Employee generated learning means the people who hold the knowledge build the training. The instructional design (ID) team sets standards, coaches authors and reviews the work. You will also see it written as employee-generated learning, or EGL. It works well for fast-changing, low-risk know-how and poorly for high-stakes or behavior-heavy topics.
Course design is already spread across many job titles. In ATD research released in July 2024, only 25 percent of respondents said instructional design was their primary responsibility. Another 27 percent said it was a small part of their work. Whether subject-matter expert (SME) courses match ID-built ones is less clear. Most of the evidence we found comes from vendors.
Most pages that rank for this term are written by authoring-tool vendors. We are one too: SimpliAuthor is made by SimpliTrain, the same company that publishes this site, so weigh our view accordingly. We have not tested any tool or run an SME program. The facts below come from published sources as of November 2026, and the decision table and checklist are our own judgment.
Key takeaways
- Employee generated learning moves course creation to SMEs. ID staff become coaches, template owners and reviewers.
- The case for it rests mostly on vendor-reported results. We found no independent study that compares SME-built and ID-built courses.
- Experts tend to overestimate what beginners know, so every SME-built course needs a test with a real novice.
- Match review depth to risk: light for internal how-tos, full gates for regulated or safety content.
- Ten minimum guardrails, listed below, cover ownership, objectives, sources, accessibility, AI use and expiry.
What is employee generated learning?
Employee generated learning is a model in which responsibility for creating and maintaining training moves from the L&D department to employees and subject-matter experts. Instructional designers stop being the only builders. They coach authors, provide templates and review work. Easygenerator’s guide gives this definition.
The 2024 book Employee-generated Learning, from Kogan Page, is by Easygenerator co-founder Kasper Spiro and research director Videhi Bhamidi. So the term’s main promoter also sells an authoring tool. That is a reason to check claims, not to dismiss them.
The same guide admits limits. It says the model suits large-audience, high-stakes content such as compliance and safety poorly. It also fits the higher Bloom’s levels badly. It puts that kind of content at under 5 percent of training needs but gives no source for the share.
Why is employee generated learning growing now?
Three pressures push it. Central teams are stretched, expert knowledge changes faster than a build queue, and AI tools can draft a first version from documents. None proves SME-built courses work well, but together they explain why more teams are trying it.
First, capacity. ATD’s 2025 State of the Industry report used 2024 data from 539 organizations. It puts the average at 212 employees per talent development staff member, down from 352 the year before.
Second, AI. Donald H Taylor’s September 2025 survey of over 600 L&D professionals found that producing content is still the biggest single use of AI in L&D. The report page says the output is often good enough, though not as good as the best instructional designers’ work. For the wider set of shifts, see our eLearning trends for 2027.
Claims about speed, savings and approval rates come almost entirely from vendors and their customer stories, and several have no named source. Treat them as hypotheses for a pilot, not benchmarks.
What goes wrong when SMEs build courses alone?
The failures are predictable. Experts overestimate what beginners know, pack in too much content, skip objectives, run out of time and leave courses unmaintained. These are process and design problems, not character flaws, so guardrails can address most of them.
The first is the curse of knowledge. In a 2022 study in Memory & Cognition, Tullis and Feder found that knowing the answers led people to predict that more novices would know them. The authors say this may help explain why teachers overpredict student knowledge. Their participants were students learning trivia, not workplace experts, so read it as a hint, not proof.
Practitioners report the same pattern in course work. SMEs hand over huge raw documents and expect all of it taught. They reuse a meeting slide deck as the course. They treat learning objectives as optional. Time is the other limit. A 2020 Training Industry piece by OpenSesame’s CEO says building training pulls SMEs away from their jobs and that the content is hard to keep current.
Pro tip: Run the Blind Spot Test. Before the first review, ask the SME to name the three knowledge-check questions a new hire is most likely to miss. Then watch one real novice take the check. The gap between the guess and the result shows how much context the draft still assumes.
When should an ID team, an SME or both build the course?
Let SMEs build when the content is practical, changes often and carries low risk. Keep the ID team in charge when errors are costly, the audience is large or the skill is behavioral. Use a hybrid when an expert holds the knowledge but cannot easily teach it.
The table below is our judgment, drawn from the sources in the last column. It is a default, not a rule. Your regulator, client contract or risk team can override it.
| Situation | Default builder | Review level | Basis |
|---|---|---|---|
| Software steps, product updates or process changes for an internal audience | SME builds | Light: one peer SME check and an ID spot-check of the objective | Judgment. A practitioner article says technical skills suit employee-made content. |
| Regulated, safety, legal or financial-disclosure content | ID team builds, SME is the source | Full gates, a second expert and a named sign-off | Easygenerator’s guide keeps high-stakes compliance with L&D. Dillon (2026) wants tighter control of medication and financial topics. |
| Large audience with a long shelf life, such as company-wide onboarding | ID team or hybrid | Full gates | Judgment. One flaw repeats for every learner. |
| Behavioral and soft skills, such as coaching or hard conversations | Hybrid, with ID leading the design | ID reviews the practice and feedback design | One practitioner says skilled performers often cannot explain what they do. The guide says higher Bloom’s levels fit poorly. |
| An expert is about to leave the company | SME records, ID shapes the course | Medium: a reviewer from the same field | Judgment. The guide names knowledge loss as a risk. |
| The SME has no protected time | ID team builds, SME reviews at gates | Set by risk tier | The 2020 OpenSesame piece on SME time. |
For regulated content, see how one gated process works in our aviation course development process. Each stage has one owner, and nobody signs off a stage they wrote.
