ADDIE Model: The 5 Phases Explained Simply, With a Worked Example for an Online Course

Updated 31 October 2026 · Facts checked against the sources listed at the end

The ADDIE model is a step-by-step way to plan and build a course. ADDIE stands for Analysis, Design, Development, Implementation and Evaluation. First you find out who the learners are and what they need (Analysis). Then you plan the goals, tests and order of lessons (Design), make the lessons (Development), run the course (Implementation), and check whether it worked so you can improve it (Evaluation). It grew out of a 1975 training model made at Florida State University, and it is still one of the most used ways to design instruction.

Most pages about ADDIE are written for company training teams. This guide is written for coaching institutes, tuition centres and teachers in India who want to build an online or recorded course that students actually finish and learn from. It explains each phase in simple words, with the questions to ask and the output you should have at the end. It then walks through a full worked example, a four-week recorded maths course, and gives you a one-page ADDIE template, the common mistakes, and an honest note on when ADDIE is more than you need.

What the ADDIE model is

ADDIE is an instructional design model. Instructional design simply means planning teaching on purpose: deciding what learners must be able to do, then building lessons and tests that get them there. ADDIE gives that planning five phases, and each phase ends with something you can hold in your hand, such as a list of goals, a course plan or a finished set of videos.

The University of Washington Bothell describes ADDIE as a device to help you think through a course's design. It also says the model looks like a straight line but does not have to be followed rigidly, especially when you already have course material. That point matters for Indian coaching institutes, because most already have notes, question banks and recorded classes. ADDIE does not ask you to throw them away. It asks you to check them against clear goals.

Phase	The main question	What you have at the end
Analysis	Who are the learners, what must they be able to do, and what stands in the way?	A one-page analysis: learners, goal, gaps, time, budget, devices
Design	What exactly will learners be able to do, how will we test it, and in what order will we teach it?	Learning objectives, a test for each objective, and a course map
Development	What do we need to make, and does it work?	Finished videos, notes and tests, checked in a small pilot
Implementation	How do we run the course well for real students?	Trained teachers, enrolled learners, a launch and support plan
Evaluation	Did learners reach the goals, and what should change?	Results, feedback and a list of changes for the next batch

The five phases of the ADDIE model, as described by UW Bothell and the 1975 IPISD manuals

ADDIE is a planning habit, not software. You can use it with paper notebooks, a spreadsheet or any learning management system. The phases are the same.

Where the ADDIE model came from

The ideas behind ADDIE come from military training in the United States. In 1975 the Center for Educational Technology at Florida State University wrote a five-part manual called the Interservice Procedures for Instructional Systems Development (IPISD), sponsored by the Naval Education and Training Command. Its aim was to train people to do a particular job well, and it says the same steps can be applied to any curriculum development work.

The five phases in that manual were Analyze, Design, Develop, Implement and Control. The last phase was later called Evaluate, which gives the familiar ADDIE. Wikipedia's summary notes that the origin of the ADDIE label itself is not clear, and that the dynamic, less step-by-step version used today appeared by the mid-1980s. So when a page says one person invented ADDIE, treat that with care: it grew over time from a team's work.

Why does this history matter to a tuition centre? Because the original manuals were built around one practical question that suits coaching well: what must the learner be able to do at the end, and how will we know?

Phase 1: Analysis, or finding out what is really needed

In the Analysis phase you clarify the problem, set the overall goal, and learn what your learners already know and what they lack. UW Bothell lists the questions: who your learners are, what overall goals you are trying to reach, what knowledge and skills must be taught, how much content is needed, and what resources are needed and available.

The 1975 Analyze manual adds a step many institutes skip: analyse existing courses. Before you record anything new, look at what you already teach and ask whether it was built in a way that meets today's needs. Often the honest answer is that the content is fine but the order, the practice or the tests are weak.

Questions to answer, for an Indian coaching course

Who exactly is this for? Class, board, medium of instruction, and the level of a typical student, not the best one.
What problem are we solving? Weak basics, no time for revision, students who joined late, students in another town?
What must a student be able to do at the end that they cannot do now? Write it as an action, such as solve, explain, write or speak.
What do students already know? A short entry test on the first day of a batch tells you more than a guess.
What devices and data do they have? A phone with limited data changes video length, file size and the use of PDFs.
How much time do they have each week, around school and other classes?
Who will teach, record and answer doubts, and how many hours can they give?
What do we already have: notes, past recordings, question sets, worksheets?

Output: a one-page analysis. If you cannot fit the answers on one page, the course is probably trying to do too much. Split it.

Phase 2: Design, or planning the goals, tests and order

Design is where most of the real thinking happens. UW Bothell says this phase deals with learning objectives, assessment, exercises, content, lesson planning and media, and that it should be systematic and specific. The 1975 Design manual breaks it into four steps, and they still work well.

