Creativity and Innovation: Ideas Got Cheap. Adoption Didn't.

Somewhere in your organization there is a folder of things that worked.

A better way to sequence the picking waves, tested in one warehouse over six weeks, measured, proven. A revised handover template that took a customer onboarding from eleven days to six. A small change to how one regional team handles short shipments, invented by a planner who never told anyone because nobody asked. Each of these was creative in the only sense that matters commercially. Someone looked at a process that everybody else experienced as a fact, recognised it as a decision, and made a different decision that turned out better.

Most of them are still in the folder.

This is the shape of the creativity problem in 2026, and it is not the shape most of us were trained to expect. In the past, we looked at organizations that fail to innovate and thought they lacked imagination, and that the remedy is to get more ideas into the room. Then, we held workshops, innovation challenges, and offered a prize for the best suggestion in the Idea Box.

Since then, in the space of about three years, ideas became free.

Any manager with a laptop can now generate forty options for restructuring a distribution network before the coffee gets cold. The AI will, with great confidence, offer options which are plausible, well-argued, and formatted in a table (or a list of bullet points depending on what it gleaned from your preferences). Some of the ideas will be genuinely good. The generative constraint that shaped a century of management thinking about creativity has been lifted almost overnight, and what has become visible underneath is the constraint that was always doing most of the damage.

Ideas do not fail in your company because there are too few of them. They fail on the way into operations.

That distinction matters because the two problems have almost nothing in common. A shortage of ideas is solved with stimulation: new inputs, different people in the room, permission to be wrong. A failure of adoption is solved with carefully listening to the people doing the work, sequencing their needs, and patiently removing obstacles which happen to be mostly social. Organizations reliably diagnose the first and are reliably suffering from the second, which is why so much money is spent on the wrong intervention by people acting in good faith.

The bottleneck moved

The World Economic Forum's Future of Jobs Report 2025 ranks creative thinking fourth among the core skills employers say workers need today, behind analytical thinking, resilience and flexibility, and leadership and social influence. It also sits in the report's top-right quadrant: skills that are core now and expected to become more so by 2030. The report adds a detail that reads differently for anyone working in supply chain than many other departments. Slower economic growth and tighter trade restrictions are among the forces pushing creative thinking up the list. Creativity is not rising because the world got more playful. It is rising because supply chains/ trade is ever more constrained (tariffed, sanctioned and rerouted), and constrained systems have to be re-thought rather than merely funded.

Now hold that alongside what is happening to adoption.

BCG's fourth annual AI at Work survey, published in June 2026, found that regular AI use among frontline employees has surged to 74 percent, up more than twenty points in two years. Forty-two percent of those regular users report saving at least a full working day every week. That is an enormous release of human capacity, and it is the closest thing to a free lunch most organizations will ever be handed.

Yet, sixty-six percent of them say they get limited or no guidance on what to do with that time, and more than half do not redirect it into anything strategic. BCG's phrase for what happens next is worth borrowing: the time saved leaks out of the organization. It does not become better work. It becomes slightly earlier evenings, slightly longer water-cooler talk, and a productivity gain that no one can find in the numbers.

The same survey found that strategic clarity lifts measurable business impact by around 25 percentage points, while better tools, absent that clarity, move it by roughly five. Companies that actually redesigned their workflows were 24 points more likely to see measurable improvement. The gap between the two figures is the entire subject of this article. Capability is cheap now. Reshaping the work to absorb the capability is not.

There is a more sobering number in circulation, and it deserves both citation and caution. A 2025 report from MIT's NANDA initiative, The GenAI Divide: State of AI in Business, examined more than 300 enterprise deployments and concluded that roughly 95 percent of corporate generative AI pilots had produced no measurable return. The figure travelled around the world in a week, and it is worth saying plainly that the report was not peer reviewed and that the definition of return does a great deal of work. But the authors' diagnosis is the part many of us see happen from the inside. The pilots were not failing because the models were weak. They were failing because organizations had no reliable way of turning something that worked in one place into how the work is done everywhere.

Which is a description of an innovation problem that has nothing whatsoever to do with a shortage of ideas.

