The problem with advice
If you have ever tried to help someone change a habit, a friend who wants to quit smoking, a partner who keeps skipping the gym, you probably gave them advice. You told them what to do. You explained the benefits. You reminded them of the risks. You may have even laid out a plan.
It probably didn't work.
This is not a failure of the relationship or the effort. It is a failure of the method. Advice, however well-intentioned, activates a psychological mechanism called reactance: the more someone pushes you toward a change, the more you resist it. This is not irrational. It is how humans protect their autonomy.
Motivational Interviewing was developed precisely because researchers noticed this pattern. The question they asked in the 1980s was: what if instead of telling people what to do, you helped them find their own reasons?
What Motivational Interviewing actually is
Motivational Interviewing (MI) is a collaborative, person-centered counseling style developed by clinical psychologists William Miller and Stephen Rollnick. The first formal description was published in 1983 in the journal Behavioural Psychotherapy. The method was originally developed for alcohol addiction treatment, a context where traditional confrontational approaches had consistently failed.
The core insight of MI is simple and counterintuitive: people are more committed to change they argued for themselves than change someone argued for on their behalf.
The practitioner's job in MI is not to persuade. It is to ask questions that help the person surface and articulate their own reasons for changing. This is called eliciting change talk, and it is the mechanism by which MI produces behavior change.
The MI principle in one sentence: "The righting reflex": the coach's instinct to fix the problem, is exactly what makes coaching fail. MI is the practice of suppressing that instinct and asking instead.
The OARS framework
MI practitioners learn four core micro-skills, known by the acronym OARS:
O: Open questions
Open questions cannot be answered with yes or no. They invite the person to explore and elaborate. "What made today difficult?" is an open question. "Did you find today difficult?" is a closed one. The difference sounds small. In practice, it determines whether a coaching conversation goes anywhere.
The most powerful open questions in MI are about the person's own values and reasons, not the coach's agenda. "What matters most to you about making this change?" is more effective than "Have you considered why this matters?"
A: Affirm
Affirming is not cheerleading. It is acknowledging specific effort, strength, or insight. "You've got this!" is not an affirmation in the MI sense, it is noise. "You showed up three days this week even when it was hard, that tells me something about how much this matters to you" is an affirmation. It is specific, it is earned, and it references what the person actually did.
R: Reflect
Reflecting means mirroring back what you heard before adding anything. This is the most commonly skipped step, and the one that matters most. A reflection says: I heard you. I am not rushing to fix you. You can continue. Simple reflections repeat what was said. Complex reflections add meaning: "So it sounds like the problem isn't motivation, it's the design."
S: Summarize
Summaries are only used when the person has said enough to summarize. They pull together what has been shared and hand it back. "What I'm hearing is: you've tried three times, each time breaks at week three, and you think the problem might be what happens on weekends, does that sound right?" Summaries signal that the coach has been paying attention and invite the person to correct or build on what was heard.
What the research actually shows
MI is not a theory. It is one of the most studied behavioral interventions in existence. A 2019 Cochrane review of 48 randomized controlled trials found MI significantly outperformed brief advice for:
- Alcohol and substance use reduction
- Physical activity and exercise adherence
- Medication adherence in chronic disease
- Dietary behavior change
- Smoking cessation (in combination with other methods)
Source: Lindson et al., Cochrane Database of Systematic Reviews, 2019
A meta-analysis of 119 studies (Lundahl et al., 2010) found MI produced meaningful behavior change in 74% of studies. Effect sizes were largest when MI was delivered by trained practitioners, but the effect was also present when MI-style conversations were delivered by non-specialists trained in the method.
Source: Lundahl et al., Journal of Consulting and Clinical Psychology, 2010
This last finding is critical: you do not need to be a licensed therapist to use MI effectively. You need to use the method correctly. This is what makes it viable for an AI coaching application.
The four processes of MI
Beyond OARS, MI has a four-stage process that structures how coaching conversations develop:
1. Engaging
Building the relationship and establishing trust. In AI coaching, this means the first exchange should feel like being heard, not assessed. No advice, no evaluation. Just: I'm listening, tell me more.
2. Focusing
Agreeing on a specific direction for the conversation. "What would be most useful to talk about today?" This prevents coaching conversations from becoming vague rambles about everything.
3. Evoking
This is where the core MI work happens. The practitioner asks questions that help the person articulate their own reasons for change, their values, their hopes, what they have tried before, what worked. Change talk sounds like: "I want to," "I could," "I have before," "It would matter because." The coach's job is to notice and amplify it.
4. Planning
Only after change talk is strong does MI move to planning. Most coaching gets this wrong, it starts with the plan. MI ends with it. And the plan comes from the person, not the coach.
Why Avenn uses MI as its core method
Most AI tools, including general-purpose language models like ChatGPT, default to advice when asked about habits. "Here are five strategies to build a morning routine." This is the exact opposite of what MI prescribes, and it is why generic AI responses on habit topics often feel hollow or useless.
Avenn is built on the MI principle: reflect before you ask, ask before you advise, and never advise until you understand the obstacle.
Every response Avenn gives reflects what was said, then asks one specific question. It does not give a list of strategies. It does not tell you to "be consistent." It asks: "When was the last time this worked? What was different?"
That question, the exception question, is pure MI. And 40 years of research say it works better than anything else.
Try MI-based coaching free
Avenn uses Motivational Interviewing, BJ Fogg Tiny Habits, and Solution-Focused Brief Therapy in every coaching conversation. One full coaching session per week, free. No card.
Start coaching freeFrequently asked questions
What is Motivational Interviewing?
Motivational Interviewing (MI) is a collaborative, person-centered counseling style developed by William Miller and Stephen Rollnick in the 1980s. It is designed to strengthen a person's own motivation for and commitment to change by exploring and resolving ambivalence. It has over 40 years of randomized controlled trial evidence and is used by therapists, addiction counselors, and habit coaches worldwide.
What does OARS stand for in Motivational Interviewing?
OARS stands for Open questions, Affirm, Reflect, and Summarize. These are the four core micro-skills of MI. Open questions invite exploration. Affirming acknowledges specific effort. Reflecting mirrors what the person said before adding anything. Summarizing pulls together what has been shared when enough has been said.
How is Motivational Interviewing different from giving advice?
Advice tells. MI asks. The core principle is that people are more committed to change they argued for themselves than change someone else argued for. An MI practitioner never argues for change directly, they ask questions that help the person surface their own reasons. This elicits "change talk," which is significantly more effective than direct persuasion.
Can an AI use Motivational Interviewing effectively?
Yes. Multiple studies have validated AI-delivered MI-style conversations for smoking cessation, physical activity, and medication adherence. The key is training on annotated MI sessions and building in the OARS structure, specifically ensuring responses reflect before they question, and never advise before understanding the obstacle. Avenn is trained on MI practitioner sessions and applies OARS structure in every exchange.