The Approval Cliff · Problem

The Approval That Wasn’t

Why stakeholder sign-off isn’t the same as confidence. A signature records that nobody in one room objected. Confidence is whether the direction survives the next one.

Quick answer

Approval is a record of one room. Confidence is a prediction about the next one, the room where somebody who wasn’t at the meeting sees the work for the first time and asks why. A decision is confident when the person who signed off can defend it there, on their own, and the skeptic across the table concedes.

A signed approval proves one thing about a creative direction: nobody in that meeting stopped it. That’s worth having. It just isn’t what most teams think they’re holding when they walk out with it.

Most approval processes are built to produce the first thing and assume the second comes with it. It doesn’t, and the gap between them is where approved work goes on to fail.

What does stakeholder sign-off actually certify?

Stakeholder sign-off certifies that the direction cleared every objection people were willing to raise, in that meeting, with what they knew at the time. Each part of that sentence is a limit.

The tools around sign-off treat it as a closing step. Search “design sign off” and the first page is sign-off sheet templates from Workamajig and Smartsheet and a sign-off widget in the Figma community: forms that record that a decision happened. They’re useful for what they do. None of them asks what the signature rests on.

“Willing to raise” is the biggest one. A yes in a creative review is made of several different things, and only one of them is a judgment about the work. Some people agree because the most senior person in the room already did. Some agree because the meeting is in its second hour and the next one starts at three. Some have a reservation they can’t put into words, and in a room that rewards articulate objections, an unspoken one counts as consent. Some disagree and have decided the fight isn’t worth it on this project.

Every one of those people shows up on the sign-off sheet the same way.

Figure 1 · What a yes is made of

Five signatures that look identical on the sign-off sheet, and what each yes was actually made of: only one is a judgment about the work A sign-off sheet runs across the top with five slots, each labelled approved and each carrying a handwritten signature that looks like the others. Under the heading what each yes was made of, a dashed line drops from each signature to a card. The first card, in green, reads a judgment about the work. The other four, in red, read the most senior person had already said yes; second hour, the next meeting starts at three; a reservation they couldn't put into words; decided the fight wasn't worth it. A closing line reads only one of the five is a judgment about the work; all five show up on the sheet the same way. THE SIGN-OFF SHEET five yeses, identical on paper APPROVED a judgment about the work APPROVED the most senior person had already said yes APPROVED second hour; the next meeting starts at three APPROVED a reservation they couldn’t put into words APPROVED decided the fight wasn’t worth it WHAT EACH YES WAS MADE OF Only one of the five is a judgment about the work. All five show up on the sheet the same way.
The sheet records five yeses and treats them as five judgments. Four of them are something else, and nothing on the form can tell you which four.

“What they knew at the time” is the second limit. The room knew the brief, the deck, the designer’s rationale and each other’s opinions. It almost never knew how the audience the work was made for would read it, because nobody in the room was that audience.

We’ve called the cost of this gap the Approval Cliff: the distance between the moment a direction is approved and the moment the team finds out approval never meant alignment. This piece is about the moment just before the cliff, when the signature is fresh and it still feels like certainty.

Why does a room full of yeses feel like confidence?

A unanimous approval feels like evidence because everyone in the room assumes someone else checked.

The brand lead assumes the agency tested its thinking. The agency assumes the client knows its customers. The CEO assumes the people who’ve lived with the project for two months have pressure-tested it. Each person reads the others’ agreement as confirmation, so the agreement of the whole room feels much stronger than any one person’s judgment, even though nobody added any information.

The yes is also shared, and a shared yes has no owner. When the direction is questioned later, “we all agreed” is a true statement and a useless defense. It tells the skeptic how many people liked the work and nothing about whether it’s right.

And the feeling is pleasant. A meeting that ends in agreement is a relief after weeks of revisions, and relief is very easy to mistake for conviction.

What happens when the work reaches the second room?

Every approved direction gets carried into a second room, and the person who signed off usually goes in alone.

Sometimes it’s the CEO who skipped the review. Sometimes it’s a board, a sales team seeing the new identity at a kickoff, a regional office, a donor committee, a faculty senate. Eventually it’s the market itself. Whoever it is, they weren’t in the meeting, they didn’t hear the rationale, and they have the one thing the approval room lacked: a first impression with no investment in the outcome.

