Insights
Learners need to know where they are, how they did and what is next. Administrators need courses, enrolment and reports that mean what they say. Learners need an eLearning platform to answer three questions at every point: what am I working towards, how am I doing, and what should I do next.
Learners need an eLearning platform to answer three questions at every point: what am I working towards, how am I doing, and what should I do next. That has to hold when they come back after weeks away, on a phone, or with assistive technology. Administrators and instructors need to build and change courses without fighting the tool, enrol and group people in bulk, set exceptions for individuals, and report on records that mean what they say.
The two lists are joined. Much of what a learner sees on a learning platform is something an instructor or administrator configured, so the learner experience cannot be fixed on its own.
Two sides, one platform
Each learner need has a counterpart on the administrator's side. If the counterpart is missing, the learner-facing screen cannot deliver the need, however well it is designed.
| What the learner needs | What has to exist on the administrator and instructor side |
|---|---|
| To see where they are and what comes next | A course structure the instructor can build, reorder and label without workarounds |
| Feedback that says how they did and what to try next | Criteria and feedback the instructor can attach to an attempt quickly enough to be useful |
| A way back in after a gap | Deadlines and resume points that can be adjusted, and a report that tells a stalled learner from a struggling one |
| Access on any device and with any disability | An authoring tool that prompts for accessible content, and per-learner settings such as extra time on a test |
| Motivation that does not manipulate | Progress rules (what counts as done) that someone decided on purpose |
| A record they can trust | Roles, data settings and reports that separate access, activity, completion and evidence of learning |
The rest of this page takes each side in turn, then shows where they meet and how to test a platform against both.
What learners need
To know where they are and what comes next
John Hattie and Helen Timperley's 2007 review of feedback research, The Power of Feedback, frames effective feedback as answers to three questions that a student or a teacher may ask: where am I going, how am I going, and where to next. In an interface, those questions become concrete. There is a visible objective for the course or module, a progress view that shows what is done and what remains, and one obvious next action on the screen the learner lands on.
Platform makers have reached the same priority. Moodle's own prototype notes for its navigation redesign state that the main aim was to reduce the cognitive load of finding one's way around a Moodle site, and describe a course index that also gives teachers an easy view of the course structure. The redesign was aimed at both sides at once.
Feedback on the work, not on the person
More feedback is not automatically better. Avraham Kluger and Angelo DeNisi's 1996 meta-analysis of feedback interventions covered 607 effect sizes. The average effect on performance was positive, yet more than a third of the interventions made performance worse, and their preliminary analysis suggested that feedback worked less well the more it drew attention away from the task and towards the self.
For a platform, our reading is this: automated feedback should point at the attempt and the next step rather than the person. "Incorrect" answers how am I going and leaves where to next empty. "Well done!" answers neither. A useful message names what went wrong in this attempt and what to try, or routes the learner to the material that addresses it.
A way back in after a gap
Stopping part-way is common in open online courses. In Justin Reich and José Ruipérez-Valiente's analysis of six years of Harvard and MIT courses on edX, The MOOC Pivot, low completion rates did not improve over six years, and the vast majority of learners never came back for another course after their first year. That is open online courses, not every platform, but it is reason enough to design the return rather than assume continuous attendance.
Learning research suggests what the return should do. In their 2013 review of ten study techniques, Dunlosky and colleagues rated practice testing and spacing study over time as the two techniques of high utility, and rereading as low. Our reading: a returning learner is a spaced learner by accident. A short check on what they last covered uses the gap; sending them to reread the last module wastes it.
The return also has to avoid punishing the gap. A post on Duolingo's blog says that breaking a streak can feel quite demotivating and that fear of losing one could stop learners from starting one, which is why it lets learners freeze a streak. That is the company's own account rather than independent research, but the design point carries: on return, show exactly where the learner left off, what has changed or fallen due since, and a realistic next step, not a wall of overdue warnings.
Motivation without manipulation
Points, badges and leaderboards have measurable effects on learning, but small ones. Michael Sailer and Lisa Homner's meta-analysis of gamification in learning found small positive effects on cognitive, motivational and behavioural outcomes. The cognitive effect held in the more rigorous studies; the motivational and behavioural effects were less stable.
