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

Education administration & admissions automation

LEADING EDGE

TRAJECTORY

Advancing

AI that automates educational administrative processes including admissions screening, enrolment management, and institutional workflows. Includes application evaluation support and process automation; distinct from learning analytics which analyses student performance rather than administrative operations.

OVERVIEW

Education administration and admissions automation has split into two distinct maturity tracks. Routine operational tasks — chatbot-based recruitment, transcript processing, enrolment communications — now run in production at forward-leaning institutions with documented ROI, placing that slice of the practice firmly in deployable territory. High-stakes algorithmic screening of applicants remains far more constrained: peer-reviewed research consistently documents demographic bias in ML admissions models, and institutions face reputational and legal exposure that keeps human oversight non-negotiable. This bifurcation defines the practice's leading-edge status. The tooling ecosystem for low-stakes automation is proven; Georgia State's chatbot handles 99% of 50,000 student messages at 94% satisfaction, and CRM-driven workflows deliver measurable efficiency gains. But the harder problem — letting algorithms influence who gets admitted — is advancing cautiously at best. Most institutions have not begun, and those experimenting keep humans firmly in the loop. The question for adopters is not whether to automate administrative workflows, but how far toward consequential decision-making that automation can responsibly extend.

CURRENT LANDSCAPE

Low-stakes admissions automation has reached production maturity at a vanguard of institutions. Chatbot platforms like Mainstay (formerly AdmitHub) operate at scale across dozens of campuses, while AI-powered transcript processing tools report productivity gains exceeding 500% with 99.3% accuracy. Concrete ROI is documented: Bakersfield College saved over USD 2M, Long Beach City College achieved 10x return, and Southeast Missouri State recovered 182 staff hours monthly. Caltech and Virginia Tech deployed AI for essay evaluation and interview scoring in early 2026, though both maintain mandatory human review — a pattern that defines how institutions are navigating the boundary between operational efficiency and consequential judgment.

The vendor ecosystem is consolidating around CRM and SIS platforms with AI-driven features now standard. Workday Student manages 5.8M+ student records across 200+ institutions globally; Q1 2026 saw 26 new Ellucian SaaS SIS/ERP go-lives (Aurora, Mohamed bin Zayed AI University, Norwich), with named migrations ongoing at Metro State Denver, Clemson (July 2026), Johns Hopkins (summer 2027), and Barnard (2028). Ellucian Student platform launched April 2026 as unified AI-native SIS/HCM/Finance integration. Salesforce EDA cut admissions response times by 26%, while HubSpot CRM delivered 10x event registration uplift at University of San Diego. March 2026 market surveys show leading enrollment platforms (Slate 9.1/10, Element451 8.8/10) bundling AI lead qualification, workflow automation, and CRM integration as baseline expectations. Institutionally, adoption sentiment has shifted measurably: a March 2026 survey across 300+ institutions found 66% institutional AI adoption (up from 49% one year prior), with 51% citing Marketing/Admissions/Enrollment as their top benefit area. In student-facing research, 44% of admissions offices now deploy AI for recruiting/admissions with reported ROI of 6-10 staff hours saved weekly and 1-6 month payback. Crucially, demand is being driven from both supply and demand sides: EAB survey of 5,000+ high school students finds 46% now use AI in college search (up from 26% in spring 2025—a 77% YoY increase), with 18% removing colleges from consideration based on AI recommendations, forcing institutional adaptation of enrollment funnel strategies. Globally, India's state-level deployments — CSM Technologies' SAMS platform across Odisha and Bihar, Uttar Pradesh's Samarth portal — signal that automation demand extends well beyond North American higher education. Market analysts project the admission management system market growing from USD 1.58B in 2024 to USD 4.55B by 2034, driven by cloud adoption and AI feature standardization.

May 2026 activity confirms system-level adoption momentum: the University System of Georgia (25 public institutions) selected Workday ERP for July 2028 go-live, signaling major multi-institutional commitments driving vendor consolidation. Concrete workflow automation outcomes continue: a community college achieved 3x scholarship applications per student (1.6 to 5.1), $2.4M additional aid, and 58% reduction in counselor time via SIS-integrated automation—demonstrating that targeted workflow automation drives measurable revenue and efficiency gains. Third-party research (EAB, Civitas, Gartner) validates scaling: institutions automating enrollment workflows achieve 28-42% workload reduction, 11-18 percentage-point retention improvement, and documented incremental revenue impact. However, institutional barriers are now well-documented: integration complexity (15-30 direct system integrations per institution, underestimated by 2-3x) is the largest hidden cost in SIS migrations; consulting intelligence based on 18 university leader interviews documents $20M cost examples and 80% SaaS migration failure rates, with institutional debt—not vendor lock-in—as the primary adoption constraint. A critical market insight emerged from analyzing 7M+ student journeys across 50+ institutions: enrollment success correlates with multi-source engagement patterns, not single-source acquisition, challenging the first-source attribution methodology that many automation systems rely on and revealing that 53% of prospective students have already narrowed college options before institutional CRM systems capture them. These findings suggest that post-inquiry automation, while delivering measurable efficiency gains at the workflow level, operates downstream of the major decision-making funnel.

