The Gate

Two ways to die

Imagine two people walk into the same review meeting on the same morning.

The first has a genuinely new idea, a way of doing something the company has never done, using a method nobody in the room recognises. Call her the radical.

The second has a small improvement to something that already exists, a better way to run a process the company runs every day, that saves a bit of time and money and steps on nobody’s strategy. Call her the incrementalist. This is Maria.

Both of them will very likely walk out with a no. And here is the thing that took me an embarrassingly long time to understand, and that most of the writing on corporate innovation still gets wrong: they will get their no for completely different reasons. The radical and the incrementalist are killed by two different mechanisms, and confusing them is why so much well-meaning advice about fostering innovation misses.

Almost everything written about innovation being stifled is about the radical, and about the bias against the genuinely new. That bias is real, and we will spend the first half of this chapter on it, because it is well documented and it matters.

But it has nothing to do with why Maria’s idea dies. Maria’s idea is not novel, and that is the whole point of it, because it is a sensible, legible, low-risk improvement to a known process. The anti-novelty machinery does not even engage. Something else kills her idea, something the innovation literature almost never names, and it is the more important half of this chapter because it is the half that explains the fleet.

Let us take them in order.

The first death: the bias against the new

Start with the radical, because hers is the famous story.

There is a comfortable belief, especially among leaders, that organisations want creativity — that the new idea’s problem is a shortage of it, and that if people would just be braver and think bigger, innovation would follow. It is a nice belief and it is almost the reverse of what the evidence shows.

In 2012 Jennifer Mueller and colleagues ran a study with a title that gives the game away, “The Bias Against Creativity.” What they found is that people hold an implicit bias against creative ideas, a genuine and measurable aversion, that surfaces precisely when they are feeling uncertain. And a decision meeting evaluating a novel idea is, by definition, a room full of uncertainty. The more genuinely new the idea, the more uncertainty it generates, and the more the bias fires. So the ideas that are most creative are the ones most reliably rejected, by the same people who will tell you sincerely, over coffee afterwards, that they wish their teams were more creative. They are not lying. They simply cannot feel the bias operating, because it runs underneath the reasons they give themselves.

The finding has held up and sharpened since. Kevin Boudreau, Eva Guinan, Karim Lakhani and Christoph Riedl ran a grant proposal process at a research university and randomised which evaluators saw which proposals, generating over two thousand evaluator–proposal pairs. Two things came back, and both are quietly brutal. Evaluators marked down proposals that were highly novel. They also marked down proposals close to their own expertise. Expertise did not correct for the novelty penalty; it carried a penalty of its own. Novelty was not a tiebreaker in the idea’s favour. It was a penalty.

And Paola Criscuolo, Linus Dahlander, Thorsten Grohsjean and Ammon Salter — Criscuolo and Salter again, from the bootlegging study in Chapter 1, this time with two new collaborators — took the question inside the corporation, studying how the panels at a large professional services firm actually chose which R&D projects to back.

What they found was not a simple prejudice against the new, and the shape of it matters. Some novelty helped: a proposal that departed moderately from what the firm already did received a larger share of the funds it asked for than a wholly conventional one. Past a point the curve turns over, and too much novelty was selected against, because it was difficult to appreciate. The ideas that lost out were not the mediocre ones. They were the ones furthest from what the panel already knew how to think about.

And where the curve turns depends on the panel rather than on the proposal. A panel carrying a heavy workload became less willing to back novelty. A panel with a spread of expertise became more willing. The same proposal, in other words, is worth different amounts depending on who is in the room and how much else they have to get through that day.

So the gate is not a neutral filter that occasionally lets a good radical idea through. It has a thumb on the scale, the thumb presses hardest on exactly the ideas that depart most from what is already understood, and how hard it presses is a property of the room rather than of the idea.