Our reading: game elements are a modest addition to clear structure and useful feedback, not a substitute for them. A badge for finishing a module only means something if finishing the module means something, which is a decision about progress rules (see [where the two sides meet](#where-the-two-sides-meet)).
Access on any device and with any need
Learning platforms collect interactions that accessibility standards single out. Several criteria added in WCAG 2.2 land directly on them:
- Dragging movements: drag-and-drop quiz questions and course reordering need a single-pointer alternative.
- Target size: answer options, progress dots and media controls need a minimum target size, or enough space around them.
- Accessible authentication: signing in should not depend on remembering or transcribing a code without help such as paste or a password manager.
- Consistent help: where help is offered on several pages, it sits in the same relative position on each.
- Redundant entry: learners are not asked to re-enter information they have already given in the same process.
CAST's Universal Design for Learning Guidelines (version 3.0) add the learning side of the same idea: design multiple means of engagement, of representation, and of action and expression, so that the format of a course is not itself the barrier.
On a phone, the most useful thing to show is the next activity, or the place the learner stopped. A course page that shrinks a desktop outline onto a small screen makes that harder to find.
If the platform serves public institutions, accessibility is also a buying condition. The US Department of Justice's Title II web rule covers public schools, community colleges and public universities, requires WCAG 2.1 Level AA, includes content that vendors provide on their behalf, and after a 2026 extension applies from 26 April 2027 for public entities serving a population of 50,000 or more, and 26 April 2028 for smaller ones and special districts. In the UK, the public sector accessibility requirements ask for WCAG 2.2 AA and a published accessibility statement, and cover intranets and extranets; schools and nurseries are partially exempt but must still publish a statement. Neither sentence is legal advice; whether a given platform is in scope depends on who provides it and how.
What administrators and instructors need
To build and change courses without fighting the tool
The instructor's course is the learner's interface. In a study of how university students experience Moodle, Maslov, Nikou and Hansen found that, in many respects, the experience depended on how teachers had structured their courses. It was a small study (interviews and a survey with 20 students), but the mechanism is plain: the same platform produces different learner experiences in different hands.
So course building needs the same design attention as the learner view. In practice that means:
- reordering sections and activities directly, with a keyboard alternative to dragging;
- duplicating and reusing a course for a new cohort without rebuilding it;
- previewing exactly what a learner will see, in the learner's role;
- changing a live course without breaking the records of people part-way through it.
Accessibility starts here as well. The W3C's Authoring Tool Accessibility Guidelines (ATAG 2.0) name learning management systems among the tools they apply to and include a part on how the tool should help authors produce accessible content. A course builder that asks for alternative text when an image is added, and for captions when a video is uploaded, prevents defects that no learner-side redesign can repair later.
To enrol, group and make exceptions at scale
Administrators work in batches: a new intake, a department, a client's staff. Enrolment needs to come from a file or from the source system rather than one person at a time, and groups and roles need to follow. Where courses use external tools, the LTI standard's Names and Role Provisioning Services let a tool receive the course roster and roles from the platform instead of asking an administrator to maintain a second list.
Exceptions are a separate job: one learner's extended deadline, another's extra time on a test because of a disability. Moodle, for example, lets dates, timing and the number of attempts be changed for individual users or groups. The design point is visibility: an exception should be easy to set, easy to see on the learner's record, and easy to remove, rather than buried in a settings page that only one administrator knows about.
Reports that say what they mean
A learning record can mean at least five different things: the learner opened the material (access), did something with it (activity), met a completion rule (completion), scored on an assessment (performance), or showed the understanding the course was for (evaluated learning). Reports and dashboards that blur these let a click pass for evidence.
Research on dashboards gives reason for caution. A systematic review of empirical studies of learning analytics dashboards by Matcha, Ahmad Uzir, Gašević and Pardo found little grounding in learning theory, no basis yet for recommending dashboards as support for learners' own monitoring of their learning, and weak evaluations.
What an instructor needs first is narrower than a dashboard: who needs help this week, and why. A useful report separates "not started", "started and stalled" and "attempted and struggling", because each calls for a different response. Where learners are students, Jisc's code of practice for learning analytics adds that analytics should benefit students, be transparent about what is collected and why, and be visible to the students themselves.