High-stakes screening remains the sticking point. AAAI 2026 research demonstrated that ML admissions models degrade when applicant pool composition shifts, adding technical reliability concerns to the established fairness objections. A new complication has emerged: a longitudinal study of 81,663 applications found LLM use in admissions essays accelerating sharply, with lower-SES applicants using them more but facing steeper admission penalties — blurring the line between applicant authenticity and algorithmic evaluation. With applicant volumes up 32% since 2020, the operational pressure to automate is real, but institutions remain unwilling to cede screening authority to systems whose failure modes carry legal and reputational consequences. Research also confirms that response speed is a material conversion driver — institutions automating lead routing and initial communications see measurable enrollment uplift, reinforcing demand for low-stakes workflow automation even as high-stakes screening remains constrained.

TIER HISTORY

ResearchJan-2019 → Jan-2019
Bleeding EdgeJan-2019 → Jan-2024
Leading EdgeJan-2024 → present

EVIDENCE (110)

— EAB analysis of 7M+ student journeys across 50+ institutions reveals multi-source engagement, not single-source acquisition, predicts enrollment conversion—challenging first-source attribution methodology and automation assumptions.

— Consulting firm survey (1,000+ higher ed stakeholders at Ellucian Live) identifies top implementation barriers: process preservation, customization handling, integration complexity, discovery scope, and partner selection.

— Strategic analysis (Capture Higher Ed's Enrollment Engagement Report): 53% of prospective students narrow college options before inquiry submission, revealing pre-funnel research invisible to institutional CRM tracking systems.

— University System of Georgia (25 public institutions) selects Workday ERP with July 2028 implementation timeline and Deloitte partnership, signaling major ecosystem shift toward cloud-based administrative systems.

— Carthage College (private liberal arts) deployed Workday Student for production student-facing onboarding automation across financial aid, meal plans, emergency contacts, parking, and medical forms.

— Intelligence brief (18 primary interviews with university leaders) documents $20M integration cost barriers and 80% SaaS migration failure rate—institutional debt, not vendor lock-in, is primary adoption constraint.

— Research university implemented Workday for unified student experience, reporting enhanced registration, improved financial aid processes, streamlined HR workflows, and better data analytics across operations.

— Community college (4,800 students) achieved 3x scholarship applications per student (1.6→5.1), $2.4M additional aid, 58% counselor time reduction via SIS-integrated automation and SMS reminders.

