An SMB HR lead asks a vendor: "How many minutes should our module be?" The vendor says five. HR pays, employees sit bored for three minutes after they already got the point, and HR blames the author. The right question is not how many minutes. The question is what the employee must be able to do at the end, and the minimum content that gets them there.
This guide splits the useful range into three brackets — 2 to 3 minutes, 4 to 6 minutes, and 7 to 10 minutes — and shows which material belongs where.
Why a minute target is the wrong starting point
Microlearning as an approach rests on two foundations. The first is Sweller's cognitive load theory from 1988: working memory is limited, so removing unnecessary material frees capacity for the key concept. The second is Ebbinghaus's forgetting curve from 1885: a single event decays, but repetition at intervals slows the decay.
Neither theory defines the "right" length in minutes. Both point in another direction:
- One idea per module. Two concepts belong in two modules, even if the total is under five minutes.
- Design for repetition. One 15-minute module ranks lower than three 5-minute modules spread over three weeks.
Minutes are a consequence, not a target. Define what the learner must be able to do first, and the length shows up on its own.
What the data actually says
The most-cited empirical source is the 2014 edX video analysis (Guo, Kim, Rubin). It found that total watch time peaked for videos of 6 to 9 minutes; longer videos were often abandoned before the end. The authors themselves noted that watch time cannot confirm whether a viewer was actively following the content.
That is not proof that 7 minutes is the optimum. It only shows that attention falls off for most users past 10 minutes, and that under 2 minutes there is no room for a scenario or an exercise. Teams under 100 people see a similar pattern in practice: completion rates often sit under 60 % for modules over 15 minutes and over 80 % for modules under 6 minutes, when the topic itself is not off-putting.
Two caveats:
- Studies on students and MOOC viewers transfer only partly to employees in an SMB context.
- A high completion rate does not mean the employee can apply the knowledge on the job.
Three brackets and what each one does
2 to 3 minutes — reminder, policy, short nudge
Two to three minutes is enough to refresh a concept the learner already met once. Examples: the annual phishing reminder, a password-policy update, a short note about a new exemption in a safety law. The module has one screen of text, one image or diagram, and optionally one trivial quiz question.
Do not try to teach a complex procedure in three minutes. The employee remembers only the first sentence.
4 to 6 minutes — one new concept with one example
Four to six minutes is the top of the useful range for real microlearning content. Structure:
- Minute 1. A problem from practice (one concrete scenario, not a definition).
- Minutes 2 to 4. The solution and why it works. One principle, not three.
- Minute 5. A quick exercise or quiz with feedback.
- Minute 6. What to do if an edge case appears.
Anything beyond that is a long module. If you cannot cut below 6 minutes, you are probably covering two concepts.
7 to 10 minutes — step-by-step procedure
The "long" microlearning module (the term is an oxymoron, but the industry uses it) has a place for procedures where steps cannot be split. Examples: a sign-off procedure under an ISO process, an intro to a GDPR DSAR response, the basics of a helpdesk escalation.
Attention drops past 10 minutes. If a procedure needs more than 10 minutes, split it by logic, not by time. Three 7-minute modules beat one 20-minute module.
Example 1: GDPR refresh for a 50-person agency
A digital agency, 50 employees. The previous year's GDPR training was a 45-minute e-learning. Completion rate 54 %, average post-test score 32/50. HR switches to a microlearning structure.
New design:
- Module 1 (3 min): Reminder of what counts as personal data — with three examples from the team's own work (CRM, email database, campaign recordings).
- Module 2 (5 min): A DSAR example — a concrete request email, what to check in 24 hours, and whom to escalate to.
- Module 3 (5 min): Third-party vendor checks — the contract clauses a PM must see before signing.
- Module 4 (3 min): Breach notification — one scenario, one decision, one correct action toward the supervisory authority.
Sixteen minutes of material in total, spread over four weeks. Completion per module: 87, 83, 81, and 79 %. Average post-test score 42/50. HR spends 11 hours on preparation, five hours less than the prior year's equivalent long-form module.
The point: the 3-minute modules did not re-explain basics because the context was known. The old long module repeated the basics each time, which bored anyone who already had them.
Example 2: Forklift safety for a 120-person warehouse
A logistics centre, 120 forklift operators. The annual refresher was a 90-minute classroom session. Attendance 100 %, but an internal audit flagged 23 skipped safety steps in the three months after training.
The L&D lead splits the content into six modules:
- Module 1 (5 min): Pre-start forklift check — photo plus interactive checklist.
- Module 2 (8 min): Pallet transport on marked lanes, two scenarios of incorrect loading.
- Module 3 (4 min): Floor markings and pedestrian priority.
- Module 4 (6 min): Fault procedure — one real recording from last year's inspection, with what went wrong.
- Module 5 (3 min): End-of-shift and parking routine.
- Module 6 (5 min): A 10-question quiz, pass at 8/10.
Material runtime 31 minutes, employee time 35 minutes across three weeks instead of 90 minutes in one day. A three-month audit shows 6 skipped steps instead of 23.
Operators rated module 2 — the eight-minute one — as the hardest, which confirms that over 7 minutes the content asks more effort. The concept, however, was inseparable and splitting it would have been artificial.
For the difference between microlearning and classic LMS modules see microlearning vs LMS.
Common length mistakes
- A 7-minute module with two concepts. Split into two 4-minute ones.
- A 2-minute module for one complex procedure. Impossible scenario. Move to 5 to 7 minutes.
- A 15-minute module because the manager asked for it. The manager's time budget is not the learning outcome. Show the completion rate from three shorter modules and compare.
- One length for every topic. A password reminder does not need 5 minutes; a DSAR response does not fit in 3.
- Microlearning without repetition. A single 4-minute module is not microlearning. A series at two-week intervals is.
That last point is the most common: without the pulse, microlearning is only a shorter version of the old e-learning.
Checklist for picking a length
Before publishing, answer five questions:
- How many concepts does the module hold? One → up to 6 minutes. Two → split.
- Is the content a procedure with ordered steps? Yes → 7 to 10 minutes. No → shorter.
- Has the learner met the basics already? Yes → 2 to 3 minutes as a reminder. No → at least 4 minutes for the intro.
- Is it a series or a one-off? Series → each part shorter. One-off → max 10 minutes.
- Is there an exercise or quiz in the module? Yes → add 1 to 2 minutes. No → the module ends before the goal.
For setting the learning objective before the length, see learning objectives that work.
What to do this week
Take one existing module over 15 minutes and measure two things: its completion rate and the number of unique concepts in it. You will likely find three to five concepts crammed together and completion under 60 %. A split into three shorter modules at two-week intervals is the first iteration.
Microlearning in an LMS like Mentor is not only about shorter modules; it is about designing the series with a built-in rhythm and measuring completion per part. The length decision becomes a consequence of structure, not the starting point. Once you do that once, you never ask again how many minutes a module should be.