies 87% of sales reps missing quotas due to setup misalignment and data quality gaps; Promptolis interviews with 30+ sales leaders reveal reps find frequent prompts distracting and generic suggestions useless, with practice-then-perform models (AI roleplay before calls) outperforming live interruptions. Innovativegroup.io analysis confirms operationalization stalls beyond pilots—workflow invisibility, incentive misalignment, and data readiness gaps prevent scaling. The capability-adoption mismatch persists: where deployments are properly implemented, outcomes reach 70% coaching effectiveness gains and 60% manager productivity improvements. New quantified ROI research (June 2026) from Allego and CSO Insights documents deployment effectiveness where adoption succeeds: 77% of orgs report AI revenue gains, 84% report improved coaching quality, 83% report reduced onboarding; organizations with disciplined real-time coaching deployments see 32.1% win-rate improvements, 27.9% quota attainment lifts, and 8.4% higher year-over-year revenue growth. Financial barriers deepen the adoption challenge: ZonFlip analysis (May 2026) documents 88% of AI pilots fail to reach production, with operations and integration consuming 80-85% of total cost of ownership (not model cost); data preparation alone runs $100K-$380K per project, with 96% of enterprises reporting costs higher than expected. ByteIota research confirms structural barriers: 79% of enterprises face significant adoption challenges; 88% of pilots never reach production with measurable ROI; only 21% have governance maturity for autonomous agents. Successful deployments follow a disciplined 4-week rollout pattern: Week 1 validates use case with top performers (proving time savings), Week 2 secures data governance, Week 3 pilots with daily standups, Week 4 locks habit and sunsets legacy processes. The capability-adoption mismatch persists: real-time guidance success depends primarily on organisational discipline, implementation rigor, data infrastructure readiness, and governance discipline to contain hallucination risk in agentic systems—not technological maturity.

TIER HISTORY

ResearchJan-2022 → Jan-2022
Bleeding EdgeJan-2022 → Jan-2024
Leading EdgeJan-2024 → Jul-2024
Good PracticeJul-2024 → present

EVIDENCE (106)

Wingman (Clari Copilot) - GTM LabzProduct Launches

— Third-party review: Wingman listens to live calls via Zoom/Teams/phone and displays real-time coaching cues on-screen (battlecards, monologue alerts, deal intelligence). New rep ramp acceleration 30-40% with annotated call libraries; 5X cheaper than enterprise alternatives.

— Production deployment across 900+ industrial distributors: AI-powered Sales Excellence agent delivers contextual in-call guidance via Teams, reducing information search 30-50%, redirecting 170,000 hours annually to customer engagement, accelerating onboarding 20-30%.

— Named customer (Renaissance) demonstrates in-call coaching with content cards delivering guidance during live calls: 26% win rate increase, 45% deal size increase, 11-day cycle reduction, 60% seller productivity gain.

— CRO-authored market guidance: Pavilion 2026 benchmark of 268 GTM teams shows real-time coaching now standard offering across six major vendors (Gong 38%, Chorus 22%, Clari 14%, Mindtickle 11%, Outreach 8%, Second Nature 7%). Single-tool commitment achieves 31% higher adoption.

— Practitioner deployment model for 2027: AI agents operate as always-on second pair of eyes (drift detection, skill diagnosis, next-action recommendation). Documented outcomes: 14-22% quota lift, 2.1x faster new-AE ramp, 9-13pp forecast accuracy improvement.

— Establishes real-time coaching as distinct emerging category: live coaching linked to 30% quota attainment increase over reactive training. Contrasts with 87% content forgetting rate and shows reinforcement loops explain superior outcomes of continuous guidance.

— Independent research (318 revenue enablement leaders): 77% report AI revenue gains, 84% improved coaching quality, 83% reduced onboarding, top adopters report 45% higher win rates and 50% cycle reduction. Demonstrates coaching adoption maturity and quantified ROI.