HISTORY

  • 2019: Early adoption of chatbot-based recruitment and retention platforms (AdmitHub 50+ customers) and first-generation financial aid verification automation (Campus Management 50% processing time reduction at UW-Madison). Parallel emergence of concerns about algorithmic bias in admissions decisions as a regulatory and equity risk.
  • 2020: Continued vendor expansion (AdmitHub 90+ institutions, $7.5M Series A from Google/Salesforce) and deployment of chatbot-based student support during COVID-19 pandemic (Common App partnership). Critical setback: University of Texas at Austin discontinued its GRADE algorithmic screening system after 7 years despite documented 71% efficiency gains, citing bias concerns. Signals market divide between viable low-risk administrative automation and constrained high-stakes screening use cases.
  • 2021: Consolidation of market into two adoption tracks. Low-risk chatbot and CRM-based admissions automation scaled: AdmitHub $14M Series B, 3M+ student interactions, 30% summer melt reduction at Georgia State; Indian edtech platform achieved 30% lead-to-enrollment gains via LeadSquared; Australian deployment via Salesforce EDA. Infrastructure reliability concerns emerged: UC/Cal State admissions portals crashed during peak application deadlines. Regulatory scrutiny intensified: New America documented how predictive analytics perpetuate racial inequities; UK Government AI Barometer identified bias as primary adoption barrier. High-stakes algorithmic screening remained constrained by equity concerns.
  • 2022-H1: Continued low-risk automation scaling (AdmitHub/Mainstay 100+ institutions); emergence of alternative admissions innovation (Concourse portfolio-based system, 658+ pilot participants, $18.1M scholarships). Academic research renewed focus on algorithmic bias: MIT documented fairness gaps up to 21% in ML explanation methods for admissions; Brookings analysis showed enrollment algorithms reduce aid for low-income students. Institutional response: NSF/Amazon funded $1M project for fair algorithms in graduate admissions. Market divide between low-stakes automation (scaling) and high-stakes screening (constrained by fairness requirements) solidified, with investment shifting toward equity-focused solutions.
  • 2022-H2: Institutional adoption of chatbot automation continued (Texas A&M System deployment of AdmitHub), confirming sustained multi-institution rollout through year-end. However, adoption barriers persisted: survey research found majority of HEIs not ready for AI chatbot deployment, highlighting persistent organizational, technical, and trust gaps in the market despite vendor traction.
  • 2023-H1: Low-risk automation platforms demonstrated sustained deployment scale (Family Legacy Foundation AdmitHub chatbot for 1,000+ students annually, Antioch University text automation achieving 62% lead-to-enrollment gains). Institutional investment in enrollment data infrastructure accelerated (Northeastern University multi-year modernization via Snowflake and dbt). Research on fairness limitations intensified: UT Austin study found ML models for student-success prediction systematically less accurate for racially minoritized students with ineffective mitigation; Brookings documented enrollment algorithms reduce scholarships and perpetuate discrimination. Counterbalancing innovation: Cornell research demonstrated ML admissions models can outperform SAT-based screening while maintaining demographic parity. Persistent market divide: low-stakes automation scaling with clear ROI; high-stakes algorithmic screening adoption constrained by fairness and equity barriers.
  • 2024-Q1: Generative AI impact on admissions became visible: institutional caution toward algorithmic screening intensified (SMU explicitly discontinued personal statements due to ChatGPT but rejected AI-based screening), while low-risk automation continued evolving (UC Irvine integrated ChatGPT-powered chatbots with 96% accuracy, Columbia College of Missouri deployed AI funnel management improving staff morale). Landscape survey data showed early-stage adoption: Kaplan survey found only 14% of admissions officers use AI in their work, 85% lack GenAI policies, 9% use detection software—indicating institutional uncertainty despite growing vendor capabilities. Peer research confirmed persistent barriers: technical, ethical, and resource challenges limit scaling beyond specialized low-risk use cases.
  • 2024-Q2: No-code and specialized automation tools entered production deployments: AppSheet case study showed significant processing time reductions in university admissions workflows; Freedom OCR platform gained adoption for transcript extraction and SIS integration. Institutional adoption of bot-based admissions automation continued scaling per UCAS conference case studies. High-stakes algorithmic screening remained cautious: University of Miami piloted AI for application volume handling (50,000+ applications), while IE University deployed AI for major selection—but institutional hesitation persisted. Critical assessment documented negative impacts: 2024 admissions cycle revealed increased opacity in AI-driven selection processes, FAFSA processing failures, and 40% decline in financial aid completion. Market division reinforced: low-stakes automation demonstrated value; high-stakes screening constrained by equity and transparency concerns.
