Having the Conversation You Keep Avoiding

The conversation has been drafted in your head a dozen times, each version ending differently depending on your mood that day. Meanwhile the actual conversation hasn't happened, the situation hasn't changed, and every day it stays unaddressed makes the eventual conversation slightly harder to start.

It could be ending a business partnership, confronting a friend about a pattern that's wearing thin, or telling a family member something they won't want to hear. The specifics vary; the shape of the avoidance doesn't.

Why does generic advice about hard conversations fail?

"Just rip the band-aid off" treats the difficulty as purely a matter of nerve, as if the only obstacle were working up the courage to start talking. It ignores that most dreaded conversations fail not at the opening line but partway through, when the other person reacts in a way the rehearsed opening didn't account for and the conversation derails into defensiveness or shutdown.

Generic AI prompting tends to produce exactly this kind of one-sided preparation. Ask an assistant to "help me tell my business partner I want to dissolve the partnership" and it will often draft a clean, well-reasoned opening statement without any structure for what happens next, when the partner reacts with anger, denial, or silence, none of which a single prepared paragraph can navigate.

The deeper issue is that a dreaded conversation is rarely a monologue delivered and received; it's an exchange where the other person's reaction shapes what needs to happen next. Preparing only your own opening line solves a fraction of the actual problem.

The avoidance itself also compounds in a specific way: the longer a hard conversation is put off, the more evidence accumulates that something was being withheld, which tends to make the eventual conversation harder, not easier. Waiting rarely improves the conditions for a conversation that was already going to be difficult.

What does the documented negotiation-of-hard-moments record say to do?

Chris Voss: use calibrated questions to keep the other person engaged rather than defensive

Voss's tactical empathy, documented in Never Split the Difference (2016), uses open "how" and "what" questions to keep the other person talking and thinking rather than shutting down or attacking. A calibrated question invites problem-solving; a direct statement often invites defense.

Applied: Instead of opening with "we need to end this partnership," try "what would it look like if we each pursued this differently going forward?" The question opens the actual topic without forcing an immediate defensive reaction to a blunt statement, and the other person's answer often reveals how much common ground already exists.

This matters most in the opening thirty seconds, which set the emotional tone for everything that follows. A question that invites the other person into the problem, rather than announcing a conclusion they haven't been part of reaching, changes the entire shape of what comes next. Voss's own hostage-negotiation background underlines the point: when the stakes are genuinely severe, the first exchange determines whether the rest of the conversation is even possible.

Carl Rogers: hold the person's worth and the hard truth at the same time

Rogers' unconditional positive regard, documented across his published therapeutic work, holds that people engage most honestly when they don't have to defend their basic worth in the conversation. The hard truth still gets said; it just doesn't come bundled with an implicit judgment of the person's character.

Applied: Separate the specific issue from any claim about who the person fundamentally is: "this specific pattern has to change" rather than "you're not someone I can rely on." The first invites a conversation about the pattern. The second invites a defense of identity, which derails the actual issue almost immediately.

Rogers' point is that acceptance and honesty aren't actually in tension, despite how often they're treated as a tradeoff. A person who feels their basic worth isn't on trial in the conversation is far more able to actually hear and consider a hard truth than one who's busy defending their character against an implied verdict.

Virginia Satir: name the stress stance instead of fighting against it

Satir's family systems work documents four common stances people default to under conversational stress: placating, blaming, being overly logical, or distracting. Naming the stance gently, without judgment, often does more to shift a stuck conversation than continuing to argue the original point.

Applied: If the other person goes silent or shuts down, rather than repeating your point more forcefully, try naming what's happening: "it feels like this is hard to stay in right now, is that fair?" This often reopens a conversation that direct persistence would have kept stuck.

Satir's broader insight is that the stance itself is information about what the person needs in that moment, not an obstacle to push through. Someone placating needs reassurance that honesty won't cost the relationship; someone blaming often needs to feel heard before they can stop defending. Naming the stance is the first step toward addressing what's actually underneath it.

How do the three approaches compare?

What can you run before the next conversation?

Why it works: a question-first opening keeps the other person in a problem-solving posture rather than a defensive one from the very first line.

Why it works: identity-level language is often unconscious, and a dedicated check catches it before it derails the conversation into a defense of character.

Why it works: preparing for the other side's likely reactions, not just your own opening line, is what actually determines whether the conversation survives the first pushback.

What won't this fix on its own?

These approaches improve how a hard conversation opens and how it recovers when it stalls. None of them can guarantee the other person responds well, some conversations end badly regardless of how carefully they're approached, because the other person isn't ready, doesn't agree, or is dealing with something entirely outside the conversation itself.

A well-run conversation improves the odds of a good outcome considerably; it doesn't manufacture agreement that genuinely isn't there to be found. Knowing the difference between a conversation that went badly because it was handled poorly and one that went badly because the underlying disagreement was real and irreconcilable is itself a judgment call no framework makes for you, and it usually only becomes clear with some distance from the moment itself.

Chris Voss sits in the Negotiator category, and Carl Rogers and Virginia Satir both sit in the Psychotherapist category. For structuring the specific hard feedback conversation at the center of this, Difficult Feedback is a $49 tool built for exactly that.

All the copy-paste prompts from this guide, and the rest of the Situations series, live at /prompts too, free to copy without reading the guide first.