Sales Prospecting StatisticsAdoption Metrics

— CSO Insights research quantifies real-time coaching effectiveness: 32.1% win rate improvement, 27.9% quota attainment lift, 8.4% higher year-over-year revenue growth. Signals demonstrated ROI driving adoption decisions.

HISTORY

  • 2022-H1: Early commercial maturity with major enterprise platforms launching real-time guidance features (Microsoft GA April 2022), specialist vendors reporting strong adoption metrics, but significant technical barriers around conversational AI reliability and evaluation methodology limit broader uptake.
  • 2022-H2: Vendor maturity consolidates with expanded feature offerings (Dynamics 365, Revenue.io launching new coaching products) and maintained analyst confidence in 2025 mainstream adoption. Adoption gap widens: only 27% of orgs have conversational capabilities despite 75% recognizing need; most companies undeutilizing existing sales tech; conversational AI reliability remains unresolved core challenge.
  • 2023-H1: Vendor landscape expands with named customer deployments and published case studies (Salesken healthcare 60% conversion lift, banking 200%+ productivity gains). Outreach launches live battlecards for real-time competitive guidance. Market adoption accelerates: Salesken achieves top-tier G2 rankings (9.93/10, 327+ reviews). Adoption barriers persist: organizational readiness and sales tech stack integration complexity limit broad deployment despite vendor capability maturity.
  • 2023-H2: Major platform vendors (Microsoft, Outreach, Revenue.io, Salesken) ship real-time guidance features with sustained product development. However, deployment obstacles emerge: regional rollout complexities in Dynamics 365 Sales Copilot, misalignment between vendor maturity and customer organizational readiness. Industry research highlights persistent gap—vendors capable but enterprises unprepared, lacking governance frameworks, technical expertise, and integration readiness. Critical concerns surface around AI governance, 'shadow AI' proliferation, over-reliance risks, and human-connection limitations in automated guidance. Real-time guidance moves from proof-of-concept to limited enterprise deployment with clear capability-adoption mismatch.
  • 2024-Q1: Platform consolidation continues with Microsoft advancing Copilot for Sales (unified Teams panel with contextual CRM data, GA April 2024) and specialist vendors maintaining deployment momentum (Zime AI deployments at Bureau Inc., Versa Networks showing 1-hour daily time savings). Adoption barriers remain organizational rather than technical—27% deployment rate persists despite vendor capability maturity and proven outcomes, indicating integration complexity and governance readiness as key limiting factors for broader uptake.
  • 2024-Q2: Real-time conversation guidance reaches market inflection with Microsoft Copilot for Sales achieving production GA (June 2024) and Ernst & Young deploying across 100,000+ professionals globally. Forrester analyst coverage expands with Q2 2024 Wave evaluation of real-time revenue execution platforms. Yet critical DLA Piper study documents 48% AI project failure rate with data privacy, governance, and customer concerns as primary blockers—underscore organizational readiness barriers persist despite vendor capability advancement and proven deployment outcomes at scale.
  • 2024-Q3: Vendor platform expansion accelerates with Revenue.io deploying generative AI for real-time guidance across 100+ named customers (HPE, Nutanix, AWS) and ValueSelling Associates extending eValuePrompter to Dynamics 365 for in-call methodology guidance. Broader organizational appetite for AI-guided conversation support strengthens—CallMiner data shows 87% of CX leaders view generative AI as key to customer experience strategy. Yet structural adoption barriers persist: CRM implementation failure patterns (lack of internal support, underestimated cultural change, poor leadership) continue limiting broad uptake of advanced features including real-time guidance integration.
  • 2024-Q4: Specialist real-time guidance vendors consolidate platform integrations with Salesken maintaining #1 real-time coaching rankings on Salesloft marketplace and communication coaching platform Poised extending real-time feedback support to Teams and Zoom. Microsoft Dynamics 365 Sales release wave 2 focuses heavily on Copilot AI features and sales pipeline insights, signaling continued platform investment in guided selling infrastructure. Market remains characterized by vendor maturity and feature parity against persistent organizational readiness gaps in governance, integration, and change management—the capability-adoption mismatch deepens rather than narrows.