  • 2024-Q3: Production deployment of low-risk automation expanded: Georgia Tech screened 60,000+ applications annually with expanded admissions team, and CollegeVine onboarded 50 institutional partners for agentic AI recruiting. Peer-reviewed research from AERA demonstrated that algorithmic bias in student-success prediction models persists despite mitigation efforts, with disparate accuracy across racialized groups. Adoption survey data (Martin Center, James G. Martin Center) indicated 50% of US admissions offices use some form of AI, but critical assessment emphasized transparency gaps and regulatory risks. Market division solidified: low-stakes automation (lead management, chatbots, transcript processing) deployed at scale with documented ROI; high-stakes algorithmic screening remained constrained by well-documented fairness gaps and institutional caution. Vendor consolidation continued around recruitment and support tools rather than algorithmic screening.
  • 2024-Q4: SIS vendor ecosystem reached inflection point: 28 institutions deployed Workday Student in Q4, adding to $15.3B-to-$32B market growth forecast. Adoption survey (Liaison, Oct 2024) confirmed institutional reliance on conversational AI (73%) while strategic adoption of predictive (41%) and prescriptive (17%) AI remained limited. Expert panel discussion (Inside Higher Ed, Dec 2024) highlighted AI potential for credit transfer automation, tuition discounting, and student success—but implementation barriers persisted. Infrastructure reliability risks emerged during peak admissions processing (Ellucian outage, Nov 2024). Market remained bifurcated: low-stakes automation (chatbots, lead routing, transcript processing via Freedom OCR) continued scaling with documented ROI; high-stakes algorithmic screening constrained by persistent fairness research and institutional caution. No major policy shifts or regulatory changes in window, but transparency and liability concerns remained top institutional barriers.
  • 2025-Q1: Low-risk automation continued demonstrating ROI in early 2025: six institutions reported concrete gains (Empire State 25% engagement increase and 4% retention, Bakersfield College $2M+ savings, Long Beach City College 10x ROI). Mainstay (formerly AdmitHub) remained in production use with expanded feature set. Institutional sentiment shifted measurably: Acuity Insights survey of 160 admissions leaders found 51% believe AI will transform applicant evaluation, reflecting confidence growth in the market. However, fairness research persisted as a critical counterweight: new peer-reviewed study analyzed real university admissions data (11,600 students) confirming ML models exhibit demographic bias—white and non-first-gen students incorrectly admitted at higher rates despite test-optional policy. Practitioner discourse emphasized bias risks and mitigation strategies. Market outlook remained stable: low-stakes chatbot and CRM-based automation gained confidence and ROI evidence; high-stakes algorithmic screening continued constrained by documented fairness gaps and institutional caution.
  • 2025-Q2: SIS vendor consolidation accelerated with Workday Student deployments at LSU (Spring 2025) and Colby College (April 2025), though implementation challenges emerged around UX design and feature completeness despite backend efficiency gains. Podcast-driven market landscape analysis (April 2025) provided concrete deployment metrics: NYU 120k applications, UTexas +24% growth, UWash +57% growth. Peer-reviewed systematic review (June 2025) on AI in applicant screening analyzed current tools and bias concerns. Practitioner critical assessment (May-June 2025) reinforced focus on fairness audits, explainable AI, and hybrid human-review models. Market divide persisted: low-stakes automation (chatbots, transcript processing) continued scaling with proven ROI; high-stakes algorithmic screening remained constrained by peer-reviewed bias research and institutional fairness concerns. Regulatory environment and Supreme Court ruling implications (2023) kept institutional risk calculus cautious around autonomous AI screening systems.
  • 2025-Q3: Policy-level automation innovation demonstrated impact: UT Austin pilot combining automatic admissions eligibility with proactive financial aid guarantees nearly doubled enrollment for eligible low-income students (23% to 43%), validating automated policy-implementation systems. Global market expansion accelerated: CSM Technologies deployed Student Academic Management System (SAMS) in Indian states (Odisha, Bihar) for centralized admission automation; Uttar Pradesh government launched Samarth portal for state-level university admissions consolidation. Market analysis showed USD 1.58B admission management system market (2024) projected to reach USD 4.55B (2034) at 9.8% CAGR, indicating sustained institutional and vendor investment in automation infrastructure. Implementation barriers persisted: Workday deployments continued facing UX and feature completeness challenges despite operational efficiency gains. Fairness research and practitioner guidance remained consistent through Q3: focus on audits, explainable AI, and hybrid models to mitigate bias in higher-stakes screening systems.
  • 2025-Q4: Low-risk automation continued demonstrating concrete ROI with peer-reviewed research validation: October 2025 systematic review synthesized AI in academic applicant screening across undergraduate to medical residency programs, confirming widespread adoption. Institutional deployments at scale reached production maturity: Georgia State's AdmitHub chatbot processed 99% of 50,000 student messages with 94% satisfaction; Salesforce EDA implementations cut admissions response time 26%; Southeast Missouri State saved 182 staff hours monthly via AI chatbots. Enrollment software vendor ecosystem solidified: industry analysis showed 567% productivity gains and 99.3% accuracy with AI-powered transcript processing, indicating tool maturity. Market momentum accelerated: vendor comparative reviews identified 10% enrollment uplift from AI screening tools with documented 250+ annual man-days savings. Persistent implementation challenges offset efficiency wins: Workday Student rollouts continued facing UX complexity and feature gaps. High-stakes algorithmic screening remained constrained by sustained peer-reviewed fairness research and institutional caution; practitioner guidance (May-June 2025) emphasized explainable AI, fairness audits, and hybrid human-AI models as essential safeguards. Market divide solidified by year-end: low-risk chatbot and transcript automation continued scaling with proven ROI; high-stakes algorithmic screening remained cautious due to well-documented bias risks and institutional liability concerns.
