ADDIE vs SAM: Choosing the Right Instructional Design Model
ADDIE gives you structure and documentation. SAM gives you speed and early feedback. Here is how the two models differ, when each one fits, and the hybrid I use on most projects.
ADDIE and SAM are the two most common process models in instructional design. ADDIE (Analysis, Design, Development, Implementation, Evaluation) is a linear, phase-by-phase model that front-loads analysis and documentation. SAM (Successive Approximation Model), introduced by Michael Allen, is an iterative model that replaces long specifications with rapid prototypes reviewed early and often.
Neither is "better." ADDIE fits projects where requirements are stable and sign-off matters. SAM fits projects where stakeholders don't fully know what they want until they see it. Most of my real projects use a bit of both.
How ADDIE works
ADDIE has roots in systems approaches to military training in the 1970s and became the default vocabulary of the field. Each phase produces an output that feeds the next:
- Analysis: audience, performance gap, constraints, success measures.
- Design: objectives, assessment strategy, structure, design document.
- Development: storyboards, media, built course, quality checks.
- Implementation: launch, LMS setup, facilitator prep, communication.
- Evaluation: formative checks throughout, summative measures after launch.
The strength of ADDIE is clarity. Everyone knows what has been agreed and what comes next. The weakness is that the first time stakeholders experience the actual course is often late in development, when changes are expensive.
How SAM works
Michael Allen described SAM in Leaving ADDIE for SAM (ASTD Press, 2012). It has a short preparation phase, a "Savvy Start" brainstorming session with stakeholders and learners, and then repeated cycles of design, prototype and review.
The key idea is that a rough prototype is a better conversation starter than a polished document. People who struggle to react to a 30-page design document will happily tell you a scenario feels unrealistic when they click through it.
ADDIE vs SAM at a glance
- Flow: ADDIE is sequential; SAM is iterative.
- Main artifact: ADDIE relies on design documents and storyboards; SAM relies on prototypes.
- Stakeholder input: ADDIE gathers it at phase gates; SAM gathers it continuously.
- Best fit: ADDIE for compliance, regulated or fixed-scope work; SAM for new audiences, interactive or scenario-heavy work.
- Main risk: ADDIE risks late surprises; SAM risks scope creep if review cycles aren't capped.
When I choose ADDIE
I lean toward a classic ADDIE flow when content is dictated by regulation or policy, when there are many approvers who need a document trail, or when the format is well understood, such as converting an existing instructor-led session into a self-paced module with the same content.
In these cases, the documentation isn't bureaucracy. It's what lets legal, HR or compliance sign off once instead of five times.
When I choose SAM
I switch to SAM-style iteration when the learning experience itself is the uncertain part. Scenario-based courses are the clearest example. You don't know whether a branching scenario feels realistic until someone who does the job plays through it. The same applies to new audiences, new technology like AI-supported practice, or projects where the client says "we'll know it when we see it."
The hybrid I actually use
On most projects I keep ADDIE's analysis rigor and SAM's prototyping loop:
- Analysis stays upfront. A short needs analysis and an action map define the goal and the behaviors before anything is built.
- One lightweight design doc. Objectives, audience, assessment approach and a module list, on two or three pages.
- Prototype the riskiest piece first. Usually one complete scenario or one module, built quickly and reviewed with two or three real learners.
- Capped iteration. Two review rounds per module, agreed upfront, so iteration doesn't become endless.
- Evaluation plan from day one. What we'll measure after launch is defined in analysis, not invented at the end.
This keeps stakeholders confident while still catching the expensive problems early.
Where to start
If your projects keep hitting late-stage rework, don't throw out your process. Add a single prototype checkpoint after design and put it in front of actual learners. That one change captures most of SAM's benefit without retraining your whole team.
Sources
- Michael W. Allen with Richard Sites, Leaving ADDIE for SAM (ASTD Press, 2012)
- Allen Interactions, The SAM process
- Robert Maribe Branch, Instructional Design: The ADDIE Approach (Springer, 2009)