Three words we keep confusing

Before going further it is worth separating three things that get collapsed into one another, because confusing them sends leaders to the wrong remedy.

Novelty is something new to you. It is the raw material, and it is now essentially unlimited. A model can hand you a hundred units of novelty in a minute, most of it useless, some of it interesting, and none of it yet worth anything. Novelty has no cost and no value.

Creativity is novelty that is also useful in a particular context. It requires knowing the context well enough to judge, which is why it remains simultaneously rare and human. The planner who knows that the third supplier always ships late in the last week of the quarter can tell in in an instant which of the model's hundred suggestions is worth keeping. The model cannot, because the model does not know about the third supplier’s behavior and which week of the month it is, unless someone tells it. Creativity is the marriage of an outside idea to inside knowledge.

Innovation is creativity that survived. And here is the definition worth carrying into your organization: an idea has become an innovation when it is simply how the work gets done on an ordinary day, by people who were not in the room when it was invented. Who heard about it only after it was approved and piloted.

Most organizations that describe themselves as short on creativity are, on inspection, perfectly full of it and catastrophically bad at the third stage. They have folders of things that worked. They diagnose an idea shortage, run another ideation workshop, and add to the folder. (Or they are stuck discussing the same idea for months and still haven’t run the experiment.)

The engine underneath: seeing the process as a choice

If innovation is the hard part, it might seem sensible to skip straight to the mechanics of adoption. But there is a step further upstream that determines whether you have anything worth adopting, and it is quieter than brainstorming.

People cannot reimagine what they experience as a fact rather than a decision.

Almost every process you touch this week was designed by someone, at a particular moment, under conditions that were true then. The two-week freeze on the production plan exists because in 2014 the ERP batch job ran overnight and the data was not available sooner. The three approval signatures exist because of an incident in a business unit that no longer exists. The safety stock policy was set when lead times from that region were nineteen days rather than six. None of these were laws. All of them were decisions, and most of them were reasonable decisions, and a good number of them are now solving a problem that stopped existing years ago.

But that is not how they feel from the inside. From the inside they feel like the weather. And you cannot have a creative idea about the weather.

This is why the most productive creative act available to most professionals is not idea generation at all. It is archaeology. Take one routine step in your process, the one that irritates you most (or rather the new hire doesn’t get), and find out who decided it, when, and what conditions they were solving for. Three questions, one conversation, usually less than an hour. Roughly a third of the time you will find that the reason is still perfectly good and you will stop complaining, which is itself worth the hour. Roughly a third of the time nobody will remember, which is a finding. And the remaining third of the time you will discover that the original condition has evaporated, and you will be standing in front of an opportunity that was invisible ten minutes earlier – dare to delete the step.

Teresa Amabile's research points at the conditions that make this kind of thinking possible or impossible. Working with Constance Hadley and Steven Kramer, she analysed more than nine thousand daily diary entries from people doing work that required creativity, and found that the common managerial belief about pressure has it backwards. The days that produced creative thinking under high pressure were the days people described as being on a mission: focused, urgent, protected from interruption. The days that produced nothing were the days people felt on a treadmill, switching between fragments. And under low pressure, the split was just as sharp. People either felt they were on an expedition, exploring, or on autopilot, executing without engagement.

The practical implication is that everyone has creativity. Creative thinking is not a personality trait you have or lack, but the output of a particular kind of hour. If your week contains not a single focused hour, you have not discovered that you are uncreative but that your work is fragmented. You are on a treadmill, and treadmills are a scheduling problem before they are a talent problem.

Two ways creativity fails now

Two failure modes have become common enough to name, and they come from opposite directions.

The first is convergence. In 2024, Anil Doshi and Oliver Hauser published a study in Science Advances that has aged into one of the more important findings of the AI era. Writers who were given story ideas from a large language model produced work that was judged more creative, better written, and more enjoyable than writers working alone. The effect was strongest among the least creative writers, which is the good news and is genuinely good. But the AI-assisted stories were markedly more similar to one another than the unassisted ones. Individual creativity went up. Collective diversity went down.

