My Power Apps LMS says 100% complete, but I can’t prove what this person completed
A non-developer Power Apps maker has built or inherited a learning tracker that displays a completion result, but cannot reconstruct the claim across stable learner identity, enrolment or attempt, exact course version, required activities, completion rule, saved result and original valid time. Course edits, re-enrolment, replay, recalculation or correction make that missing chain visible, usually at the worst moment.
Context
The blocker, in a nutshell
If this blocker is unfamiliar, start here.
A learning management system (LMS) can show progress, completion or a percentage. Those labels describe the system's current result; they do not, by themselves, show the records behind it. A defensible completion claim needs a traceable chain from a stable learner identity and enrolment or attempt to the exact course version, the requirements and rule that applied, the saved result and the relevant event time. Different products store that chain differently.
Key Terms
Industry jargon explained
Click any term to see its definition.
The Reality
A day in their life
Business professional at a 50–1,000-person organisation who has been asked to build an internal learning tracker or LMS in Power Apps, despite not being a developer.
08:30 — The lab is finished. Four screens, six tables, one completion row for my fictional learner. It works, and I am pleased with it.
10:15 — Then I add a third lesson to the course, because that is what will happen in real life. I go back and look at my finished learner. I genuinely do not know what I expect to see, which is the moment I realise I never decided.
12:40 — I run the other three tests. Half a course does not produce a completion row, which is right. Repeating the final lesson does not create a second one, which is also right. Two learners with the same name stay separate, because I joined on the ID rather than the name.
15:10 — The version test is the one that needs work. I write down what I want to happen, rather than what happened by accident, and I note the two lines I still cannot answer from an actual record.
16:30 — What I want next time is simple: build it so that adding a lesson next month does not quietly rewrite what somebody finished last month, and be able to point at the row that proves it.
The People
Who experiences this blocker
Business professional at a 50–1,000-person organisation who has been asked to build an internal learning tracker or LMS in Power Apps, despite not being a developer.
25–55 • Beginner to early-intermediate Power Apps maker. Comfortable with tables and reports, but new to historical completion modelling, course versioning and learning-record standards.
Skills
Maps business processes
Builds basic Power Apps screens and Dataverse tables
Uses reports, spreadsheets and test records
Can follow relationships when they are explained in plain English
Can inspect a saved row rather than trusting a screen
Frustrations
- The dashboard says complete but does not show the records behind the number
- Course names and learner names are reused, so reports look ambiguous
- A course edit or retake can make an old result hard to explain
Goals
- Know what a completion percentage really represents
- Design records that survive course changes
- Explain one completion result to a manager or app owner
Learning operations lead, LMS buyer or training reviewer
Stakeholder asking the maker or supplier to substantiate a completion result
Also affected by this blocker. Often shares the same frustrations or creates additional pressure.
Top Objections
- Doesn’t the certificate already prove completion?
- Why do we need a separate whole-course result if lesson progress already exists?
- Can’t we simply recalculate the percentage whenever we need it?
- Do we really need course versions for a small internal LMS?
- Does passing this review mean the system is compliant or production-ready?
How They Talk
Use These Words
LMScourselessonlearnerenrolmentprogresscompletereportDataversetest record
Avoid
completion provenanceevent sourcingtemporal data modelbitemporal auditnormative conformance
Learning Pathway
Trustworthy Learning Completion Records
From trusting a percentage or certificate to knowing which records substantiate one completion claim, what happens to it when the course changes, and which questions are still open.
Showing 1 of 1 recommendation
Briefing
Briefing Built
◆◆◆◆◆Excellent Fit
What Does That Completion Row Actually Prove?
From a working tracker whose completion numbers have never been questioned, to knowing exactly what they store, how they behave when the course changes, and which questions are still open.
You'll build: Four fictional-data tests run on the maker's own learning tracker, plus a short dated note recording what each completion record stores, what happens when the course is edited, and anything still UNKNOWN.
Includes: Four fictional-data tests · What to store checklist · Six-line completion note · Supplier walkthrough questions
completion claimstable learner identityenrolment pinned to a course version+6 more
Root Cause
Finding where this blocker actually starts
We traced backward through five layers of "why" until we hit the source. Here's what's really driving this.
1
Why can’t the maker trust the 100% display by itself?
The screen presents a current outcome but does not necessarily expose the saved records and relationships that produced it.
2
Why are the visible name and course title not enough?
Display names and titles can be duplicated, renamed or reused. Stable system identities are needed to distinguish the learner, course/run/version and participation record.
3
Why can today’s result differ from yesterday’s meaning?
