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 generates onboarding content and creates personalised learning paths for new employees based on role and experience. Includes automated welcome material creation and adaptive onboarding journeys; distinct from L&D which serves ongoing development rather than initial onboarding.
AI-driven onboarding automation has reached good-practice maturity: vendor platforms (Workday Sana, SAP SuccessFactors Joule, 15+ competing solutions) ship GA features for personalised journeys, automated provisioning, and task completion at scale. Named deployments confirm ROI: McCarthy Building Companies achieved 10% vs 18% industry turnover through SAP integration; HR Path (2,500-person firm) saves 20 hours/month on leave automation; financial services zero-touch provisioning reduced account setup from 3 days to 2 hours. Research synthesis (Brandon Hall, SHRM, Gartner) shows 25-40% time-to-productivity gains, 82% retention improvement, and per-hire cost reductions ($4,100→<$1,500). Yet the practice stalls at adoption plateau: Q1 2026 adoption metrics show 48% exploring/piloting, 31% operationally deployed, 25% paused or discontinued in past 24 months. The real constraint is no longer technology—it is organisational execution: governance gaps (97% of automation decisions lack documented judgment frameworks), adoption plateau (only 8.6% have agents in production; 95% of pilots yield zero P&L impact), and sustainability barriers (40% of non-managerial employees report AI adds no time savings despite platform deployment). The practice is accessible, proven, and economically justified—but scaling from pilot to sustained delivery requires operational discipline most organisations lack.
Platform ecosystem maturity is undeniable: Workday Sana (11,500+ customers, 300+ onboarding-specific skills, 90% early-adopter penetration in 40 days), SAP SuccessFactors Joule (70-87% time savings, named at Timken, Delta, American Honda), and 15+ competing vendors deliver GA-level capabilities across enterprise, mid-market, and SMB tiers. Named deployments confirm near-term efficiency gains: McCarthy Building Companies (8,000 emp) reduced turnover from 18% to 10% via SAP; HR Path (2,500-person consulting firm) saves 20 hours/month on leave automation and 2x job description speed; financial services deployments achieve 45% onboarding productivity lift through zero-touch IT provisioning (3 days→2 hours). Market economics validate unit ROI: per-hire costs compress ($4,100→<$1,500 with AI), time-to-productivity gains reach 21-31 days, and SHRM data shows 89% 90-day retention (vs 72% unstructured), with $2.3M avoided turnover offsetting $2.5M investment over 3 years. Yet adoption has stalled at the pilot-to-production boundary: Q1 2026 survey shows 48% exploring/piloting, 31% operationally deployed, and critically, 25% paused or discontinued within the past 24 months. Production-readiness remains the real barrier: only 8.6% of organizations have AI agents in production vs 63.7% with no formalized initiative; 95% of generative-AI pilots deliver zero measurable P&L impact (MIT NANDA). The core limitation is governance: successful deployments rely on documented judgment frameworks defining what agents can decide autonomously vs escalate, but 97% of automation decisions currently lack this rigor. Workday's shift to metered consumption (Flex Credits per action) introduces new friction: customers spending 2 years building better automation workflows now face per-interaction costs, creating budgetary friction that slows organizational adoption. Practitioner evidence highlights the execution gap: knowledge quality (not model sophistication) is the #1 implementation variable; accuracy thresholds of 90%+ are non-negotiable (below that, adoption collapses permanently); and organizations report bottleneck shifting rather than elimination—setup automation reduces per-hire administrative time (9 hours→2 hours) but context knowledge transfer (architectural questions, cultural integration) still demands 5-10 hours of senior-staff time. Data quality degrades in production (95%→62% accuracy), integration costs remain substantial ($140K-$350K over 4-6 months), and identity lifecycle coordination remains constrained (47% struggle with infrastructure access, 43% report >1 week provisioning delays, 70%+ manually re-key data into multiple backend systems). The human dimension remains critical and unmeasured: only 12% of employees strongly agree their organization onboards well, and automation cannot replicate the recognition, belonging, and safety signals that reduce early attrition. Regulatory headwinds add friction: Illinois, Colorado, and EU AI Act provisions create compliance uncertainty and adoption hesitation (57% of affected HR professionals do not understand requirements). The practice exemplifies bifurcated adoption: vendor capability and early-adopter case studies advance rapidly, but broad organizational scaling, measurable sustained productivity impact, and regulatory compliance remain constrained by governance gaps, data readiness, and organizational execution discipline rather than platform availability.
— MeBeBot field analysis of 12 months HR AI deployment: 90%+ accuracy non-negotiable (below threshold adoption collapses); knowledge quality not model sophistication is #1 variable; onboarding identified as high-ROI area with unexpected benefits. Only 34% enterprises report measurable financial impact.
— Q1 2026 survey of 250+ HR leaders at 1,000+ employee orgs: 48% exploring/piloting, 31% operationally deployed, 15.5% not using AI, 25% paused/discontinued in past 24 months. Shows adoption plateau: roughly half exploring, third deployed, quarter actively pausing initiatives.
— Brandon Hall Group, SHRM, Gartner synthesis reports 25-40% time-to-productivity reduction with AI onboarding vs manual (Brandon Hall), 82% retention improvement, only 32% of HR teams adopted as of 2025 (lags recruiting 51%), per-hire costs $4,100→<$1,500, 30-50% HR admin hour reduction.
— Practitioner analysis of critical governance gap: Workday provides rails (security, policies) but customers must define judgment layer (business rules, escalation, context logic). Flex Credit metered consumption forces ROI discipline. Messy judgment becomes credit waste at scale—root cause of automation failure.
— McCarthy Building Companies (8,000 employees) deployed SAP SuccessFactors Onboarding achieving 10% turnover vs 18% industry average (55% reduction). Solved critical visibility gap; management now has complete skills and project history visibility enabling better staffing decisions.
— Named financial services firm reduced account provisioning 3 days→<2 hours, achieved 45% onboarding productivity lift, 80% error drop. Gartner: 48% of zero-touch adopters achieved 60% time reduction; Forrester: 80% improved first-week productivity; IDC: 72% support ticket reduction.
— HR Path Group (2,500 employees, 28 countries) deployed SAP SuccessFactors Joule for onboarding: simplified new-hire guidance, 20 hours/month leave automation savings, job descriptions 2x faster (1 hour→minutes). Full rollout across 2,500 employees with measured efficiency gains.
— Critical maturity signal: only 8.6% of companies have AI agents in production vs 63.7% with no formalized initiative; MIT NANDA found 95% of gen-AI pilots deliver zero P&L impact. Five structural barriers: diagnosis gap, architecture deferral, data quality, governance overhead, missing production infrastructure.