Read that as an operations leader rather than as a novelist. If every planning team in your industry is asking similar models similar questions about network design, the range of answers being considered across the whole sector is narrowing while every individual team experiences itself as more inventive than before. Everyone gets better ideas. Everyone gets closer to the same idea. In a market where advantage comes from doing something your competitors are not doing, that is a slow and almost undetectable problem, and the people experiencing it will report high satisfaction throughout.

The defence is a matter of sequence, and it costs about ten minutes. Write your own three or four options before you open the model, not after. Once you have read a fluent, confident, well-structured list, your own thinking bends towards it, and the odd idea you would have had is quietly never had. Generate first, then consult. Use the model to stress-test, extend, and find the flaw in what you already thought, which it does superbly, rather than to hand you the starting position.

The second failure mode is more familiar and more expensive. Call it creativity theatre.

It has a recognisable form. A day is booked. Someone brings colored pens, colored papers and a whiteboard. The energy in the room is real, and people who never normally speak in front of the regional director say something interesting. Photographs are taken of walls covered in notes. A summary deck circulates the following week with fourteen themes on it. Then, the next Monday, nothing changed. No standard procedure is amended, no system configuration changed, and eleven months later a new consultant observes that the organization struggles with innovation and proposes a workshop.

The tell is simple and you can apply it to any innovation activity you are asked to attend. Ask what document, system, or standard will be different at the end. If the honest answer is a deck, you are attending theatre. Theatre is not worthless, because it can build relationships and surface information. But it should be budgeted as engagement, not as innovation, and it should never be the organization's answer to why nothing changes.

The cure is not a better workshop. It is a smaller one with a different ending. Run the session for ninety minutes rather than a day, take the two most promising ideas rather than fourteen, and refuse to close the meeting until each has a named owner, an implementation date, and the specific artefact that will be different. Everything else in the room goes on a list that everyone understands is a list, which is more honest than the deck and takes less time to produce.

The finish line is boredom

Innovation is done when it has become a standard operating procedure.

Here is the reframe that does the most work in an operational setting, and it is the opposite of what most innovation language implies.

An idea has arrived when it becomes boring. When it is written into the standard operating procedure, configured as the system default, included in the induction for new starters, and referred to by people who have no idea it was ever anyone's idea. The last mile of innovation is not a launch. It is a documentation task, a training slot, a parameter change, and the patient removal of the old way so that it stops being an option.

This is unglamorous work, and it is systematically under-owned. The person who had the idea finds it tedious. The person running operations did not ask for it. The improvement team has moved to the next thing. So the idea sits in a state of permanent pilot, alive in three sites, notional in fifteen, requiring a champion to keep breathing, and it dies quietly the month that champion changes role.

The same reframe has a second edge, and it is a caution against enthusiasm. Not everything should be reimagined. An organization that innovates everywhere innovates nowhere, because every change consumes attention, retraining, and error tolerance from the same finite pool. The core of your operation, the thousand things that must simply work, earns its keep by being standardised, boring, and left alone. That is precisely what funds creativity at the edges. When somebody proposes reinventing the invoicing process because it feels old-fashioned, the right question is not whether a better invoicing process is imaginable. It is whether this is where you want to spend your organization's limited capacity to absorb change.

Deliberate boredom in the core. Deliberate novelty at the edge. Anyone who tells you to be creative about everything is describing a company that will change nothing.

Why proven ideas stall

So a thing has been tried, it worked, and it has not spread. What actually stops it? In practice three frictions do nearly all the damage, and each has a counter-move.

The first friction is that people reject novelty while sincerely believing they want it.

Jennifer Mueller, Shimul Melwani and Jack Goncalo demonstrated this in a 2012 study with an unusually honest title, The Bias Against Creativity: Why People Desire but Reject Creative Ideas. Participants consistently rated creativity as desirable and then, when placed in conditions of uncertainty, evaluated creative ideas less favourably than practical ones. The bias was not a matter of taste or stubbornness. It switched on when people needed to reduce uncertainty, and it operated below the level at which they could report it. Ask them and they will tell you, truthfully, that they value new thinking.

This explains an experience nearly every reader will recognise: presenting something genuinely new to a leadership team that has just finished saying it wants bold ideas, and watching it be received politely and buried. They were not lying in the first half of the meeting. They were uncertain in the second.

