Exposure is not the outcome.
A patient can receive an appointment reminder and still not attend. A user can open onboarding and still not complete setup. An employee can read a change announcement and remain unable to explain what changes on Monday. A leader can view a dashboard and act on the most vivid number rather than the most diagnostic one.
These are often compressed into a single phrase—“the communication failed.” But several different transitions may have failed:
Violet calls each meaningful movement in this chain a transition. The point is simple: find the first movement that is not happening.
Evidence at one stage does not establish success at the next. An open is not comprehension. Comprehension is not agreement. Agreement is not action. Action is not durable benefit.
Start with ordinary questions. Learn the Violet terms as you go.
The same message can work for one person and fail for another because people bring different goals, knowledge, trust, and constraints. Violet names the parts so a team can diagnose the difference without blaming the person or rewriting everything.
You do not need every answer. Record what is known, label what is assumed, and focus on the first broken transition.
The audience is not a demographic label.
Receiver state means what the person brings to this moment. A useful Receiver State Map includes only what could affect the next step:
- Role and identity
- Which responsibilities or affiliations are active in this situation?
- Expertise
- Which concepts and distinctions can be processed efficiently?
- Active job
- What progress is the person trying to make right now?
- Trust posture
- What does the receiver expect about the source’s competence and incentives?
- Memory and exposure
- What has been seen before, and what happened afterward?
- Constraints
- What limits attention or action: time, access, cost, stress, authority, or workflow?
The same person can occupy different states across a journey. A clinic visitor may first be asking whether the service is relevant, then whether it is credible, then what it costs, and only later whether to book.
The appointment time and preparation instructions are likely to receive weight.
The cancellation policy, consequence, and route become more relevant—even for the same patient.
Replace “engagement” with one observable next state.
A target transition is the next observable movement you want to support. “Improve engagement” hides too many possible outcomes. “Increase informed cancellations at least 24 hours before an appointment” is specific enough to diagnose and measure.
The receiver’s goal must also be checked against the organization’s goal. If a user first needs to compare risk, establish eligibility, or understand recourse, “Book now” may be a premature transition.
Notice what surrounds the message.
The contextual field is the situation around the message: where and when it appears, who is present, the device being used, privacy, social rules, risk, and which actions are possible. The words may stay the same while the practical meaning changes.
They can open a calendar, discuss transport, and call the clinic without being overheard.
Privacy, motion, noise, and the inability to call may postpone action until the message is forgotten.
Opening health details or stepping away to call may be inappropriate even when the reminder is understood.
The contextual field is always present. Sometimes the honest audit answer is simply that it is unknown.
What happens just before can change what people notice next.
An induction cue is a question, instruction, image, comparison, or earlier interface step immediately before the main message. It can briefly shape what the person looks for next. Violet calls that possible temporary mindset an operative state; it is a hypothesis unless it has been measured.
A receiver who chose “Reschedule it” may now give greater weight to the rescheduling route and deadline.
The prompt may help organize attention, but it can also narrow or manipulate it. It should surface a legitimate goal, not manufacture urgency or hide alternatives.
Five memorable questions reveal different kinds of failure.
Position, order, contrast, novelty, repetition, urgency.
Example Put “Thursday at 10:30 a.m.” in the first line, not beneath a generic greeting.Truth, specificity, completeness, diagnostic relevance.
Example Name the clinic, visit type, date, and rescheduling deadline.Expertise, relationship, incentives, accountability, legitimacy.
Example Send from the clinic’s recognized number and identify who can answer questions.Length, ambiguity, missing knowledge, steps, emotional burden.
Example Replace a phone-tree search with one direct reschedule link.Reward, avoided loss, safety, status, identity, practical relevance.
Example Explain that early cancellation releases the slot and avoids a missed-visit fee, if true.The 5S terms are questions, not universal scores. The same detail can add useful Substance for an expert and create unnecessary Strain for someone who needs the decision first.
One unsupported jump can break an otherwise clear journey.
People rarely move directly from seeing information to a final result. Each step must give them what the next step requires.
Violet calls fixing the largest unsupported jump cognitive path smoothing. In plain language: make the missing step visible, remove an unnecessary detour, or put needed information where the decision happens.
The receiver must infer where rescheduling lives, remember credentials, and search the portal.
Each step supplies what the next step requires while preserving a route to assistance.
More information is not always the bridge. Guidance that helps a novice can become redundant for an expert. The path is receiver-relative.
An “efficient” AI review can sound like unaccountable automation.
“We are introducing an AI system to review customer applications and improve processing efficiency.”
Will a machine decide? Who is accountable? Can I challenge an error? What happens to my data?
“AI will help trained reviewers sort applications; it will not make the final eligibility decision. A reviewer remains accountable. If your application is declined, your notice will include a route to human reconsideration.”
The redesign does more than make the sentence friendlier. It identifies who remains accountable (Source), explains what the system does and does not do (Substance), and preserves a route to challenge an error.
Evidence for the parts, not proof of every Violet prediction.
Violet brings together established research on attention, goals, trust, effort, framing, memory, social influence, and step-by-step learning. The evidence cases below show representative pieces. They do not prove that every Violet diagnosis will be right.
The framework should be narrow enough to fail.
What Violet proposes
- What people bring to a situation should affect what they notice and use.
- Different next steps should require different combinations of cues.
- Fixing the actual broken step should work better than adding the same amount of information elsewhere.
- Repeated exposure may make a later step easier even before a person reports changing their mind.
What Violet does not claim
- That internal receiver state can be known with certainty.
- That every behavioural failure is a communication failure.
- That a click or conversion proves understanding or benefit.
- That AI is a new cognitive mechanism or evidence for the theory.
Change what is responsible for the break.
Not everything affecting the next step can or should be changed. Separate what you control, what you can influence, what you must design around, and what you still need to learn.
Wording, order, prominence, evidence, direct link, send time.
Channel, device support, privacy, meeting format, location.
Expertise, trust posture, prior experience, active job.
Where the reminder is opened, why action is delayed, which step fails.
The goal is not maximum attention or compliance. It is to make an appropriate, informed next step more likely while preserving truth, real alternatives, and the person’s ability to decline.