Write the learning objectives. The manual describes an objective as what is to be learned in terms of expected student performance, under stated conditions, to an accepted standard. In plain words: what the student will do, with what help or tools, and how well.
Write the tests next, before the lessons. The manual's second step is to develop tests that match the objectives. If an objective says solve, the test must ask the student to solve, not to choose a definition.
Describe the entry level. The third step looks at what students bring with them, which sets the limits of the course and shows where a short bridge or revision lesson is needed.
Fix the sequence. The fourth step orders the objectives so that supporting skills come before the skills that depend on them.

How to write a good objective

Weak objective	Why it is weak	Better objective
Understand trigonometric ratios	You cannot see or test understand	Given a right triangle with two sides marked, find sin, cos and tan of an angle, with at least 8 of 10 correct
Know the tenses	No action, no standard	Rewrite 10 sentences from present to past tense with no more than 2 errors
Learn GST basics	Too wide, no conditions	Using a given bill, work out the GST amount at a stated rate, correct to the rupee
Improve spoken English	No measurable result	Introduce yourself for one minute without notes, covering name, studies and goals

Each better objective has an action, a condition and a standard, as the 1975 Design manual describes

Output: a course map. For each module list its objectives, the test that checks each one, and the lessons in order. This map is the blueprint for the next phase, and it is the page you will return to when you evaluate. Our guides on formative assessment and rubrics for assessment help with the testing side.

Phase 3: Development, or making and checking the material

Development is where you build what the design asked for: scripts, slides, videos, notes, worksheets and test questions. UW Bothell describes it as creating and assembling the content planned in the design phase, including writing content and designing graphics, and working on the technology if the course is online.

The 1975 Develop manual has two steps that save a lot of effort. One is to review and select existing materials: check what you already have against the objectives, and reuse what fits. The other is to validate the materials, which means trying them out to make sure the teaching works and students reach the objectives. Wikipedia's summary of the phase calls this a pilot test with key people before launch.

A practical build order

Write the test questions first, from the course map, with an explanation for each answer.
Mark which old recordings and notes still match an objective. Keep those.
Script only the missing lessons. One lesson, one objective, so videos stay short and easy to find again.
Record a sample lesson and watch it on a cheap phone with low volume. Fix sound before anything else.
Run a pilot: give one module to five or six real students and watch where they get stuck.
Fix what the pilot showed, then build the rest.

If you teach with short lessons, our guide to microlearning explains how to cut a chapter into small pieces that still make sense together.

Phase 4: Implementation, or running the course for real

Implementation is the launch and the day-to-day running. UW Bothell says it starts with a plan to prepare both the facilitators and the learners, and that teacher preparation should cover the curriculum, the learning outcomes, the way the course is delivered and how testing works. Wikipedia's summary adds that preparing learners includes training them on any new tools and registering them.

For a coaching institute this phase is often where good courses fail. The videos are ready, but no one told the students how to start, when tests are due, or where to ask doubts. Use this short list.

Brief every teacher on the course map: which objectives, which tests, and who answers doubts and how fast.
Give students a first-day message with the course link, the weekly plan and how to ask for help.
Open the course with a short orientation lesson: how to use the app, how tests work, how progress is shown.
Release lessons at a steady pace if students tend to rush or binge, and keep a fixed weekly time for a live doubt class if you run one.
Watch the first week closely. Low starts in week one are easier to fix than drop-outs in week four.

If your course mixes recorded lessons with classroom teaching, our blended learning guide covers how to divide the work between the two.

Phase 5: Evaluation, or checking what worked

Evaluation has two parts. Formative evaluation happens during every phase, while you are still building and running the course. Summative evaluation looks at the finished course and its results. UW Bothell calls evaluation the least understood phase and says ADDIE is better seen as a continuous circle than a straight line, because what you learn here feeds back into analysis.

In the 1975 manuals this phase was called Control, and it had three steps: an internal evaluation of how learners performed in the course, to find teaching that was missing or not needed; an external evaluation of how well people performed after the course; and a revision of the system based on that evidence.

The Kirkpatrick levels, adapted for a coaching course

Many designers use Donald Kirkpatrick's four levels during this phase. Kirkpatrick Partners describe them as Reaction, Learning, Behavior and Results, and advise starting your planning from the results you want. Here is how they translate for a tuition or coaching course.

