Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for elearning developers

Course Drop-Off Analysis Agent

The biggest drop-off points are found, fixed and shown to have improved.

AnalysisGovernmentEducation
See how it worksGet access
Course Drop-Off Analysis Agent: what goes in, what the agent does and what you get

What it does

Completion rates fall, but the data shows only that learners leave, not why. The agent reads screen-level completion data and finds where learners leave, such as a long video or a hard quiz. It compares segments, like new hires versus experienced staff and mobile versus desktop, to see who leaves where. It then reads the content at those points to find likely causes, such as length, unclear steps or a technical fault, and proposes edits. After the developer changes the course, it waits for new learner data and tests whether the drop-off at that screen shrank. If not, it tries the next likely cause. The developer approves every edit. Edge case: a drop-off that appears only on mobile may be a layout fault, not a content problem.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
ApprovedYes, continueNo 1 STARTS WHEN Monthly run 2 USES A TOOL Read screen-level completion and exit data 3 DOES Find screens with exit rates well above the courseaverage 4 DOES Compare exits by segment and device 5 DOES Read the content at those screens for likely causes 6 DOES Propose edits for each cause 7 YOU APPROVE Developer approves edits 8 USES A TOOL Apply edits and record the change date 9 USES A TOOL Read new learner data after the change 10 CHECKS THE RESULT Did the exit rate at that screen fall by the target? If not: Try the next likely cause and propose a newedit. Back to step 3. 11 RESULT Drop-off report and change log
Read the steps as a list
  1. Monthly run
  2. Read screen-level completion and exit data
  3. Find screens with exit rates well above the course average
  4. Compare exits by segment and device
  5. Read the content at those screens for likely causes
  6. Propose edits for each cause
  7. Developer approves editsThe agent waits here for your OK.
  8. Apply edits and record the change date
  9. Read new learner data after the change
  10. Did the exit rate at that screen fall by the target?If not: Try the next likely cause and propose a new edit. Back to step 3.
  11. Drop-off report and change log

How it decides

It ranks screens by exit rate against the course average and tests causes in order of likelihood.

  • Exit rate above twice the course average is flagged
  • A device-only drop points to a layout fault
  • Wait for at least 50 learners before testing
  • Change one thing per screen at a time

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Exit rate multiple (default 2x)
  • Minimum learners to test
  • Segments to compare
  • Review frequency
  • Target improvement

What keeps you in control

It always asks you first

  • Developer approves each edit
  • Developer approves any change to required content

Hard limits

  • Never edit course content without approval
  • Never test with fewer than the minimum learners
  • Never expose learner identities

It stops when

  • Done: flagged screens meet the target exit rate
  • Stop: not enough learner data

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA 25 screen course loses 18% of learners on screen 9, a 7 minute video. The agent sees the exit is twice as high on mobile and proposes splitting the video and adding captions. After the change, 60 more learners show exit down to 9%. Screen 14 still has a high exit, so the agent proposes shortening its quiz.

More agents for elearning developers

Learning · Education

Learning Misconception Practice Agent

Each learner overcomes their specific misconceptions, with evidence the instructor can see.

8 steps1 check13 jobs
Monitoring · Automation

Course Resource Link Repair Agent

Every course resource works, still matches the lesson and is accessible.

8 steps2 checks2 jobs
Creating · Education

Accessible Lesson Production Agent

Accessible versions always match the latest approved lesson.

7 steps1 check1 job
Learning · Education

Language Conversation Repair Agent

Learners use the target language skill correctly in conversation.

7 steps1 check4 jobs