Records that stand up to compliance
In workplace learning, the report may have to satisfy someone outside the team: who completed which version of which course, when, with what result, and when it expires. That requires version history on courses, expiry and recertification rules, and an audit trail of manual changes.
The record format limits what any report can say. SCORM tracks a fixed set of data elements that the course reports while it runs in the browser. The xAPI specification, now published as an IEEE standard (9274.1.1-2023), defines a JSON format for statements about learning experiences, sent to a learning record store, and the cmi5 specification defines how an LMS launches and tracks courses using xAPI. Choosing among them is a product decision about what you will need to prove later, not only a technical one.
Platforms used by children carry a further obligation. The UK Information Commissioner's Office states that its Children's code applies to edtech likely to be accessed by children on a direct-to-consumer basis, and to edtech provided through schools where the provider influences the nature and purpose of the processing, for example by using children's data for its own product development. Privacy defaults and data settings belong in the administrator experience from the start.
Where the two sides meet
The table at the top of this page is the practical argument. Three places show it most clearly.
Completion rules. If viewing a page counts as completing it, the learner's progress bar reaches 100 per cent and the instructor's report says nothing about understanding. The rule is set on the administrator side; its effects appear on both.
Course structure. A learner's "what next" is the instructor's outline. If the authoring tool makes restructuring painful, outlines stay as they were first built, and the learner view inherits them.
Exceptions and deadlines. A learner returning after a gap sees whatever the platform does with missed dates. If adjusting them is hard for staff, the learner meets a wall of overdue items.
Our judgement from this: a redesign that reskins the learner front end while leaving course building, rules and reporting untouched risks moving the problem rather than solving it. Scope the work across both sides, even if the visible change is mostly on one.
How to test a platform against both sides
Task-based testing with real users of each role checks the points above directly. These tasks are a starting set; adjust the numbers and content to your platform.
| Who | Task | A failure looks like |
|---|---|---|
| New learner, on a phone | Find the first activity and start it | Scrolling through a full course outline to find a starting point |
| Learner returning after three weeks | Resume, and say what is now due | Landing on the dashboard with no resume point, or a list of overdue items with no next step |
| Learner using only a keyboard or a screen reader | Complete a quiz that includes a drag-and-drop question | No non-drag alternative; focus lost inside the question |
| Learner after a failed attempt | Explain what to do differently next time | Feedback that only says the answer was wrong |
| Instructor | Move a module, duplicate the course for a new cohort, preview it as a learner | Rebuilding by hand; no learner preview |
| Instructor | Give one learner extra time on one test | Changing the test for everyone, or asking an administrator |
| Administrator | Enrol a new intake into three groups from a file | One-by-one enrolment, or a file format nobody can produce |
| Administrator | List who has not completed the current version of a mandatory course, and whose certificate expires next month | Exporting to a spreadsheet and reconciling by hand |
| Instructor | Name who needs help this week, and why | A completion percentage with no way to tell stalled from struggling |
Usability testing of this kind answers different questions from an expert review; UX audit, heuristic evaluation and usability testing compared explains which method fits which question.
The Wandering Raven: one Moodle platform, both sides
The Wandering Raven is an EdTech startup offering a creative-writing platform for students aged 8 to 18, built on Moodle. Tcules interviewed the client's educators and a sample of the target audience, and ran a heuristic evaluation of the existing site. The audit found mobile inaccessibility, unstructured content, a disorganised interface and little sense of community or play.
The learner-side work covered information architecture, characters drawn for the platform, writing prompts, interactive exercises, peer feedback, discussion forums, progress tracking and badges, with hybrid wireframing in which the client was closely involved. The administrator side was part of the same job: Tcules installed selected Moodle plugins, configured the platform and built a custom theme, and took on installing, customising, integrating and supporting Moodle so the client could concentrate on creating content.
What the published case does not show is just as important. It documents design and implementation responsibility. It does not publish measures of learning outcomes, engagement or accessibility conformance, and nothing on this page should be read as claiming them.
Where Tcules fits
Tcules is a digital product design studio creating ambitious new software products and making established ones better. Product Design is the foundation of the practice, extended by Design Engineering and Software Development, with a particular focus on SaaS. For learning products, that means working on learner and educator experiences together, including AI-assisted support and implementation in Moodle or custom software. How we approach that work is set out on our EdTech and learning expertise page.