Level	What it measures	How a coaching institute can check it
1. Reaction	Whether learners find the course favourable, engaging and relevant	A three-question feedback form after each module; ask about relevance first
2. Learning	Whether learners gained the intended knowledge, skills, confidence and commitment	Compare an entry test with the module tests on the same objectives
3. Behavior	Whether learners apply what they learned in their own setting	School test answers, homework quality, or a teacher's check a few weeks later
4. Results	Whether the outcome you aimed for actually happened	The goal you set in Analysis, such as fewer students failing a unit test

Level names and meanings from Kirkpatrick Partners; the coaching examples are ours

Kirkpatrick Partners also suggest quick checks while a programme is running, rather than waiting for one survey at the end. In a course, that can be as simple as looking at test results and where students stop watching after the first week.

A worked example: ADDIE for a four-week recorded maths course

Here is the whole model applied to one realistic case. A tuition centre in a small city teaches Class 10 maths in the evening. Many students lose marks in trigonometry, and the teachers want a short recorded course that students can use at home, alongside their classes. The example is made up to show the method; the numbers in it are planning choices, not research results.

Phase	What the centre decided	Output
Analysis	Learners: Class 10 students of average level, English and Hindi medium, mostly on shared phones. Problem: they mix up the ratios and cannot use identities. Entry test of 15 questions on day one. Time: about 30 minutes a day at home.	One-page analysis and entry test results
Design	Four modules, one a week: ratios, values of standard angles, identities, heights and distances. Each objective written with an action and a standard. A 10-question test at the end of each module, written before any video.	Course map with objectives, tests and lesson order
Development	Old class recordings reviewed: two kept, the rest re-recorded as short single-topic videos. Each test answer has an explanation. Module 1 piloted with six students for three days.	Videos, notes and tests; pilot notes
Implementation	Teachers briefed on the map and on answering doubts within a day. A first-day message to students and parents. Lessons released one module per week, with a Saturday live doubt class.	Enrolled batch, weekly plan, doubt routine
Evaluation	Feedback after each module. Module test scores compared with the entry test. Watching data checked each week. After the school's unit test, the teachers compared trigonometry marks with the term before.	A list of changes for the next batch

A worked example; the case and its numbers are illustrative

What the pilot changed: in the example, the six pilot students skipped the long introduction video and most got the same question wrong on the module test. The centre cut the introduction into a short opening, and re-recorded one lesson with a slower worked example. This is what the validation step is for: an evening's fix before launch instead of a lost batch after it.

What the evaluation changed: the watching data showed many students stopping in the middle of the identities module. The teachers split the longest lesson in two and added a practice test between them. Without the course map from Design, they would not have known which objective those lessons served or which test to check.

A one-page ADDIE template you can copy

Copy this into a notebook, a document or a spreadsheet, one block per course. Fill it in before you record anything.

Course name and who it is for (class, level, language, device):
The problem in one sentence:
The overall goal: at the end, learners will be able to ...
Entry check: how we will find out what they already know:
Time and people: hours a week for learners; who teaches, records and answers doubts:
Modules (3 to 8), each with 2 to 5 objectives written as action, condition and standard:
For each objective, the test question type that checks it:
Existing material we will reuse, and what must be made new:
Pilot plan: which module, with how many students, for how many days:
Launch plan: first-day message, release pace, doubt routine:
Evaluation: what we will check after each module, and after the course:
Changes for next time (fill in after the course):

If a line takes you more than a few minutes, the decision has not been made yet. That is worth knowing before you record.

Using the ADDIE model with an LMS

A learning management system does not design a course for you, but it makes some ADDIE phases far easier, because it holds the material, runs the tests and records what learners do. Here is where an LMS helps, phase by phase.

Design: sections in the LMS can mirror the modules of your course map, with a test at the end of each.
Development: tests with an explanation for each answer turn every test into a small lesson.
Implementation: free preview lessons, lessons that open a set number of days after enrolment, and announcements keep a batch moving together.
Evaluation: test results and data on where learners stop watching show which lesson or objective needs work.

On Sikhami, coaching institutes and tuition centres get a free LMS that handles these steps: video courses arranged in sections, multiple-choice tests with explanations and a pass mark, lectures that unlock some days after a student enrols, certificates of completion, announcements, doubts asked at a moment in a video, and course analytics that include test results and where students stop watching. You can also turn a YouTube playlist you already have into a course, which suits the reuse step of Development. See the free LMS page and the Help Centre article on how to create a video course to start.

India's own national platform gives a useful model of course structure. The Ministry of Education's SWAYAM says its courses are built in four quadrants: video lectures, reading material that can be downloaded or printed, self-assessment tests and quizzes, and an online discussion forum for clearing doubts. Those four parts map neatly onto the Development phase: for each module, ask whether you have a lesson, something to read, a test and a place for doubts.