The counter-move follows directly from the mechanism. If uncertainty is what triggers the rejection, then reducing uncertainty is what protects the idea, and you reduce it by anchoring the new thing to something familiar and already trusted. Lead with the precedent rather than the novelty:

“This is the same logic as the returns process we changed in 2023, applied to inbound. Two sites have run it for nine weeks and here is what broke. If it fails, here is exactly what we lose and how we revert by Friday.” None of this dilutes the idea. It removes the specific thing that kills it.

The second friction is that the cost of adoption lands on people who received none of the credit.

This one is almost never stated out loud, because saying it sounds petty, and so it operates undisturbed. The team that invented the improvement got the visibility, the mention in the quarterly review, and the satisfaction. The fifteen teams asked to adopt it get a fortnight of doing their jobs badly while they learn something they did not choose, at a moment when their own targets have not been adjusted to accommodate the dip. Their reluctance is not resistance to change in any psychological sense. It is an accurate reading of an incentive.

The counter-move is to make the switching cost visible and then to actually pay it. Name the dip out loud, including its likely size and duration, because a predicted dip is tolerable and an unexpected one is evidence that the idea does not work. Move the target for the transition period, or state clearly that you will not hold the shortfall against them. Send the people who already did it to sit with the people about to do it, which converts a mandate into help. And give the adopting sites, not the inventing site, the credit for the second and third implementations, because in operational terms those are harder than the first.

The third friction is that nothing enters daily work without an owner inside the existing operating rhythm.

Ideas that live in a programme, a steering committee, or a transformation office are visitors. They are discussed in forums that exist to discuss them, by people whose job is to discuss them, and they never touch the meeting where the actual week is planned. An idea becomes real at the moment it appears as a line in the recurring operational review, owned by the person who owns that review, with a number attached that they are accountable for.

The counter-move is to stop asking for a decision and start asking for a slot. Before you next propose something, find out which standing meeting governs the process you want to change, who chairs it, and what metrics appear on its agenda. Then frame your idea in the vocabulary of that meeting and that metric. An improvement that cannot be expressed in terms of something already on somebody's dashboard will not survive its first busy month, however good it is.

Push it or park it: three questions

This leaves us with the routine of a regular week: You have an idea, or someone has handed you one, and you have limited energy and limited credibility to spend. Three questions, asked in order, resolve most cases.

Does it survive a bad week? Every improvement looks good on a calm day with full staffing and clean data. The real test is the week when two people are sick, a supplier misses, and the system is slow. Does your new way of working still function then, or does it require conditions that hold about sixty percent of the time? Ideas that are only robust under good conditions do not fail loudly. They get quietly abandoned during the first crisis and never resumed, and everyone concludes the idea was bad rather than that it was fragile. If the honest answer is that it needs calm conditions, that is not necessarily a no. It is an instruction to simplify it until it survives the bad week, which usually means giving up the most elegant twenty percent of it.

Who absorbs the switching cost, and have you asked them? Not who benefits, which is the question proposals always answer. Who pays, in learning time, in error rate, in a temporarily worse number on their own scorecard? If you cannot name those people, you do not yet understand your own idea. If you can name them and have not spoken to them, that conversation is the highest-value hour available to you this week, and it will usually improve the idea as well as its chances. The person who will have to live with your improvement knows three things about it that you do not.

What becomes the default if it works? Push through to the boring end before you start. Which document is amended, which system setting changes, which step disappears from the induction, and critically, what stops being permitted. Ideas that are added without anything being removed become the twelfth item on a list of eleven things people already cannot finish, and the organization absorbs them by ignoring them. If you cannot say what the new default is and what it replaces, you are proposing an addition, not a change, and additions are how operational complexity accumulates while nothing actually improves.