The question in the second room is always some version of “why this?” Most approvers walk in with one of four answers:

  • The committee liked it. That counts heads and says nothing about the work.
  • The designer’s rationale, and the skeptic is free to find it unpersuasive.
  • Their own taste, the exact thing the skeptic is questioning.
  • “It tested well internally,” when internally meant a meeting.

The rationale deserves a closer look, because it’s the strongest of the four and the one designers work hardest on. A design rationale is the documented why behind each choice, and a good one names the problem, the approach, the trade-offs and the evidence. In most creative reviews the evidence line is the one left thin. The reasoning can be excellent and still be the designer’s argument, which a skeptic can decline without ever engaging the work.

None of those four survives a determined skeptic. The approver either wins on seniority or loses on it, and either way the decision gets reopened, or it holds for reasons that have nothing to do with whether it’s right.

Figure 2 · Two rooms

The approval room, where everyone said yes with what the room knew, and the second room, where the approver carries the work in alone and the four usual answers fail Two cards. On the left, the approval room, what the signature records: six small figures in a row, each with a green yes above it, and under them what the room knew, the brief, the deck, the rationale, each other's opinions, with a dashed red chip reading not in the room: the audience. An arrow labelled carried in alone leads to the right card, the second room, what confidence has to survive: the approver faces the skeptic, who asks why this, and who wasn't at the meeting and has no investment in the outcome. Under them the four answers, each with a red cross: the committee liked it, the designer's rationale, my own taste, it tested well internally, and the line none of the four survives a determined skeptic. A closing line reads approval is a record of one room; confidence is a prediction about the next one. THE APPROVAL ROOM what the signature records yes yes yes yes yes yes WHAT THE ROOM KNEW the brief the deck the rationale each other’s opinions not in the room: the audience carried in alone THE SECOND ROOM what confidence has to survive why this? the approver the skeptic wasn’t at the meeting, no investment in the outcome THE FOUR ANSWERS the committee liked it the designer’s rationale my own taste it tested well internally none of the four survives a determined skeptic Approval is a record of one room. Confidence is a prediction about the next one.
The left card is what every approval process is built to produce. The right card is the one it gets judged in, and the only person from the left card who is there to answer for it walks in with four arguments and no evidence.

That’s the test worth running before launch. Ask whether the decision makes the approver’s position more defensible in a room where it will be challenged. If the only support is the approval itself, the answer is no, however many signatures are on the sheet.

How can you tell you have approval without confidence?

You can usually tell before the second room arrives. These are the tells.

Ask three people who approved the direction why it’s right, separately, and you get three different answers. Agreement on the outcome with no agreement on the reason means each person approved a slightly different thing.

The reasons people give are about craft. “It’s clean,” “it feels modern,” “the type is beautiful.” All of that can be true of work the audience reads completely wrong.

The most senior person’s comment became the rationale. If the deck for the next meeting quotes the CEO’s reaction as the case for the direction, the room deferred, and the deferral is now the evidence.

Somebody says “let’s see how it lands.” That sentence is a plan to find out after launch.

And the one that matters most: nobody in the room can say what the audience will see. Ask what a customer will take from the work in the first few seconds and the room describes what it intended. Those are two different answers, and only one of them gets tested at launch.

If you’ve walked out of an approval meeting with a signed-off direction and an uneasy feeling, that feeling was doing its job. It’s the part of you that already knows the signature didn’t settle anything, and it deserves a better answer than another round of opinions.

How do you defend a design decision to someone who wasn’t in the room?

You defend it with evidence the skeptic can look at, that came from the audience the work is for, and that existed before anyone’s opinion was on the line. A decision with all three carries into the second room. A decision missing them is only as strong as the person presenting it.

Opinions fail that standard on every count, however senior the person holding them. Vague feedback fails it too: the phrases stakeholders use are honest reactions, but they’re translations of something seen into words, and every translation loses some of it.

Audience perception data passes. A team tests a range of ideas and elements with the audience the work is for, the direction emerges from where responses cluster, and the result is a map anyone can read. That is audience perception mapping: an audience responds to images instead of describing them, and the pattern shows where the group agrees, where it splits, and why. About 30 responses will show a pattern. For a brand decision, 300 to 500 is the range. Results come back in under 48 hours once the test is live.