  • 2026-Jan: January deployments at Caltech (VIVA AI interview tool) and Virginia Tech (AI essay evaluation) demonstrated continued adoption of AI evaluation tooling alongside human review. International student survey (1,600+ respondents) found 17% use AI for university search with high satisfaction (96% met/exceeded traditional sources), signaling student-side adoption shift. Peer-reviewed research (AAAI 2026) documented technical limitations: ML admissions models show degraded performance when applicant pool composition shifts, confirming reliability constraints in algorithmic screening. Practitioner guidance emphasized ongoing challenges in AI-generated writing detection for academic integrity in graduate admissions.
  • 2026-Feb: Research on AI-generated essay writing in applications (81,663 applicants, 2020-2024) showed LLM adoption accelerating in 2024 with concerning equity patterns: lower SES students used LLMs at higher rates but faced stronger admission probability declines. Augsburg University demonstrated positive alternative: automatic admission program reduced application friction and improved student confidence without algorithmic screening. Private school survey indicated 87% portal adoption and 77.8% online scheduling, confirming partial automation of routine intake tasks while preserving human touchpoints. Admissions system modernization continues with reported 25% cycle acceleration and 35% manual task reduction via CRM platforms. Security and reliability concerns persisted: guidance documents cited institutional errors (mistaken acceptances) and data breaches (Columbia). Market momentum remained stable: low-stakes automation demonstrating continued ROI and scaling; high-stakes algorithmic screening constrained by both fairness research and emerging AI-generated writing authenticity challenges.
  • 2026-Mar: Market surveys confirm AI workflow automation and CRM integration as baseline expectations across leading enrollment platforms (Slate 9.1/10, Element451 8.8/10 in 2026 vendor rankings). Institutional adoption sentiment has shifted measurably: Ellucian's survey of 779 professionals across 300+ institutions shows 66% institutional AI adoption (up from 49%), with 51% citing Marketing/Admissions/Enrollment as top benefit area; Comm100 report documents 44% AI deployment in admissions with 6-10 staff hours saved weekly and 1-6 month payback. Response-speed automation confirmed as a material conversion driver with research-backed evidence linking faster lead routing to enrollment uplift. Low-stakes automation continues scaling with strong ROI evidence; high-stakes algorithmic screening remains constrained by fairness concerns and institutional caution.
  • 2026-Apr: SIS vendor ecosystem maturity confirmed at new scale: Workday Student now manages 5.8M student records across 200+ institutions globally; Ellucian Student launched as a unified AI-native platform integrating SIS, HCM, and Finance (April 2026 GA), with a record 26 SaaS go-lives in Q1 2026 (including Aurora, Mohamed bin Zayed AI University, and Norwich). Gartner Magic Quadrant 2026 recognized both Workday and Ellucian as leaders, with Workday's Pensacola State deployment achieving 43% enrollment growth. AI evaluation tools matured operationally: Virginia Tech's essay reader processes 250k applications in <1 hour, Caltech's VIVA bot tests intellectual ownership, California Community Colleges deploy fraud detection—all with human oversight preserved. Practitioner landscape research identified growing equity concerns: EAB survey of graduate enrollment managers shows 87% have tried AI but only 23% of institutions formally use it, signaling organizational adoption lag; Foundry10's first full-cycle study finds students from families earning $75k–$100k adopt AI in applications at 150% higher rate than lower-SES peers, raising fairness questions. Governance advancement: ACM FAccT 2025 introduces hyperFA*IR algorithm addressing finite-pool selection bias in university admissions systems. EAB enrollment leaders survey shows 30% prioritize "Actually-Existing Admissions AI," seeking hype-busting examples and real ROI validation. Market remains bifurcated: low-stakes automation (chatbots, lead routing, transcript processing) proven and scaling; high-stakes screening constrained by fairness research, AI-generated essay authenticity questions, and regulatory emergence (EU AI Act August 2026 classification as high-risk).
  • 2026-May: The University System of Georgia (25 institutions) selected Workday ERP for a July 2028 go-live, the largest multi-institutional commitment yet, confirming continued consolidation around a small number of enterprise SIS platforms. EAB analysis of 7M+ student journeys across 50+ institutions reveals that 53% of prospective students narrow their college list before any institutional CRM contact, challenging the assumption that enrollment automation operates at the critical decision point.

TOOLS