Watch the filter work on something ordinary. A customer service lead has found that letting agents resolve credit notes below a certain value without approval cuts resolution time by roughly two days. She piloted it with six agents for a quarter and the numbers hold. Does it survive a bad week? Mostly, but not quite as designed: on high-volume days the agents skipped the two-line justification note, which is the only control the finance team has. So the idea gets simplified. The justification becomes a dropdown with four options rather than free text, which survives a bad day. Who absorbs the switching cost? Not the agents, who gain. It is the finance team, who lose visibility they currently have and inherit a monthly reconciliation they did not ask for. She had not thought about them at all, and the twenty-minute conversation that follows produces the dropdown categories, which is the fix to the first problem. What becomes the default? The approval step is deleted from the workflow tool rather than made optional, because an optional old path is the old path. The threshold goes into the agent handbook and the induction, and the finance report gains one line.

Three questions, one afternoon, and an idea that would otherwise have joined the folder is now simply how credit notes work.

Three yeses: push it, properly, with allies and evidence and the stamina to still be talking about it in month four. Any no: park it deliberately, write down what would have to change for the answer to flip, and put your energy somewhere it can land. Parking an idea consciously is a creative act. Losing one by attrition is not.

Five practices

Compressed into things small enough to start this week.

1. Do the archaeology on one thing. Pick the process step that annoys you most and find out who decided it, when, and what they were solving for. One conversation. You are not looking for a culprit, you are looking for a condition that has expired. This is the cheapest source of genuinely creative options available to you, and almost nobody does it.

2. Think before you prompt. Write your own three options before you brainstorm with a model. Then use it to attack them, extend them, and find what you missed. This costs ten minutes and it is the difference between a tool that widens your thinking and one that quietly narrows it towards everyone else's.

3. Book one unfragmented hour. Not for output. For the expedition. If your calendar contains no such hour, that is the finding, and it is fixable at your level in a way that most things are not.

4. Anchor the new to the trusted. Before presenting anything novel, find the precedent it resembles and lead with that. Name the failure mode and the reversal plan yourself, in the first two minutes, before anyone has to ask. You are not weakening your case. You are removing the uncertainty that would otherwise kill it.

5. Finish something boring. Take one idea already proven in your organization, ideally not your own, and drive it the last mile: into the procedure, into the system default, into the induction, into the standing review with a named owner and a number. One completed idea is worth more than a folder of proven ones, and the folder is where most of your organization's creativity currently lives.

The last of the six

Long-time readers will remember where TransNord Logistics left this thread. In the June session, a junior analyst suggested that the company share delivery networks with its competitors, and the room went silent. It was the best idea anyone had that year. The interesting question was never whether it was clever. It was whether an organization built on the promise of delivering certainty could carry an uncertain idea far enough for it to become ordinary.

That is the question the whole series has been circling. Creativity and Innovation is the sixth and final skill of the NorthStar Critical Skills Journey, and it turns out not to be a separate subject at all. Seeing a process as a decision rather than a fact is critical thinking. Finding out what the adopting team knows that you do not is deep listening. Knowing whether the improvement is real is analytical judgment. Surviving month four is resilience. Getting fifteen sites to change how they work is influence. Creativity is where the other five arrive.

Which is a more demanding conclusion than the one the innovation literature usually offers, and also a more encouraging one. You are not waiting on a rare gift. You are waiting on an hour, a conversation, and the willingness to finish something after it stops being exciting.

We will put this under pressure with real cases in Session 6: Creativity and Innovation, on Thursday 20 August at 2.00pm CET, 8.00pm Singapore time. Bring the idea in your folder.

References

Amabile, T. M., Hadley, C. N., & Kramer, S. J. (2002). Creativity Under the Gun. Harvard Business Review, August 2002. hbr.org.

Boston Consulting Group (2026). AI at Work: Strategy Matters More Than Tools, fourth edition, June 2026. bcg.com.

Doshi, A. R., & Hauser, O. P. (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances, 10(28). science.org.

Challapally, A., Pease, C., Raskar, R., & Chari, P. (2025). The GenAI Divide: State of AI in Business 2025. Accessible via MIT Medialab’s NANDA project and MLQ.AI. Note that this report has not been peer reviewed.

Mueller, J. S., Melwani, S., & Goncalo, J. A. (2012). The Bias Against Creativity: Why People Desire but Reject Creative Ideas. Psychological Science, 23(1), 13-17. journals.sagepub.com.

World Economic Forum (2025). The Future of Jobs Report 2025, Chapter 3: Skills Outlook. weforum.org.