Why? Not stupidity and not cowardice, which is the usual disappointing answer. A genuinely novel idea is by definition the one with the least evidence behind it, because nobody has done it before. It is the hardest to justify, the hardest to forecast, the easiest to attack, and the one most likely to make the person who championed it look foolish if it fails. An evaluator behaving with perfect rationality about their own risk will mark it down. The bias is not a malfunction of the evaluation process. It is the evaluation process, working as designed, applied to something the design cannot price.

That is the radical’s death, and it is worth understanding. But it is not Maria’s death, and if you only understand this one you will misdiagnose the fleet completely. You will conclude that the problem is fear of the new and you will prescribe more courage, more vision, more tolerance for moonshots, none of which would have saved Maria’s supplier-deviation tool. Her tool is not a moonshot. It is the opposite of one. It is a small, ordinary, sensible improvement to a process the plant already runs every day, the kind of idea nobody would call brave or visionary or creative. And that ordinariness, which ought to be exactly what makes it safe to surface, turns out to be irrelevant to whether it survives, because a different mechanism is waiting for it.

The second death: territory

Here is the mechanism nobody talks about, and once you see it you cannot unsee it in any large organisation you ever walk into again.

Maria’s idea is small, sensible and low-risk. It threatens no strategy and generates almost no uncertainty. By every measure in the anti-novelty literature it should sail through.

And it dies anyway. Ask yourself why, honestly, and the answer is not about the idea at all. It is about whose idea it is, and whose patch it lands on.

Because Maria’s tool reviews supplier deviations, and there is a team, call it Quality Systems, whose job is supplier deviations. That is their remit. Their budget line. The thing they are staffed to own. And Maria, a quality engineer three rungs down and one department over, has just built something that works inside their territory, without asking, and it is better than what they have.

Watch what happens when she surfaces it. The Quality Systems lead does not say this is a bad idea. He can’t, because it is plainly a good one, and saying otherwise would expose him. What he says is subtler and you have heard it a hundred times. Interesting, though this really sits within our roadmap. Or: We looked at this, actually, and it’s more complex than it appears at the plant level. Or the purest form, the one that ends things without a single word of disagreement: Great initiative. Let’s make sure it’s aligned with the programme.

None of those is a rejection of the idea. Every one of them is a defence of the border.

What the research calls this

For a long time this had no proper name in the management literature, which is part of why it goes undiagnosed. It was folk knowledge. Everyone who has worked in a big company knows about turf, but folk knowledge is easy to wave away as office politics, a personality problem, a few insecure managers rather than a structural force.

Then, in 2005, Graham Brown, Thomas Lawrence and Sandra Robinson published a paper in the Academy of Management Review that gave it a spine. They called it territoriality, and their central move was to show that people in organisations become territorial not only over physical space, the office and the parking spot, but over ideas, roles and responsibilities. The intangible territory. The remit. And they mapped the behaviours that defend it: marking, where you signal that something is yours, and defending, which comes in two flavours, anticipatory, heading off an incursion before it happens, and reactionary, repelling one that already has.

Maria’s surfaced tool is an infringement in the exact technical sense, and what she triggers in the Quality Systems lead is reactionary defending, a response that the research shows has essentially nothing to do with the merit of what infringed. Brown and Robinson followed this up in 2011 in Organization Science, looking directly at how people react to territorial infringement, and the reaction is what you would expect from the word. The interloper is repelled and the border reasserted, regardless of whether the interloper was carrying something valuable.

But the finding that should stop you is a later one. In 2015 Graham Brown and Markus Baer published work in the Journal of Applied Psychology with a title that could serve as a subtitle for this whole chapter, “Protecting the Turf: The Effect of Territorial Marking on Others’ Creativity.” They showed experimentally that when people mark their territory it measurably suppresses the creativity of everyone else around it. The mere act of one party staking a claim causes other people to hold back, contribute less, stop bringing things forward.

Read that against the fleet and it lands hard. The Quality Systems team does not have to reject a single one of Maria’s ideas for the territorial effect to work. Their ownership of the patch, marked and visible and understood by everyone, is already, before Maria builds anything, a reason for her to keep it quiet. The border does not just repel the ideas that reach it. It stops most of them from ever being surfaced at all, and territoriality therefore does its heaviest work invisibly, in the ideas that die in the mind of someone who correctly reads whose patch they would be stepping on.