  • 2025-Q1: Broader sales AI adoption accelerates with surveys showing 92.5% of sales professionals using AI daily and 81% of teams investing in AI tools; reps report 2-3 hours daily savings from AI-assisted workflows. Real-time conversation guidance specifically shows no major new deployments or vendor launches in early 2025, suggesting market maturation with incremental feature additions rather than category-defining innovations. Adoption remains concentrated in enterprise and mid-market segments; organizational governance and integration complexity continue limiting broader uptake despite vendor feature maturity.
  • 2025-Q2: Vendor platform expansion continues with Microsoft Dynamics 365 Sales 2025 Wave 1 advancing real-time guidance and manager coaching features (May 2025), Revenue.io launching Prompt Favorites for in-call prompt reuse (April 2025), and new entrant Scorchsoft introducing SalesChamp Live real-time coaching (May 2025). Expanded tooling support via Copilot Studio for building custom conversation training agents signals evolving implementation maturity. Consulting projections cite 15-25% conversion improvement and 30-50% faster rep ramp times, yet emerging mid-2025 analysis warns of AI vendor lock-in risks, underscoring persistent organizational concerns about platform dependency alongside unresolved adoption barriers in governance and integration.
  • 2025-Q4: Late-2025 data confirms persistent capability-adoption mismatch despite continued vendor platform maturity. Quantified deployment outcomes show real-time coaching can improve coaching effectiveness by 70% and manager productivity by 60%, yet industry-wide research documents 70-85% of AI deployments fail to meet expectations with only 1% of organizations scaling beyond pilots. Implementation failures cluster around data quality, goal-practice misalignment, insufficient change management, and wrong success metrics—all organizational rather than technical. Vendor roadmaps remain committed to real-time guidance features while organizational barriers to scaling remain structurally unresolved, confirming that real-time conversation guidance remains a competitive advantage for well-resourced, mature sales organizations rather than industry-standard practice.
  • 2026-Jan: Early 2026 data intensifies evidence of adoption breadth constraints. Clari + Salesloft report reveals 64% of revenue-generating employees not utilizing full AI tool capabilities, with $56B annual productivity drain; generational divides in adoption widen performance gaps. Industry analysis documents 95% of enterprise AI sales pilots fail to deliver meaningful ROI, citing "Admin Drag," integration complexity, and organizational workflow misalignment as structural barriers. Vendor ecosystem remains stable with incremental feature additions (Prompt Favorites, Copilot Studio scenario libraries) while deployment maturity stagnates—capability-adoption gap persists despite vendor platform maturation.
  • 2026-Feb: Vendor feature launches accelerate with Revenue.io introducing AI-Powered Sales Scorecards for real-time performance evaluation during calls. Independent verified deployments of Poised platform across Eventbrite and Scoop confirm ongoing adoption. Critical assessments of real-time guidance maturity highlight platform maturity (real-time coaching effectiveness, objection handling support) but document persistent barriers: Salesforce lock-in constraints, audio reliability limitations, complex integration requirements, and MIT research confirming 95% of AI sales spending produces zero measurable ROI. Market convergence signals suggest real-time guidance integrating with forecasting and autonomous agent ecosystems. Structural adoption barriers remain unresolved despite continued vendor platform innovation.
