AI in hiring. Learning that moves business KPIs. Assessment integrity. Real practitioner playbooks for L&D and TA teams.
A research note on the nine shifts that have reshaped what learners expect, what credentials mean, and what it now takes to run a course business independently — and what they imply for creators evaluating where to host their next program.
In-depth playbooks and research projects, broken into readable parts.
Synchronous and asynchronous learning aren't competing approaches — they're different infrastructure patterns optimised for different outcomes. An honest framework for when each model wins, why most enterprise L&D programmes need both, and how to evaluate vendor capability across both delivery patterns without accepting feature-checklist answers.
Most LMS vendors now claim AI capability, but most claims describe commodity AI integrations that don't change learning outcomes meaningfully. An honest framework for separating the AI capabilities that genuinely matter (grounded RAG, KPI-driven pathways, voice-preserving multilingual translation) from the bolt-on features designed to satisfy procurement checkboxes — and how to actually evaluate vendor AI claims against your real content.
Most isolation requirements that procurement frames as needing single-tenant LMS are satisfied by well-architected multi-tenant platforms with strong logical isolation, dedicated resource pools, and customer-managed encryption. An honest framework for what isolation actually requires, when single-tenant is genuinely necessary, and what to ask vendors during procurement.
Native mobile apps for enterprise LMS genuinely help for frontline workforces, distributed operations, and learning that depends on offline access. For most other contexts — office-based knowledge workers, scheduled training, occasional learners — a strong mobile-responsive web experience delivers the same outcomes at lower cost. An honest framework for when each fits, what good mobile actually looks like, and how to evaluate vendor mobile capability hands-on.
Open-source LMS platforms like Moodle, Open edX, and Canvas have no licence fee — but the real TCO at mid-market scale often exceeds modern SaaS LXPs once infrastructure, engineering capacity, customisation, partner support, and AI capabilities are honestly counted. A practical framework for when open-source genuinely fits, and when the free LMS framing breaks down.
For most organisations, buying a learning platform is the right call — but the build conversation keeps surfacing, and most internal proposals underestimate the cost of building an in-house LXP by 5-10x. An honest analysis of when each path makes sense, what the real TCO of building looks like, and how to evaluate vendor platforms when you've decided to buy.
Outsourced assessment services vs in-house platforms is genuinely an operating-model decision, not a vendor selection. An honest framework for when each makes sense, when the hybrid path is the right answer, and how to think about assessment as ongoing operational infrastructure rather than as a project-delivered service.
Browser-only proctoring was sufficient until AI tools moved below the browser. ChatGPT, Claude, Copilot, and Gemini now integrate at the operating-system level, where browser proctoring has no visibility. An honest analysis of the architectural gap, what OS-level proctoring closes, and what to ask vendors when assessment integrity actually matters.
For most organisations running between a handful and a few hundred requisitions a year, a lite ATS - particularly bundled with assessment infrastructure - is genuinely the right answer. Enterprise ATS earns its cost only at specific scale and complexity thresholds. An honest framework for sizing the ATS to your hiring operation rather than aspiring above it.