Completion can depend on configured criteria, current course content, recalculation, overrides, package behaviour or a new enrolment or attempt. A percentage calculated on demand always measures today's course, so editing the course silently changes what earlier results appear to mean.
4
Why do repeated actions and corrections create more uncertainty?
A replay, retake, correction, void or re-enrolment can create another event or date. Without an explicit history model, nobody can say which result was the first valid completion and which later records supersede or qualify it.
5
Why does the problem survive a polished dashboard or a working build?
Demos and builds naturally focus on the current learner journey and visible result. Historical reconstruction is a different test that requires saved-record read-back and deliberately modelled relationships.
Root Cause
Current-state progress and historical completion evidence are different jobs. The problem appears when a product or custom build shows the current result but does not expose, or has not modelled, the stable identities, version context, rule, saved outcome and history needed to reconstruct what the result originally meant. A percentage calculated on demand is the common shortcut that causes it.
The Numbers
How this stacks up
Key metrics that determine the opportunity value.
Overall Impact Score
6/100
Urgency
6/10
Moderate pressure to solve
Build Difficulty
9/10
Complex, needs deep expertise
Market Size
5/10
Healthy demand exists
Competition Gap
7/10
Moderate competition
"A user sent a screenshot of the 'Well done! The course is completed' screen for a course. The problem is: I can't find any record of this completion in Docebo."
Posted by dhillock. A practitioner had a learner screenshot showing completion but could not find the result in the listed reports and audit surfaces. A reply the following day from mstrom, a Docebo Pro Certified Admin, offered a plausible cause: a SCORM module left open for a long period can lose its connection to the platform, so the learner sees the success screen and the platform records nothing. Re-verified by live fetch on 2026-08-18. One reported case, not prevalence. — Docebo Community — Can't find any record of a course that user says they completed, 2025-08-20
More Evidence
What others are saying
"Now, I'm not seeing last year's completion details in my transcript."
Posted by Allison.M. A practitioner archived and re-enrolled for annual compliance training, then could not locate the prior year's completion in the transcript. The thread later identified archived-enrolment reporting, so this proves retrieval friction rather than permanent loss. CORRECTED 2026-08-18: this entry previously recorded the quote as 'Where can I see that I completed this training last year?', which is not the text on the page. Verbatim text re-verified by live fetch on 2026-08-18. — Docebo Community — Archive and Re-Enrollment: Annual Course Renewals, 2026-06-12
"SCORM status has now changed to incomplete"
A practitioner supplied repeatable steps where an updated package and resume action changed a completed status. Extracted 2026-08-04. NOT RE-VERIFIABLE: the URL returned HTTP 403 on 2026-08-18, so downstream content must reference this as a reported case without presenting the wording as a confirmed verbatim quotation. Product and update specific; it does not show that all LMS or package updates behave this way. — Articulate Community — Recent update to Storyline breaks LMS resume, 2025-11
The Landscape
What solutions exist today?
Current market solutions and where there are opportunities.
T
Trust the dashboard, percentage or certificate
Approach: Accept the visible completion result as sufficient.
Weakness: It may answer the current-state question while hiding which learner identity, course version, criteria, attempt and saved result support it.
C
Calculate the percentage on demand
Approach: Work out completion from current lesson progress whenever it is displayed.
Weakness: It always measures today's course. Adding or removing a required lesson silently changes what every earlier result appears to mean, without anybody editing a completion record.
E
Export a standard LMS completion report
Approach: Use the product's learner, course or learning-history report.
Weakness: Reports can be useful, but somebody still has to check whether stable identities, versions, repeated attempts and historical criteria are included or joined correctly.
U
Use native archived-enrolment, attempt or version history
Approach: Rely on product-specific historical records and reports.
Weakness: This can be strong evidence when configured and available, but the location, retained fields and correction behaviour vary by product.
The Gap
Why existing solutions keep failing
The pattern they all miss — and how to beat it.
Common Failure Mode
Most products can show current progress, reports or certificates. The gap appears when somebody must reconstruct one historical claim across stable identity, enrolment or attempt, course version, applicable requirements, rule, saved result and event time. A screenshot or title is not the same as that chain.
How to Beat Them
Write for the maker who just built the thing, not the reviewer evaluating somebody else's. Answer the two objections that decide whether they keep reading, that a percentage calculated on demand always measures today's course, and that versioning may be optional but knowing what happens to a finished course when you edit it is not. Then lead with four fictional-data tests they can run on their own lab in fifteen minutes: partial progress, replay, version change and same name. Reframe the six record jobs as what to store rather than what to interrogate, keep the record to a few honest lines with UNKNOWN allowed, and keep the reviewer material as a short closing section.
