Costs
Part of Control and influence: the questions that matter
Control and influence mistakes when a request stalls
Control and influence mistakes when a request stalls: reframe before pushing, test the no, check the record, add no control by reflex, and escalate last.
A stalled request invites the same error every time: treating influence as a weaker grade of control and applying it harder. It is a different instrument. It changes the odds that someone decides differently, and it runs out. What follows is the order in which to repair the mistakes, once you notice you are in one.
What to take away
- The productive question is never how hard to push. It is what would have to be true for this person to say yes, and whether you can make it true.
- A record of having raised something is not a result. If the warning changed no decision, it either never reached a decider or arrived as information rather than as a choice.
- Most declared walls have never been tested. One question to the person who owns the rule is cheaper than months of working around a door.
The repair order
- Reframe instead of pushingask what would have to be true for a yes.
- Test the no before you accept itfind out who owns the rule.
- Check the recorda warning that changed no decision is not a result.
- Before you add a control, ask whether it would have caught the last failure.
- Escalate last, and say you are doing it before you do it.
Reframe before you push
Mistake 1 is pushing harder on something that only responds to reframing. When influence is slow, the instinct is more contact and more evidence. If the objection is about risk to them, none of that touches it.
Force vs Reframe
Pushing harder
- Trigger
- Influence feels slow
- Instinct
- More contact, more evidence
- Result
- Objection untouched
Reframing
- Trigger
- Objection is about risk
- Instinct
- Ask what must be true
- Result
- Route to a yes
The productive question is never "how hard should I push" but "what would have to be true for them to say yes", and then whether you can make it true. That is the logic of interest-based bargaining, which federal mediators describe as resting agreements on mutual and individual interests rather than positions.
Test the no
Mistake 8 is reading a no as a fact about the answer.
A no is a fact about a specific request, from a specific person, at a specific moment, framed a specific way. Change the framing, the timing, or who asks, and the answer changes often enough to be worth testing.
The follow-up is not "are you sure" but "what would have to change for this to be a yes". It either gets you a route or gets you an honest never, and both are more useful than a closed conversation.
Mistake 9 is announcing where the line is instead of finding out. People declare things uncontrollable on the basis of an assumption nobody tested: that the rule has no exception, that the budget is fixed, that the decision is made. Some of those are true. Checking costs one question to one person.
FMCS describes its Intentional Negotiations approach as helping parties move beyond positions toward shared understanding, practical outcomes and durable agreements. The follow-up question does the same job in one sentence.
Check the record
Mistake 3 is confusing "I raised it" with "I changed it". A record of having warned somebody is not a result, whatever it does for you afterwards.
If a warning did not change a decision, one of two things happened: it never reached anyone who could decide, or it arrived as information rather than as a choice with consequences attached. Both are fixable. Neither is fixed by having sent the email.
Before you add a control
Mistakes 4 to 6 share a pattern. When influence stalls, people reach for something they can control instead, and the substitute feels like progress. Three tests separate real control from its substitutes.
Before adding another control
- Is the outcome mine to own or only to nudge?
- Is this the causal variable or the visible one?
- Would it have caught the last failure?
- Was this going to happen anyway?
- Am I spending tolerance I will want later?
- Owning outcomes you can only nudge.Carrying full responsibility for something you influence is exhausting, and it makes you worse at the job. Once the outcome is your identity you argue for your plan because it is yours, and you stop noticing the version that would work better with someone else's name on it.
- Controlling the visible variable instead of the causal one.Meeting cadence, status formats and checklists are all controllable, all satisfying, and frequently unrelated to why the thing is late. Process is what you reach for when the real variable belongs to someone else. Ask whether the new control would have caught the last failure.
- Spending control on what was going to happen anyway.Approving something already in motion, or mandating what people were already doing, looks like leadership and produces nothing. It also uses up the tolerance people have for being directed, which you will want later.
Escalate last
Mistake 2 is escalating the first time influence is slow. Escalation works.
It also spends something: the person you went around now knows you will do it again, and adjusts by telling you less and agreeing to less in advance.
Keep it for the case where the answer genuinely matters more than the relationship, and say you are doing it before you do it. Told first, in plain words, it costs far less than when it arrives as a surprise.
The alone test
Pick a problem you are stuck on and write down what you would have to do, entirely alone, to fix it.
If the list has items, you are in control and have not used it. If the list is empty, you are in influence, and the work is finding the person who decides and understanding what saying yes would cost them.
Both answers are useful. What is not useful is the middle position where you act as though you have authority, get treated as though you have none, and describe the gap as a communication problem. The sorting exercise itself is in control and influence, and the honest-instrument argument is in control or influence in leadership.
Common questions
How many times should I reframe before accepting that the answer is no?
Twice is usually the honest limit. The first attempt tells you the stated objection, the second tests whether it was the real one. If it changes again, the cost to them is real and no wording removes it. Take the trade to whoever owns both sides.
Is escalation always expensive?
No, and treating it as forbidden is its own mistake. It costs when it arrives as a surprise. Told first, in plain words, it usually costs very little, and occasionally the sentence alone resolves the thing, because it makes continued deferral visible without threatening anyone.
I keep being told to be more proactive, which sounds like adding process.
That is mistake five almost exactly. Before adding a control, ask whether it would have caught the last failure you actually had. If it would not, it is activity in the controllable column standing in for work in the influence column. Control and influence examples shows the sorting.
How do I ask "what would have to change" without sounding like I am refusing to hear no?
Ask it once, and accept the answer including when the answer is nothing. The question is a request for a route, not a second attempt at the same request, and it stops working the moment it is used as one.
