And the fastest way to fix it isn't what you think.

We've all been there. A client sits across from you, ambivalent about change, and something in your gut says just explain why this matters. You do it warmly. You do it with empathy. You might even throw in a reflection first. But here's the uncomfortable truth backed by decades of fidelity research: if you're the one arguing for change, you're not doing MI. You're doing persuasion with a softer voice.

And that's not a moral failing — it's a predictable drift that happens to nearly everyone.

The Righting Reflex Is Running the Show

The reason most of us slide into "polite persuasion" has a name: the righting reflex. It's that automatic helper impulse to fix, correct, warn, or prescribe the moment we hear risk or suffering. Under time pressure — which is to say, always — that reflex shows up as early advice, "educational" corrections, or subtly steering the conversation toward what we think is best.

Here's why this matters specifically for MI: the approach is defined as directional, yes, but directional through evocation. The Motivational Interviewing Network of Trainers is explicit that MI is not a method for "getting people to change." Unsolicited advice, confronting, instructing, and warning are all flagged as incompatible with the model.

MI asks you to suppress the urge to explain long enough for the client to build — and hear themselves build — their own case for change. That's a fundamentally different clinical move than being warm while you recommend.

The Fastest Fidelity Win: Stop Doing the Wrong Things

If you want to get better at MI quickly, the highest-yield move isn't adding more OARS to your repertoire. It's deleting MI-inconsistent behaviors — the reflexive moves that reliably push clients into defending the status quo, even when you deliver them kindly.

In the MITI 4.2.1 coding system (the gold standard for MI fidelity), MI-inconsistent behavior boils down to two categories: Confront and Persuade. "Polite persuasion" — the thing most of us default to — still lands squarely in the Persuade code. Sometimes it crosses into Confront. The warmth doesn't change the classification.

The data here is striking. A 2024 implementation study found that coached providers reported significantly fewer MI-inconsistent behaviors than non-coached providers (means of 1.21 vs. 3.92). Even more telling, provider-reported MI-inconsistent behaviors predicted worse observer-rated MI fidelity down the line, even after controlling for post-training scores. In other words, what you fail to subtract matters more than what you try to add.

A Three-Minute Self-Audit

Pick your top two reflexes and replace each with one sentence that supports autonomy and invites evocation before education:

Reflex: "Tell them the plan" (coded as Persuade) → "Would it be okay if I share a couple of options — and then hear what fits for you?"

Reflex: "Correct their thinking" (often coded as Confront) → "You've got good reasons to keep things as they are — and also some reasons you're considering change. What feels most important right now?"

That's it. Two substitutions. Try them for a week and notice what shifts.

The Real Active Ingredient Isn't Warmth — It's What Happens to Client Language

Here's where the mechanism research gets fascinating. Across multiple decades of MI process studies, the dominant finding is that clinician MI skills influence client language about change, and client language predicts outcomes.

A large 2017 meta-analysis of MI process models found that MI-inconsistent clinician behaviors were associated with more sustain talk from clients — but not with more change talk. When sustain talk and change talk were examined separately, sustain talk predicted worse outcomes while change talk alone wasn't a significant predictor. The ratio between the two — the balance of pro-change versus anti-change language — was what mattered.

William R. Miller himself, writing in 2023, summarized it this way: sustain talk tends to be a better predictor of outcome than change talk alone, and the change-talk-to-sustain-talk ratio is related to the probability of subsequent change.

Translation for your next session: Warmth is necessary but not sufficient. The measurable signal is what's happening to the client's own language about change. Are they talking themselves into it, or are you doing that work for them?

The Decisional Balance Trap

If you regularly use a classic "pros and cons of change" exercise, you might want to reconsider. This isn't just opinion — MI's own research literature supports the concern.

A review on decisional balance versus MI evocation concludes that with ambivalent people, decisional balance tends to decrease commitment to change, while evocation promotes it. The logic is intuitive once you see it: if you equally evoke and reinforce both sides of ambivalence, the expected result is… continued ambivalence.

A more MI-consistent alternative: Keep the acknowledgment of both sides, but shift where you invite elaboration. Reflect sustain talk accurately without dwelling on it. Then preferentially invite the client to expand on their own reasons for change and the costs of the status quo. If you must do a pros-and-cons activity, structure it asymmetrically — explore the cons of staying the same and the pros of change — so you're not giving sustain talk equal airtime.

The question isn't whether ambivalence is normal. It is. The question is whether your strategy resolves it or accidentally stabilizes it.

Brief MI Is Real — But "Brief" Means Session Length, Not Learning Curve

For those of you working in primary care, prescribing, or time-limited settings: brief MI works. A 2025 study found that a 4-hour interactive MI workshop plus brief guides led to significantly improved patient biomedical outcomes compared to controls, including meaningful reductions in HbA1c and diastolic blood pressure.

And for those integrating MI with other evidence-based treatments, the data is encouraging. A pilot RCT combining MI and brief prolonged exposure for veterans with co-occurring PTSD and harmful drinking achieved 70% treatment completion and substantially larger reductions in PTSD severity compared to usual care.

But here's the honest caveat: training produces real post-training gains, but skills tend to erode over about six months when feedback and coaching are absent. A modest dose of post-workshop coaching — often described as three to four sessions over six months — can sustain what training builds.

Brief MI is possible. Brief MI mastery is not. That's not discouraging. It's just realistic about what skill development looks like.

What You Can Do This Week

MI competence isn't about "sounding warm." It's about making different choices in the moments where persuasion is most tempting. Here's your starting point:

Record one session this week (with appropriate consent). Listen back for just ten minutes. Count how many times you explain, recommend, or correct — versus how many times you ask the client to voice their own reasons for change. Don't judge the ratio. Just notice it.

That single act of measurement puts you ahead of the drift curve. Because the gap between trained clinicians and competent MI practitioners isn't about bad intentions — it's about insufficient feedback. And feedback starts with looking.

Have a colleague who swears they "already do MI"? Forward this their way — with love. And if you want to keep sharpening your clinical edge, we'll be back next week with more research you can actually use.

Note: Clinical examples in this post are illustrative and fictional. As always, consult supervision and follow local regulations when implementing new approaches. MI fidelity measurement should complement, not replace, individualized clinical judgment.