  • 2026-Apr: Market evidence sharpens the practice-then-perform vs. in-call guidance debate: critical assessments document reps finding real-time prompts distracting and suggestions generic, with AI roleplay before calls producing more reliable outcomes than live interruptions. Microsoft MSX deployment of Copilot at scale (6x user growth in two weeks, 400k+ annual interactions, 89% of prompts triggered contextually) stands as the clearest at-scale production signal, while the enablement market reached $4.2–6.6B at 17–22% CAGR. The "HUD model" transition from post-call autopsy to real-time assistance is broadly cited as the defining shift, with a sub-2-second latency requirement confirmed as the technical threshold for maintaining conversational flow.
  • 2026-May: Microsoft Dynamics 365 Wave 1 shipped Sales Close Agent with Delta-first guidance — autonomous prescriptive suggestions on deal-advancing actions — marking a shift from reactive prompts to agentic deal orchestration. Liminal launched Coach AI grounding real-time guidance in verified competitive intelligence, while market analysis of Siro's halftime-mode vs. continuous coaching architectures confirms reps benefit most from in-moment error prevention over break-based prompts, with native CRM integration as the key to revenue attribution. Revenue.io Moments (patented system with three granted US patents) published RSC Health as a named customer with 50% faster ramp and 95% compliance-fine risk reduction, adding documented IP defensibility to the GA vendor landscape. Adoption data from Careertrainer.ai confirms 35% of sales teams now use AI for real-time interaction guidance with 10-20% conversion uplift, but Gartner data shows only 39% of chief sales officers report quota improvement from AI sales initiatives — meaning 61% report no improvement despite tool maturity — ZonFlip documents 88% of AI pilots never reaching production (operations and integration consuming 80-85% of TCO), and a D&B survey of 10,000 organisations finds only 5% say their data is adequately ready despite 97% running active AI initiatives, reinforcing that the binding constraint is infrastructure and execution discipline, not technology capability.
  • 2026-Jun: Critical reliability analysis shifts focus to hallucination risks in agentic real-time guidance systems. Seekr whitepaper (May 27) documents hallucination rates rising sharply in agentic multi-step workflows, contradicting vendor marketing; OpenAI o3 hallucinates 33% of PersonQA prompts; GPT-5.5 marketing claimed 60% reduction but system card reports only 3% response-level reduction. Coevera's enterprise CRM guide predicts 40% agentic AI projects cancelled by end 2027 due to escalating costs and unclear ROI; identifies data readiness as critical gap (97% with initiatives, 5% say data ready). Empirical mitigation evidence emerges: 480M-output study (Jan-Apr 2026) shows multi-model verification reduces hallucination 61% (8.3% → 3.2%), with sector-specific improvements (legal +63%, financial +62%, healthcare +59%). New product entrants (Spiky.ai 10,000+ users, Curvo AI live MEDDIC tracking) confirm ecosystem maturity, while Darwin AI's 8-tactic framework documents specific outcomes (+31% objection handling, +27% battle card effectiveness, 45% late-stage conversion lift). MIT GenAI Divide shows 95% enterprise AI projects failed to deliver ROI within 6 months; 74% CIOs report role at risk if AI gains don't materialize. Large-scale production evidence reinforces deployment momentum: LTM's AI-powered Sales Excellence agent across 900+ industrial distributors redirected 170,000 annual hours from information search to customer engagement with 30-50% search-time reduction and 20-30% onboarding acceleration; Outreach customer Renaissance documents 26% win-rate increase, 45% deal size lift, and 11-day cycle reduction with real-time content cards. Allego's survey of 318 revenue enablement leaders confirms 77% report AI revenue gains and 84% improved coaching quality. Pavilion's 2026 benchmark of 268 GTM teams establishes Gong (38%), Chorus (22%), Clari (14%), Mindtickle (11%), Outreach (8%), and Second Nature (7%) as the six-vendor landscape for real-time coaching, with single-tool commitment yielding 31% higher adoption than multi-tool strategies. The inflection is clear: real-time guidance capability is mature and proven (34%→58% win rates in deployments), but adoption scaling depends entirely on data infrastructure readiness, governance discipline, architectural reliability engineering (multi-model verification, RAG, tool-grounding), and organizational willingness to sustain 4-week disciplined rollout patterns—not on further vendor innovation.

TOOLS