Learning in the Flow of Work: From Events to Infrastructure
68% of employees prefer to learn at work, in the flow of their tasks. 70% of learning happens on the job rather than through formal training. 58% want to learn at their own pace. Yet most enterprise L&D budget and design effort goes into formal event-based programmes that extract employees from their workflow for blocks…
1. The 70-20-10 Model Revisited: Why the Numbers Still Matter
The 70-20-10 framework that 70% of development happens through on-the-job experience, 20% through social learning and feedback, and only 10% through formal training, was never intended as a prescription or a precise measurement. It is an observation about where development naturally occurs. What is useful about it in 2026 is the implications it creates: if 70% of development happens through experience, then the most significant untapped L&D opportunity is designing experiences better, not building more programmes.
70%
of employee skills are learned on the job rather than through formal training sessions — the foundational observation that makes workflow-embedded learning design the highest-leverage L&D investment available (eLearning Industry Employee Training Statistics 2025)
68%
of employees prefer to learn at work and in the flow of their tasks — not in separate training sessions, making flow-of-work design alignment with learner preference as well as learning science
58%
of employees want to learn at their own pace — the flexibility preference that flow-of-work microlearning and performance support serve better than scheduled classroom or even scheduled eLearning
Only 10%
improvement in retention when employees learn at their own pace in their workflow context — the range across multiple studies connecting learner-controlled access to measurable retention improvement
Key Distinction
The gap between where development naturally occurs (70% on the job) and where most L&D investment goes (primarily formal programmes) is not an argument against formal programmes. It is an argument for deliberately designing the on-the-job 70% rather than leaving it to chance. The difference between organisations where on-the-job development is structured, supported, and reinforced, and those where it is accidental, uneven, and unreinforced is the gap between consistent capability development and the lottery of who happens to have a good manager and a challenging assignment.
2. Flow-of-Work Learning Formats — What Works for What
| Content Type | Flow-of-Work Format | Moment of Need |
|---|---|---|
| Embedded performance support (job aids, in-system guides) | Infrequent or complex tasks where the employee needs help at the moment of execution | One objective per module. Requires retrieval practice, not re-exposure. Triggered by task type, time interval, or performance indicator. |
| Workflow-triggered microlearning | Spaced reinforcement of training recently completed; knowledge currency for evolving domains | One objective per module. Requires retrieval practice, not re-exposure. Triggered by task type, time interval, or performance indicator. |
| Manager coaching in regular one-to-ones | Capability development on the specific challenges the employee is currently navigating | Manager equipped with coaching skills and a development conversation framework. Specific objective agreed with the employee. |
| Peer knowledge sharing (communities of practice, knowledge bases) | Domain currency, problem-solving support, collective intelligence on evolving challenges | Defined community purpose and contribution expectations. Curated knowledge base rather than a passive information repository. |
| Stretch assignments with structured reflection | Developing capability through progressive challenge in the real work context | Assignment matched to the specific capability gap. Reflection structure and feedback mechanism designed alongside the assignment. |
3. Designing for the Flow — Practical Architecture
“The organisation that asks ‘where in our employees’ workflow does a capability gap arise, and what could we put there to support performance and development at that moment?’ is designing for the flow of work. The one that asks ‘what course should we build for this topic?’ is designing for the training catalogue. The first produces development where 70% of learning happens. The second produces development for the 10% that formal training accounts for.“
- Map the workflow before designing the learning. For each role, identify the points in the workflow where capability gaps arise most frequently, where errors or quality shortfalls most commonly occur, and where performance support or reinforcement would most reduce friction. These workflow hot spots are the design targets for flow-of-work learning. They are more valuable than any subject matter expert briefing about what people should know — because they identify where the gap is actually costing performance, not where a trainer believes knowledge is insufficient.
- Design the formal programme and the flow-of-work reinforcement simultaneously. The most effective learning architecture designs the formal event and its workflow-embedded reinforcement as a single system — not as a programme followed eventually by a reinforcement afterthought. The formal event develops foundational capability. The workflow-embedded microlearning at Days 7, 30, and 90 reinforces the specific decisions from the formal event in the work context where they will be applied. Designed together, they produce compound retention. Designed separately, they produce the forgetting curve punctuated by programme delivery.