The old, honest version of it

None of this is new, of course, since nothing in this book is. The territorial reflex has a well-worn cousin you will already know by name, which is Not Invented Here, the reflexive rejection of ideas that came from outside the group.

Ralph Katz and Thomas Allen documented it properly back in 1982, studying fifty R&D project groups. They found that stable, long-tenured teams developed the belief that they possessed a kind of monopoly of knowledge in their area, and used that belief to reject outside ideas, to the measurable detriment of their own performance. They were hurting their own results to defend their ownership of the problem, and they could not see themselves doing it.

Later work by Dirk Antons and Frank Piller framed it precisely: Not Invented Here is best understood as an attitude-based bias, a psychological mechanism that leads people to reject external knowledge in an irrational, economically suboptimal way, regardless of how good the knowledge actually is. The value on offer does not enter the calculation. The only thing that matters is that it came from outside.

Which is Maria exactly. Her tool does not fail a cost-benefit test, it never reaches the cost-benefit test. It is disqualified upstream of merit by the single fact that it originated in the wrong place: three rungs down, one department over, outside the walls of the team that owns the problem.

Why this is the more important death

I have spent longer on the second mechanism than on the first, and that is deliberate, because it corrects what I think is the single most common mistake in how organisations understand their own buried innovation.

The anti-novelty bias is real, it is what everyone points to, and it kills radical ideas. But radical ideas are rare. Most of the value trapped in the fleet is not radical at all. It is Maria’s kind, meaning sensible, incremental, low-novelty, high-return improvements to things the company already does. Those ideas are not killed by the bias against creativity, because they are not creative in the sense that triggers it. They are killed by territory.

And this matters enormously for what you do about it, which is the whole back half of this book. If you believe the problem is fear of the new, you will invest in courage — innovation programmes, moonshot funds, permission to fail — and you will watch the fleet stay submerged, because you will have treated a disease the fleet mostly does not have. The increments are not hiding because the company is afraid of bold ideas. They are hiding because the company is organised into territories, and every good idea that crosses a border reads, to the border’s owner, as an incursion to be repelled, no matter how much value it happens to be carrying.

There is a second consequence and we will pick it up two chapters from now, but let me plant it here. Territory explains why the increment specifically submerges, and it does so without needing the idea to be novel. That closes a gap the standard story cannot close. If the only mechanism were fear of novelty, low-novelty increments would surface freely, and they don’t. Territory is why.

The gate is not the meeting

One last thing, because it is the piece that makes everything after this chapter make sense.

When I say “the gate,” it is tempting to picture the review meeting, the stage-gate, the steering committee, the formal go/no-go. Those are real and they do their share of the killing. But the deeper point of this chapter is that the gate is mostly not a meeting at all. The gate is the anticipation of the meeting.

Maria never presented her tool. There was no review, no panel, no formal no. The gate did its work anyway, perfectly and silently, because Maria could read the territory from her desk. She knew whose patch it was. She knew what “let’s align it with the programme” means. She had watched it happen to other people. The bias against novelty and the defence of territory did not need to be applied to her idea in a room, because she had already applied them to herself, in advance, and drawn the correct conclusion.

That is what a gate is, fully understood. Not a checkpoint that rejects ideas, but a pressure that stops them being offered at all. The rejections you can see are the small visible tip of it. The real gate is the one running in the heads of everyone who never bothered to walk up to it, because they already knew what it would say.

Which raises the question the whole back half of the book has to answer, and I will leave it hanging here. If the gate mostly operates by anticipation, then you cannot fix it by reforming the meeting. You have to change what people anticipate. Hold that thought, because it gets a good deal harder before it gets easier — next we leave the gate and go to the place the anticipation comes from, which in a manufacturing company is a very specific and very expensive kind of decision.

Next: Chapter 5. The Big Factory — where the tilt comes from, why the doors only swing one way, and the slow erosion nobody watches.