What is the ADDIE model?
The ADDIE model is a five-phase instructional design framework: Analysis, Design, Development, Implementation and Evaluation. Training teams use it to plan a course from the first needs check to the final review, with each phase producing an output the next phase builds on.
Example
A logistics company’s L&D team is asked for “forklift refresher training” after a rise in warehouse incidents. In the Analysis phase they talk to shift supervisors and read the incident reports. Most problems turn out to happen during loading at the dock, not during general driving, so the brief narrows to one task.
In Design, the team writes three objectives and outlines a 15-minute module with a short video, two practice scenarios and a quiz. Development produces the module and a supervisor checks it on the dock. Implementation puts it in the LMS and assigns it to every driver on the night shift. Evaluation compares quiz results and dock incident reports over the following quarter, and what the team learns goes into the next revision.
How it works
ADDIE is a process model for instructional design. Its phases trace back to instructional systems development models built for the US armed forces in the 1970s, and it is often the first framework new instructional designers learn. Each phase answers one question and hands a concrete output to the next.
Analysis
The team defines the performance problem, the audience, and the constraints, such as time, budget, languages and the devices learners use. Analysis also asks whether training is the right fix at all. The usual output is a training needs analysis with the gaps, the audience and the measures of success.
Design
The team writes learning objectives, chooses the format (self-paced e-learning, video, classroom or a mix), and plans the structure, activities and assessments. The output is a design document and often a storyboard for each module.
Development
Developers and designers build the materials from the design: scripts, videos, slides, e-learning modules, job aids and quiz questions. Subject matter experts review the drafts, and the team usually runs a pilot with a small group before launch.
Implementation
The course goes live. That means uploading it to the LMS, assigning it to the right people, scheduling any live sessions, briefing facilitators and telling learners why the training matters to them.
Evaluation
Evaluation runs in two forms:
- Formative evaluation happens during every phase: reviews, pilots and fixes before launch.
- Summative evaluation happens after launch and asks whether the training worked. Many teams structure it with the Kirkpatrick model, from learner reaction up to business results.
ADDIE vs SAM and agile approaches
Classic ADDIE runs like a waterfall: one phase finishes before the next begins. The common criticism is that this is slow, and that problems found late in development are expensive to fix. In practice, most teams now run ADDIE iteratively and loop back to Design when a pilot exposes a problem.
The Successive Approximation Model (SAM) takes a different route. It builds a rough prototype early, reviews it with stakeholders, and refines it in short cycles. Agile approaches borrow sprints and backlogs from software development. Both keep the same activities as ADDIE (analyze, design, build, test) and repeat them in smaller loops.
ADDIE tends to suit regulated or high-stakes training, where documented sign-off at each stage matters. Iterative models suit content that changes often or projects where stakeholders cannot say what they want until they see a draft.
Common mistakes
- Skipping Analysis because the request already names a solution, such as “we need a course on X”.
- Writing objectives after the content is built, so they describe the content instead of shaping it.
- Treating Evaluation as an end-of-course satisfaction survey only.
- Launching a course and never revisiting it, even when evaluation data shows a problem.
Why it matters for L&D teams
- A shared plan. ADDIE gives L&D, subject matter experts and stakeholders the same map of who does what, and when.
- Fewer wasted builds. Analysis catches requests where training is not the fix before anyone spends weeks building a course.
- Traceable decisions. Each phase leaves a document behind, which helps with sign-off, audits and handovers between team members.
- Built-in improvement. Evaluation data feeds the next version of the course instead of sitting in a report.
How Colossyan helps
Colossyan Learn shortens the Design and Development phases of ADDIE. Start from a document, a brief or a topic, refine the course outline of modules and lessons that Learn generates, then build the lessons with AI video, text and assessments. Implementation is a SCORM export to your LMS (a paid-plan feature), and when evaluation shows a lesson needs fixing, you edit that lesson and only its video regenerates.
See the AI course creator