What does the ID team do in employee generated learning?
The ID team shifts from builder to coach, template owner and reviewer. It sets standards, trains authors, reviews at agreed gates and keeps the library healthy. It still builds high-risk and large-audience courses itself. Easygenerator, 360Learning and Articulate all describe this same shift.
Review gates keep this from becoming a bottleneck. Our course production workflow describes six gates where SMEs review only for accuracy, completeness and what is essential, within a fixed number of rounds. The ID team should also help SMEs write objectives and cut content before anything is built. Our guide to instructional design basics for SMEs covers that starting point.
What is the minimum guardrail checklist for SME-built courses?
A workable minimum has ten guardrails. They cover ownership, risk tier, objective, template, source checks, novice testing, accessibility, AI and data rules, expiry and protected time. Skip one and a known failure returns. The table names the failure each one prevents.
Use the table as a gate list. The risk tier decides how many items need a second checker.
| Guardrail | Minimum standard | Problem it prevents |
|---|---|---|
| 1. Owner and successor | Each course names an author and a team that takes over if the author leaves. | Orphaned, outdated courses |
| 2. Risk tier | At intake, mark the course low, medium or high. The tier sets the review level. | The same review for every topic |
| 3. One-sentence objective | State what learners will do. Someone other than the SME checks it. | Information dumps |
| 4. Template | A standard opening with audience, time and objective, plus set screen types. | Uneven structure and tone |
| 5. Source check | Every fact traces to a named, dated document. A reviewer samples the facts. | Wrong or outdated facts |
| 6. Novice test | One person from the audience completes the draft and reports where they got stuck. | The expert blind spot |
| 7. Accessibility check | Captions, alt text, contrast and keyboard use are checked against WCAG 2.2. | Courses some learners cannot use |
| 8. AI and data rules | Name the approved AI tools, say which files may be uploaded and require an expert to read every AI draft. | Invented facts and leaked files |
| 9. Status label and expiry | Label each course as peer reviewed or L&D approved. Set a review date and archive stale courses. | Unclear trust and stale content |
| 10. Protected time | A manager agrees the hours for building and updating. | Abandoned courses |
Pro tip: Review the Objective Before the Draft. The cheapest review point is guardrail 3, before a single screen exists. Fixing a sentence costs a minute. Fixing a built course costs hours and a busy expert’s goodwill. Ask the reviewer to approve the objective in writing first.
Does AI make SME-built courses safer or riskier?
AI makes SME-built courses faster to start and no safer to publish. A draft from source documents lowers the barrier, but it can carry errors and expose confidential files. The same guardrails apply, and an expert should read every AI draft.
JD Dillon wrote in Training Industry Magazine’s Winter 2026 issue that approval-based governance cannot keep pace with AI-generated content. He suggests sensitivity levels for critical subjects and guardrails inside AI tools. He cites no data, so it is a practitioner’s view, but it fits the risk tiers above. Our post on drafting an AI course outline from trainer notes shows a review step in which the SME owns the final content.
What should you do next?
Run a small pilot before you write policy. Pick two or three SMEs and one low-risk topic, such as a product update for an internal team. Give them the template and the ten guardrails. Run one full cycle, then record the build time, what the reviewer fixed and what a novice got wrong.
Then decide where the line sits. Say which topics stay with the ID team, which go to SMEs with light review, and which need a hybrid. Write it as a one-page tier list your SMEs can read in five minutes.
If you need an authoring tool for the pilot, SimpliAuthor is one option, with the same-company disclosure from the top of this post. Based on the vendor’s published documentation as of November 2026, its homepage says you can build from documents, videos, web pages or SCORM packages. It says AI drafts objectives, an outline and assessments from your sources, and you edit on one canvas. The pages do not say it is designed for non-designers, and they do not state what the free option includes. You can start free on SimpliAuthor and test it with your own SMEs.
Frequently asked questions
What are the benefits of employee generated learning?
Vendors and practitioners claim faster builds, fresher content, less pressure on L&D and more recognition for experts. Most figures behind those claims are vendor-reported, and we found no independent study comparing SME-built and ID-built courses. Treat them as hypotheses. Run a pilot and measure build time, reviewer fixes and learner results.
What are the disadvantages of employee generated learning?
The sources name four. Experts overestimate what novices know, SMEs have little spare time, quality varies between authors, and courses go stale without an owner. Easygenerator’s own guide also says the model fits high-stakes compliance content poorly. Templates, risk tiers and review gates reduce these problems but do not remove them.
Do SMEs need instructional design training to build courses?
Sources disagree. 360Learning says experts need less friction, not more training, while another practitioner source recommends a few hours of ID consultation. In our judgment, a short orientation on objectives and cutting content, plus a template and a named reviewer, covers most of the gap for low-risk courses.
Does employee generated learning replace instructional designers?
No. The sources describe a role change, not removal. Designers coach authors, own templates, review work and still build high-risk or large-audience courses. ATD’s 2024 research found only 25 percent of respondents had instructional design as their main job, so a lot of design work is already part-time and benefits from that support.
How do you measure whether SME-built courses work?
Count outcomes, not output. Track whether courses are used and reused, how learners rate their relevance, and which quiz questions most learners miss, since a high miss rate can point to unclear content. Where you can, check whether job results change. Counting courses created says little about learning.