The Fix
What a solution needs to succeed
The non-negotiables and nice-to-haves for any product or service tackling this blocker.
The 3 Wishes
Tell the maker, in fifteen minutes on the app they just built, whether their completion records will still mean the right thing after they edit the course.
Must Have
An opening that names the Blueprint and starts inside the finished lab, so the asset does not read as unconnected
A friendly explanation of why a displayed completion is a claim to trace rather than proof by itself
A direct answer to 'can't we just recalculate the percentage?'
A direct answer to 'do we really need course versions for a small internal LMS?'
Four fictional-data tests the maker can run on their own build: partial progress, replay, version change and same name, each with its failure signature
What to store: course version and rule, lessons and required status, stable learner ID, enrolment pinned to a version, per-lesson progress, and a saved whole-course result freezing counts, time and version
Acceptance of equivalent product structures rather than six prescribed tables
A short honest record with UNKNOWN allowed
A plain three-way outcome
Explicit separation from competence, compliance, security and production approval
Nice to Have
One fictional learner journey
A short closing section for readers evaluating a bought system
Power Apps and Dataverse implementation notes
Pointers to xAPI or product-native alternatives without prescribing them
Out of Scope
Building the complete LMS
Teaching Dataverse table creation
Security, privacy, accessibility or licensing review
Compliance, legal or audit certification
Proving learner knowledge or competence
Quantified ROI or market demand
Success Metrics
The reader can explain in ordinary language what a completion result does and does not prove
The reader can say what happens to an existing completion when the course is edited
All four tests have been run on fictional data with an intended, not accidental, result
Anything unanswered is recorded as UNKNOWN rather than assumed
The reader treats the outcome as a data check rather than certification
The Briefing makes sense to someone who has just finished the Blueprint and to someone who arrived with a SharePoint tracker
Solution Strategy
Which approach fits you?
A Course would over-teach one decision, and a Blueprint would prescribe a build. The maker needs four runnable tests, a short list of what to store and a plain outcome. The reviewer angle is a closing section rather than the frame, because the Board places this directly after the build where the reader owns the app.
What we recommend
Use the rewritten Briefing 'What Does That Completion Row Actually Prove?' as the first and only solution recommendation, written maker-first and opening on the Blueprint.
The Future
What might make this blocker obsolete
Technologies and trends that could disrupt this space. Factor these into your timing.
medium probability
Emerging; product-specific
LMS products expose richer completion history by default
The reader would still need to verify what the native view includes and preserve the decision boundary, but fewer manual checks would be required.
medium probability
Uncertain
Learning standards carry more version and correction context
The Briefing should map equivalent facts instead of prescribing six physical tables or one standard.
medium probability
Plausible current design risk; not measured
AI-generated LMS designs look complete before their history model is tested
The four tests remain useful because they exercise behaviour rather than inspect a diagram.
For Creators
Content Ideas
Marketing hooks, SEO keywords, and buying triggers to help you create content around this blocker.
Buying Triggers
Events that make people search for solutions
- A dashboard or certificate says complete but the underlying record cannot be found
- The course is being updated while learners have progress
- A learner must retake annual training or is re-enrolled
- The same course title or learner display name appears more than once
- A Power Apps maker is choosing tables and relationships for completion tracking
- A maker has just finished a learning tracker and wants to stress-test it before real learners arrive
Content Angles
Attention-grabbing hooks for your content
- What does 100% complete actually prove?
- A learner name and course title are not a historical record
- What happens to a finished course when you add a lesson to it
- Why calculating the percentage on demand always measures today's course
- Progress, completion and competence are different claims
- Four fictional-data tests that find real completion bugs in fifteen minutes
- The saved records to request before accepting an LMS demonstration
Search Keywords
What people type when looking for solutions
LMS completion record missingprove course completion versionLMS course version completion historylearner completion evidencecompletion percentage audit trailcourse update changed completion statusLMS re-enrolment historical completionPower Apps LMS completion data model
The Evidence
Where this came from
Every claim in this report is backed by public sources. Verify anything.
1.2.3.4.5.
Can't find any record of a course that user says they completed
community.docebo.com
Archive and Re-Enrollment: Annual Course Renewals
community.docebo.com
Recent update to Storyline breaks LMS resume
community.articulate.com
1EdTech OneRoster Standard
standards.1edtech.org
Moodle Course Completion
docs.moodle.org
9 sources referenced
Source note
Blocker published by Collab365 Spaces, reviewed by Collab365 editorial team on . Cite as "My Power Apps LMS says 100% complete, but I can’t prove what this person completed", Collab365 Spaces. 9 sources referenced.
Have a question or correction?
No comments yet