Behavioural Change Toolkit · Advising Senior Leaders

Challenging Upward — Scripts

Saying the hard thing to someone senior, and still being in the room afterwards.

Five rules

1. Privately first, always. Public challenge forces a defensive position they then have to hold.
2. Consequence, not character. “This will cost us X” lands. “You’re not visible enough” does not.
3. Bring the option. Never arrive with only a problem. Two options with a recommendation is the ideal shape.
4. Say the cost of inaction, once, plainly. Then stop. Repeating it becomes nagging and you lose the room.
5. Let it be their decision. Your job is to make sure they’re deciding with full information, not to win.

The universal structure

1. Permission     "Can I raise something that's going to be uncomfortable?"
2. Observation    "Here's what I'm seeing…"          ← neutral, factual
3. Consequence    "Which means…"                     ← specific, not dramatic
4. Options        "Two ways I can see…"
5. Recommendation "I'd suggest…"
6. Handover       "But it's your call — what would you like to do?"

The permission step is not politeness. It changes how the next sentence is heard, and it gives them a moment to prepare, which materially reduces defensiveness.

Scripts

The sponsor isn’t visible enough

“Can I raise something? The thing that most predicts whether this lands is how visibly you’re behind it — more than the plan, more than the comms. At the moment, the teams I’ve spoken to can’t name anything they’ve seen you do on this.

That’s not a criticism, it’s a gap I can help close. Two options: you take two minutes at the start of each monthly ops call using the dashboard, or you spend an hour a month on site with the teams doing it. The second is more powerful and costs more of your time.

I’d suggest the first as a minimum, starting this month. Which works?”

The design is genuinely wrong

“I need to tell you something I don’t think you’ll want to hear. Six of the eight people I spoke to independently hit the same problem — [specific]. I went and did it myself and hit it too.

If we launch as designed, my honest view is we’ll get partial adoption and a lot of workarounds, and we’ll be unpicking it in six months.

Two options: we take three weeks and fix it, which moves the date. Or we launch as-is and accept we’ll be fixing it live, with the credibility cost that carries. I’d recommend the three weeks. What’s your view?”

The date is unrealistic

“I want to be straight with you about the date rather than tell you what you’d like to hear now and disappoint you in October.

On the current plan we can deliver the system on time. What we can’t do is have people actually using it properly — that’s [N] weeks of behaviour work that hasn’t started.

So the choice is: hit the date and accept low adoption for a while; move the date; or reduce the scope to the part we can land properly. All three are legitimate. What I’d like to avoid is us hitting the date and calling it done.”

⭐ “Delivered but not adopted” is the framing that gets senior attention. It names a real and recognisable outcome without requiring anyone to admit the plan was wrong.

There’s too much change

“Before we commit to this, can I show you something? Here are the seven initiatives already live for these teams this quarter.

I’m not saying we shouldn’t do this. I’m saying they can’t all land, and at the moment we haven’t chosen which ones matter most — so the teams are choosing for us, informally, and we won’t like their choices.

Could we take twenty minutes to rank these? I’ll bring the list.”

Their own behaviour is the blocker

The hardest one. Private, specific, one observation only.

“Can I mention one thing? It’s small and I think it’s costing us more than it looks.

The regional report still comes to you in the old format. I know it’s easier to read and I understand why. But it means every region keeps the old spreadsheet running, which means the new system never becomes the only source of truth — and people have noticed.

If you asked for it in the new format from next month, I think it’d do more than anything else we’re planning. Would you be willing?”

Note the shape: one specific behaviour, why it happens (charitably), the consequence, one clear ask. No general commentary on their leadership.

It’s not working and you have to say so

“I want to give you a straight read rather than a managed one. Adoption is at 31% and it’s been flat for a month. It’s not going to embed on its own.

Three options. Re-invest for eight weeks — I’d need [specific asks]. Reduce the scope to the part that has stuck and formally drop the rest. Or stop, and be clear we’ve stopped.

What I’d advise against is leaving it as-is, because a half-adopted process is worse than either version — we now have two sources of truth and people are choosing between them.

I’d recommend [X]. What do you want to do?”

They’re punishing the messenger

“Can I raise something awkward? When the low number came up last month, the response in the room was pretty sharp — understandably.

What I’ve noticed since is that reporting has got more optimistic than I think is accurate. I don’t think that’s anyone being dishonest, I think it’s people reading the room.

The risk is that you get the real picture late, when it’s expensive. Would you be willing to say something at the next one about wanting the real numbers? It’d change the temperature quickly.”

When they don’t take the advice

You’ve done your job. Now:

1. Record it, briefly and neutrally. A dated line in your notes: “Recommended three-week fix, decision was to proceed to date.” Not for blame — for your own clarity, and because in four months everyone’s memory will have reorganised itself.
2. Commit properly to the decision made. Visible half-heartedness after losing an argument costs you far more than the argument did.
3. Say it once more only if new evidence appears. New evidence is a legitimate reason to reopen; repetition isn’t.
4. Notice the pattern. If advice is consistently ignored, that’s information about the role you’ve actually been given — worth an honest conversation about what they want from you.

What buys you the right to do this

• Being right before, and being seen to be right
• Never bringing a problem without options
• Being visibly loyal to decisions you argued against
• Being the one who says “I was wrong about that” first
• Making them look good with their own peers

Credibility is a budget. Spend it on the two or three things that genuinely matter, not on every point of disagreement.

→ Executive Failure Modes · Sponsor Briefing Pack · The One-Page Board Narrative

← Back to the Behavioural Change Toolkit