ADDIE and other ways to design a course

ADDIE is not the only model. Wikipedia notes that most current instructional systems design models are variations of the ADDIE process, and names the Dick and Carey and Kemp models as other examples, with rapid prototyping as a common alternative. Some organisations extend it: the United States Navy uses a version called PADDIE+M, which adds a Planning phase at the start (goals, budget and schedule) and a Maintenance phase at the end for continuous improvement.

If your situation is ...	A good approach
A new course for a full batch, with time to plan	Full ADDIE, with a written course map and a pilot
You already have recordings and want to fix them	Start at Evaluation: check results and drop-off, then go back to Design for the weak modules
The topic changes often, such as a software skill	Short cycles: build one module, try it, improve it, then build the next, like rapid prototyping
A single workshop or one demo lesson	A light version: one objective, one activity, one check at the end
A long-running course that will last years	Add planning and upkeep, as PADDIE+M does: a budget, a schedule and a yearly review

Choosing how much of ADDIE to use

The common complaint about ADDIE is that it is slow and rigid. That happens when it is treated as a strict one-way line, with each phase signed off before the next starts. Treat it the way UW Bothell suggests, as a circle with overlapping edges, and run small tests early, and most of that problem goes away.

Common mistakes when using ADDIE, and how to avoid them

Skipping Analysis because you know your students. You know your best students. Run an entry test to know the typical one.
Writing objectives with words like understand, know or learn. Use actions you can see: solve, write, explain, label, speak.
Making tests last. Write tests in Design, straight from the objectives, so the lessons are built to reach them.
Recording everything again. Review existing material first and keep what still matches an objective.
No pilot. Even three days with five students will show the confusing lesson and the wrong answer key.
Long videos that cover many things. One lesson, one objective keeps videos short and easy to revise from.
Launching without telling students how the course works. An orientation lesson and a first-day message prevent most early drop-outs.
Only asking whether students liked it. Reaction matters, but also check learning on the tests and what changes in their school work.
Never closing the loop. Write the list of changes after each batch and act on it before the next.

When the ADDIE model is not for you

ADDIE is worth the effort when a course will be used again and again, by many learners, or by several teachers. It is less useful in some cases, and it is honest to say so.

A one-off live class or a quick doubt session. A clear aim and a closing question are enough.
Content that changes every few weeks, where a written course map would be out of date before launch. Use short cycles instead.
A teacher working alone with no time to plan. Use the one-page template, skip the detail, and still run a small pilot.
Courses where the main value is personal coaching or mentoring. The design matters less than the time you spend with each learner.

Even then, the core habit of ADDIE is worth keeping: decide what the learner must be able to do, check it, and improve. For more on choosing tools for tests, see our guide to online exam software.

Frequently asked questions

What is the full form of ADDIE?

ADDIE stands for Analysis, Design, Development, Implementation and Evaluation. Some sources use the verbs instead: Analyze, Design, Develop, Implement and Evaluate. Both mean the same five phases.

Who developed the ADDIE model?

The ideas come from the Interservice Procedures for Instructional Systems Development, written in 1975 by the Center for Educational Technology at Florida State University for United States military training, with Robert K. Branson as the first-named author. Its five phases were Analyze, Design, Develop, Implement and Control. The ADDIE label itself came later, and its exact origin is not clear.

What are the 5 phases of the ADDIE model?

Analysis finds out who the learners are and what they need. Design sets the objectives, tests and order of lessons. Development makes and pilots the material. Implementation runs the course and prepares teachers and learners. Evaluation checks results during and after the course and leads to changes.

Is the ADDIE model linear?

It is often drawn as a line, but it does not have to be used that way. UW Bothell describes it as a continuous circle with overlapping boundaries, and evaluation happens in every phase. In practice you often go back to Design after a pilot shows a problem.

How is the ADDIE model used in education?

Teachers and institutes use it to plan a course or unit: they study the learners, write objectives with a test for each, build or reuse lessons, try them with a few students, run the course, and use results and feedback to improve the next batch. It works for classroom, online and blended courses.

What is the difference between formative and summative evaluation in ADDIE?

Formative evaluation happens during each phase, while the course is being built and run, so you can fix problems early. Summative evaluation looks at the finished course and its results, such as test scores and whether the goal was met.

What can I use instead of ADDIE?

Other instructional design models include Dick and Carey and Kemp, and rapid prototyping, where you build and test small parts quickly. Many current models are variations of ADDIE. For a course whose content changes often, short build-and-test cycles may suit you better than a full plan up front.

Do I need special software to use the ADDIE model?

No. ADDIE is a way of planning, and a notebook or spreadsheet is enough for the plan. A learning management system helps with the later phases, because it holds the lessons, marks the tests and shows where learners stop or struggle.

Sources

Rules and rates change: check the source before you rely on a figure.

More guides for institutes

Get it onGoogle Play