- Develop managers as the flow-of-work development infrastructure. The manager who holds a weekly one-to-one with each direct report and uses it to provide specific feedback on capability development, ask coaching questions that build reflection on recent challenges, and make stretch assignment decisions that target specific development goals is providing the most powerful flow-of-work development available. Manager capability for development conversations is the highest-leverage infrastructure investment available to any L&D function building flow-of-work learning.
4. Why Flow-of-Work Learning Complements Rather Than Replaces Formal Programmes
- Formal programmes develop foundational capability that flow-of-work learning cannot. Extended engagement with progressive complexity, peer challenge, expert facilitation, and the psychological safety of a learning environment separate from operational performance, these are what formal programmes provide. Flow-of-work learning assumes some foundational capability to build on. A 5-minute workflow-embedded module on objection handling assumes the learner already understands what an objection is and has some framework for addressing it. The formal programme builds the framework. The flow-of-work reinforcement maintains and extends its application.
- The most effective learning architectures use both, sequenced deliberately. Formal event for foundational capability development this is what the event does well. Workflow-embedded reinforcement for application maintenance, this is what the flow-of-work design does well. Manager coaching for personalised development of the specific edge each individual is developing. This is what the 20% in 70-20-10 does well. The architecture that combines all three, with each element designed to serve its specific purpose, produces the compound capability development that neither element produces alone.
In Summary
The 70% of learning that happens on the job is not happening by accident in most organisations; it is happening because of the quality of management, the challenge of assignments, and the availability of support at the moment of need. Some employees work for managers who develop them through every interaction. Others work for managers who are too busy or too underdeveloped to provide it. This variance is the single largest source of unequal development opportunity in enterprise organisations — and it is addressable through deliberate flow-of-work learning design.
The shift from learning as event to learning as infrastructure does not require a new platform or a larger budget. It requires workflow analysis that identifies where learning is needed, manager development that equips the most powerful development lever available, and a design discipline that builds formal events and their flow-of-work reinforcement as a single system rather than two separate initiatives. These investments produce the capability development consistency that event-based training alone never achieves — because they develop where 70% of learning already happens, rather than designing better events for the 10% that formal training represents.
Qquench · 25+ Years · Workflow-Embedded Learning Design · Performance Support Architecture · Manager Development for Flow-of-Work · Formal and Informal Learning Integration · Fortune 100 · Global
Qquench designs learning architectures that combine formal capability development with workflow-embedded reinforcement, producing the compound capability development that neither formal programmes nor informal learning produces alone.
We design the workflow analysis, performance support assets, microlearning reinforcement series, and manager development frameworks that make learning infrastructure rather than a schedule of events.
Frequently Asked Questions
Q1
What does learning in the flow of work mean in practice?
Making development available at the moment of need within the work context — embedded job aids within systems, microlearning triggered at relevant task points, performance support tools in software workflows, and manager conversations that develop capability during regular operational interactions. The common thread is proximity to the work.
Q2
What is the difference between performance support and flow-of-work learning?
Performance support helps complete a current task without necessarily developing capability checklists, job aids, or decision trees. Flow-of-work learning is broader: it includes performance support plus microlearning, coaching conversations, and knowledge exchange that develop capability in the work context. Performance support reduces effort for known processes; flow-of-work learning develops capability for novel ones.
Q3
How should L&D measure flow-of-work learning impact?
Against the operational outcomes the embedded learning is designed to influence: error rate reduction, time-to-completion, quality scores for supported tasks. For structured microlearning series — skill movement at 30 and 90 days. Compare populations with and without embedded support on the specific operational metric it was designed to improve.
Q4
Is flow-of-work learning a replacement for formal programmes?
No, it is a compliment. Formal programmes develop foundational capability requiring extended engagement, peer learning, and progressive complexity. Flow-of-work learning reinforces and applies that capability, provides just-in-time support, and develops knowledge currency for evolving domains. The most effective architectures combine both, designed as a single system.
QS
Qquench Specialists
Learning Architecture and Workflow-Embedded Design Practice · Qquench
25+ years designing learning architectures that develop capability where 70% of learning happens in the workflow, supported by managers, reinforced at intervals. We write from practice, not position papers.