Figure 3 · What carries into the second room

The three-part standard for evidence that survives a skeptic, and which kinds of support meet it: opinions and vague feedback fail, audience perception data passes A grid. Across the top, the three parts of the standard: the skeptic can look at it; it came from the audience the work is for; it existed before anyone's opinion was on the line. Down the side, three kinds of support. Opinions, however senior, the exact thing the skeptic is questioning: three red crosses. Vague feedback, a reaction translated into words, with loss: three red crosses. Audience perception data, a map anyone can read, in a green row: three green checks. Under the grid, three pills: about 30 responses show a pattern; 300 to 500 for a brand decision; results in under 48 hours. A closing line reads the skeptic can still disagree; now they're disagreeing with the audience, in front of the evidence. THE STANDARD FOR EVIDENCE three parts, all required the skeptic can look at it it came from the audience the work is for it existed before anyone’s opinion was on the line Opinions, however senior the exact thing the skeptic is questioning Vague feedback a reaction translated into words, with loss Audience perception data a map anyone can read about 30 responses show a pattern 300 to 500 for a brand decision results in under 48 hours The skeptic can still disagree. Now they’re disagreeing with the audience, in front of the evidence.
The standard is strict on purpose. Anything the approver could have produced alone, after the fact, is an argument, and a skeptic can decline an argument. The bottom row is the only one the skeptic has to answer with evidence of their own.

That changes what the approver carries into the second room. “We liked it” becomes “this is what our customers responded to, this is what they rejected, and this is the quality the direction is built on.” The skeptic can still disagree, but now they’re disagreeing with the audience, in front of the evidence, and that’s a much harder position to hold. A skeptic who won’t concede needs an argument of their own that’s stronger than the map.

That last part is the accountability side of perception data. Once the audience has spoken, anyone who wants to change course has to bring evidence, and seniority stops being enough. Reading the map takes about five minutes, and it gives every later disagreement something to point at.

The approval room’s question changes too. “What do you think?” invites the performance every review turns into. “What do we do with what the audience saw?” is a question the room can answer together, and when it does, everyone has agreed, and they know why, because they can see why.

How do you change the approval process?

The creative team’s judgment stays, and so does the stakeholders’ sign-off. What changes is what the sign-off rests on.

Test before the meeting, while it’s still cheap. The test belongs ahead of direction, layered on top of the research the team already does. Then the approval meeting opens with the audience’s answer, and the agency’s proposal is built on it. Testing brand direction before launch walks through the sequence.

Write the reason into the sign-off. Next to the signature, one sentence: what the audience responded to that this direction is built on. If nobody can write that sentence, you have approval, and you know it before launch.

Send the evidence ahead of the decision, to the people who weren’t in the room. The CEO who skipped the review should see the map before they see the work. Their second-room question gets answered before they ask it.

Keep everyone in the process. Shrinking the room to get a cleaner yes buys an approval and loses the people who’ll have to defend it later. Design by committee covers why the format of the input matters more than the headcount.

Where to start

Pick the direction your team approved most recently and run the second-room test on it. Imagine the most skeptical person who wasn’t in that meeting asking why. If your answer is the approval itself, you know what you have.

When you’re ready to give the next approval something to stand on, we run done-for-you perception studies from a scoping call: one that puts the question to your leadership team, one for your whole organization, and one for the audiences outside your walls. The point of all three is the same. The person who signs off walks into the next room with something to show.

Frequently asked questions

What does stakeholder sign-off actually certify?

That the direction cleared every objection people were willing to raise, in that meeting, with what they knew at the time. A yes in a creative review can be deference, fatigue, an unspoken reservation or a decision not to fight, and every one of those shows up on the sign-off sheet the same way as a judgment about the work.

What is the second-room test?

Imagine the most skeptical person who wasn’t in the approval meeting seeing the work for the first time and asking why. A decision is confident when the person who signed off can defend it there, alone, and the skeptic concedes. If the only support is the approval itself, you have approval without confidence.

How do you defend a design decision to a stakeholder who wasn’t in the room?

With evidence the skeptic can look at, that came from the audience the work is for, and that existed before anyone’s opinion was on the line. Opinions and vague feedback fail all three. Audience perception data passes: a map of what the audience responded to and what it rejected, which the skeptic can only answer with evidence of their own.

How many people do you need to test a creative direction before approving it?

About 30 responses will show a pattern. For a brand decision, 300 to 500 is the right range, and results come back in under 48 hours once the test is live, which leaves time to open the approval meeting with the audience’s answer instead of waiting to see how the work lands after launch.