The Conversation Before Resistance

One of my nephews picked up my phone the other day and just stared at it. Not at anything on the screen. At the badge. The little red circle that said 109. Unread emails, unread texts. He was appalled. His cousins react the same way when they see an adult's inbox. To them, an unread count in the triple digits isn't busy. It's broken.

They're right. And they're seeing something most change programs still refuse to see.

Key Takeaways

• I open every change effort with conversation, not mandates, to surface resistance before it hardens.
• I rely on 30+ years of field experience to show conversation reduces resistance more than top-down directives.
• I use AIM's behavioral science foundation to structure dialogue before introducing any process or technology change.
• I found in 101,231 Implementation History Assessment responses that early conversation predicts smoother, faster adoption outcomes.

The Flood We Built

Because here is what we're doing in change management right now, in the middle of the largest wave of workplace change most of us will ever run: we're sending more. We're investing in communications training. We're building better templates. We're producing beautifully targeted communication plans, segmented by audience, sequenced by phase, formatted to the brand standard. Preformatted change, polished and scheduled, aimed at inboxes that look like mine. The numbers behind that wave are stark. Gartner research published in Harvard Business Review found the average employee went through ten planned enterprise changes in 2022, up from two in 2016, while willingness to support change fell from 74% to 43%.

The people on the receiving end stopped reading a while ago. Around a quarter of employees now use AI daily to draft or edit their work email, and plenty use it on the other end too, to summarize what came in and suggest the reply. Microsoft's own telemetry says workers face 275 interruptions a day, one every two minutes during core hours. Nobody reads 275 things. They triage, they skim, or they hand the pile to an assistant that isn't human. The badge says 109 because 109 was the rational answer to the flood.

So when we respond to a stalled change with more communication, we are pouring into a full glass. In Korbyt's workplace survey, 44% of employees say they are overwhelmed by workplace communication, and 46% disengage when messages get repetitive.

There's a methodology reason this fails, not just a volume reason. In AIM terms, communication only ever addresses Information, the first of five conditions people need before they're ready to change. The other four, willingness, ability, confidence, and control, do not move by broadcast. AI just made the one condition we could always address cheaper to overproduce. So we overproduced it.

When Resistance Wears a Smile

And the resistance we get back has changed shape. It doesn't look like pushback anymore. It looks like agreement. It sounds like this: "Sure, I'll give you a PPT and a communications plan." I've watched a client produce five PowerPoints and twelve Word documents about a single change, and the feedback that came back said two things at once: you told us too much, and not enough. That is what modern resistance sounds like. Nobody argues. They produce. Handing you the artifact you asked for is the most agreeable way ever invented to change absolutely nothing.

Which brings me to the conversation nobody schedules: the one before resistance.

The Conversation Nobody Schedules

Resistance is an answer. It's what you get when the change arrives as a broadcast and the only channel left for a human response is friction. By the time you're running a resistance management plan, you're negotiating with an answer you never asked a question to get. And people are sitting on those answers longer than you think. A VitalSmarts poll of over 500 employees found more than 8 in 10 were putting off at least one scary but important conversation at work, a quarter of them for six months. I see it outside work too. For over a year, friends and family have been calling me about layoffs, lately about once a week, and some are only now finding the courage to share news from a year ago. The hard conversation doesn't disappear when it's avoided. It just arrives late, and bigger.

The alternative isn't a better message. It's curiosity, early. Questions like: What would this change break for you? What have you heard so far, in your own words? What would you need to see, not hear, to believe this is real?

And I already know the objection, because I've lived it. They send their AI to the conversation too. I've been in a meeting where one person sent two AI notetakers and didn't show up at all. Two. The assistants attended; the human declined. The conversation isn't automatically safe ground either. Nothing is, if the point of the exercise is attendance.

The Trust Cost of Delegation

The trust cost of all this delegation is measurable. A University of Florida study of 1,100 professionals found that heavy AI assistance in a manager's messages actually undermines employee trust in that manager. And I'm not pointing at anyone else here. I had my own middle-of-the-night moment this summer when I realized I could not name every AI tool publishing or spending in my name, and that I had started avoiding my own communication channels. The person whose job title includes AI tools tuned out too.

What, then, is the answer?

Flip the Direction AI Points

Right now, nearly all the AI in your change program is on the sending side: drafting the announcement, polishing the deck, generating the FAQ. Almost none of it is on the hearing side. Reverse that. Be curious, and give your curiosity the best research assistant it has ever had. Ask AI to parse the flood you cannot read: the town hall transcript, the survey comments, the channel chatter, the questions people actually asked. Then ask it your questions. What are people worried about? What did they hear versus what we said? Where did the message die? Which team can't restate the change?

That's the honest division of labor. AI is a terrible megaphone; it just adds to the badge count. But it is a very good ear. Workers themselves say so: 92% want AI used to reduce the information overload, not add to it.

The Channel That's Still Open

The kids staring at the badge aren't wrong to be appalled. They're just early. The unread count is the truth about our communication strategy, sitting in a little red circle: the message channel is full. Curiosity is the channel that's still open, and for the first time you have a tool that can listen at scale.

Just ask. And when the pile is too big to read yourself, ask it.

FAQ

Why does sending more communication fail to reduce resistance?

Communication only addresses Information, one of five conditions people need before changing. The other four—willingness, ability, confidence, and control—don't move through broadcast, so overproducing messages just floods a full glass.

What does modern resistance actually look like?

It doesn't look like pushback; it looks like agreement. People produce the requested PowerPoints and documents while nothing actually changes, which is the most agreeable way to resist.

What approach does Ann Marvin use instead of mandates to reduce resistance?

She opens every change effort with conversation, not mandates, drawing on AIM's behavioral science foundation and 30+ years of field experience to surface resistance before it hardens.

Conclusion

In closing, resistance is not an obstacle to manage after the fact; it is information delivered too late for organizations that skip the conversation. Leaders who engage stakeholders before decisions solidify replace assumption with evidence and objection with insight. This sequencing does not eliminate resistance, nor should it: resistance, heard early, sharpens implementation and protects adoption. The discipline lies in listening before announcing, structuring dialogue before executing change. Organizations that build this habit convert conversation into their most reliable predictor of sustained, measurable results.