What Is AIDMA? Five Stages, AIDA and AISAS, and Practical Limits
Published:
Last Updated:
Category: Marketing Glossary
Authors: Shusaku Yosa
AIDMA stands for Attention, Interest, Desire, Memory, and Action. It is a classic model for organizing possible psychological stages between discovering an offer and acting on it. The five questions are: Have people noticed it? Is it relevant? Do they want it? Can they recall it when needed? Can they take the next step?
The model helps a team identify missing information and develop research questions. It does not establish that every buyer follows five steps in order. A person can search, compare, pause, return, reject an offer, or buy quickly after a recommendation. Begin with the full form, then check where the model matches real behavior and where it leaves something out.
The five stages of AIDMA
The following table translates the stages into practical questions. Activities and metrics are examples, not proof of a person's mental state. Someone may click because they are confused, and someone who never clicks may already understand the product.
Stage | Possible buyer question | Information to provide | What to investigate |
|---|---|---|---|
Attention | What is this product for? | A clear use case and introduction | Whether the intended audience recognizes the offer |
Interest | Does it relate to my problem? | Relevant situations and explanations | What people want to learn and what remains unclear |
Desire | Would the benefit justify the effort and cost? | Comparisons, conditions, demonstrations | Expected value and reasons for hesitation |
Memory | Can I find or recall it when I need it? | Useful material people can save or request | How and why they returned to the decision |
Action | How do I buy, request help, or proceed? | Pricing, requirements, and a clear next step | Completion, errors, and obstacles |
Memory means remembering an offer in this model; it is not the same as customer retention. Nor does it justify repeatedly contacting everyone who downloaded a resource. Ask what information people need, what timing they prefer, and whether they want further contact. For an operational approach to ongoing information, see lead nurturing.
AIDMA vs. AIDA vs. AISAS
AIDA comprises Attention, Interest, Desire, and Action. AIDMA explicitly includes Memory. AISAS comprises Attention, Interest, Search, Action, and Share. It draws attention to people actively finding information and sharing experiences; it is not simply AIDMA with one word replaced.
Model | Full form | Useful emphasis | Questions still to ask |
|---|---|---|---|
AIDA | Attention, Interest, Desire, Action | Motivation and information before action | What happens during a pause or after purchase? |
AIDMA | Attention, Interest, Desire, Memory, Action | Recalling an offer across time | How do comparison, search, and multiple people affect the decision? |
AISAS | Attention, Interest, Search, Action, Share | Search and sharing behavior | What about people who do not search or share? |
Dentsu's own explanation dates its proposal of AISAS to 2004 and describes Search and Share in the model. See Dentsu's AISAS explanation. That company explanation supports this attribution; it does not prove that all purchases follow the model.
Choose a model because it helps answer a decision, not because it appears newer. Start with customer needs and the value being offered, as discussed in the foundations of marketing. A framework cannot substitute for that understanding.
Worked example: a buyer choosing an office chair
Consider a fictional buyer journey. A remote worker learns about a chair from a friend, searches for its dimensions, tries a competitor's chair, pauses while checking their budget, returns to the original product, and orders it. After purchase, they check assembly instructions and the return policy.
Attention may help label the initial recommendation. Interest may be a hypothesis about the dimension search. Desire may relate to the expected comfort, and Memory may help frame the later return. However, the dimensions could also be a reason to reject the chair, and returning could result from rediscovering a comparison page rather than recalling a brand. Ask the buyer before accepting either explanation.
Observation | Initial hypothesis | Alternative explanation | Follow-up question |
|---|---|---|---|
Repeated visits to dimensions | Serious consideration | The chair may not fit | What space or requirement matters? |
A return visit weeks later | Product recall | A comparison site surfaced it again | What prompted the return? |
Leaving the order process | Lower desire | Delivery, payment, or usability problems | What prevented completion? |
Viewing returns information | Dissatisfaction | Checking conditions in advance | Is there a problem with use? |
Research by Google and The Behavioural Architects examines repeated exploration and evaluation before purchase. Google's purchase-decision research offers a useful counterpoint to a purely linear diagram. Its findings should still be interpreted within the research context rather than generalized to every country, product, or business purchase.
In B2B, the user, evaluator, budget owner, and approver may be different people. One person's five-stage diagram cannot explain the whole buying group. Build research-based personas for relevant roles and record whose question or constraint each hypothesis addresses.
Why clicks and direct traffic do not measure the stages
An impression is not a person who paid attention. A click is not the complete population of interested people. Direct traffic does not consist only of people who remembered a brand and typed its address; it can also include traffic without identifiable referral information. Renaming these metrics as psychological stages creates precision the data does not support.
Email metrics have additional limits. Mailchimp's official guidance describes the effects of Apple Mail Privacy Protection and bot activity on open and click reporting. An increase in opens is insufficient to conclude that more people remembered the product or became ready to buy.
In a fictional numerical example, an ad receives 10,000 impressions, 100 clicks, and five attributed purchases. Its click-through rate is 100 ÷ 10,000 = 1%. Those counts do not mean that 10,000 people noticed the ad, 100 entered an interest stage, and five developed desire. Counts, users, periods, and attribution rules are different questions.
Do not divide a brand-awareness survey sample by GA4 sessions to calculate a stage conversion rate. Use interviews or a suitable recall study for memory, task observation for usability obstacles, and defined event data for completed actions. Even when a behavioral metric changes, a separate comparison is needed to assess whether a particular intervention caused the change.
A worksheet for turning hypotheses into decisions
Copy this structure into a working document. Leave unknowns visible instead of completing them with invented demographic detail or assumed motivations. One useful question is enough for an initial trial.
Field | What to record | Fictional example |
|---|---|---|
Audience and decision | Who is deciding what? | A small-business employee choosing office furniture |
Evidence | Observation, statement, date, and sample | Interviewees asked when delivery could be guaranteed |
Hypothesis | Possible uncertainty or motivation | Unclear delivery timing may be delaying comparison |
Intervention | Information and available next step | Explain delivery conditions and offer a timing inquiry |
Evaluation | Audience, period, metric, and other explanations | Check understanding and useful inquiries alongside order completion |
Ownership | Who reviews the result and when? | Assign an owner, review date, and revision criteria |
A missing Memory-stage metric is not automatically a reason to buy another tracking tool. First ask whether people can retrieve the information they need and whether the proposed contact is welcome. Use CTA design to make the next step understandable and A/B testing when a suitable comparison can address the hypothesis.
Review checklist and limits
- Separate the five definitions from observations about your own audience.
- Include pauses, repeated comparison, rejection, and post-purchase use.
- Do not treat clicks, opens, or direct sessions as verified psychological states.
- Define units, periods, denominators, cost scope, and duplicate handling.
- Assign someone to revise a hypothesis when customer evidence contradicts it.
- Keep behavioral changes, financial outcomes, and causal conclusions separate.
If the model repeatedly fails to describe observed behavior, return to customer evidence rather than adding stages until the diagram fits. Organizing the associated initiatives, owners, budgets, and actual results can make reviews easier. Xtrategy's features include project and budget-versus-actual management that can support this work. Such organization does not automatically identify psychological stages or guarantee marketing results.
