How to Fix Broken Supply Chains | Dustin Burke | TED

Transcriber:

It was a strange feeling, walking into the local supermarket,

only to find empty shelves,

and, most notably, no toilet paper.

We didn't have raw-material shortages or manufacturing defects,

and we didn't discover new uses for toilet paper.

It was panic buying by everyday people.

Supply chains just couldn't keep up.

And before we knew it, the rumored shortage became a real one.

You remember that, don't you?

Well, maybe not, because I'm not talking about COVID-19.

I'm talking about the great toilet paper shortage of 1973.

And it wasn't caused by a pandemic,

but by a joke told by Johnny Carson.

But today's supply-chain challenges are no joke.

Those problems are real,

but they're problems that we've faced, and even solved, in the past.

A supply chain is the long and often complicated journey

that any item takes before it winds up in your home.

Raw materials are mined or grown

and sold to various suppliers.

Those suppliers sell them to manufacturers,

who transform those raw materials into finished goods.

And those finished goods are moved around the world

by distributors and carriers,

who, in turn, sell them to retailers,

who sell those to consumers as a final step.

Many supply chains are simple,

like when you buy strawberries at a local farmers market.

But some are almost infinitely complex.

In my 14 years working with companies on improving their supply chains,

I've seen many disruptions, from natural disasters to pandemics

and geopolitical instability.

And every time, the media talks about how, from this point,

companies can and will make their supply chains more resilient,

and the common prescriptions include diversifying risk,

better forecasting the future

and building buffers, like stockpiles of inventory

or more manufacturing equipment.

And this is good advice.

But the question I keep asking myself is:

“Why haven’t more companies taken this advice?”

The reason is that it doesn't stand up against competing priorities

and steep competition that occurs between the crises and shocks.

So if we want to build more resilient supply chains

that can withstand the next great crisis,

then we need to bring new ideas that can withstand competitive pressures.

Let's talk about sharing risk,

radical transparency and automated recommendations.

These three ideas, if we take them together,

have the potential to help break the trade-off

between resilience and efficiency.

An obvious solution to supply shortages is to build more buffers,

so that if anything happens along the supply chain,

the next recipient isn't waiting empty-handed.

Retail stores can never perfectly predict what we'll buy,

so they carry extra inventory.

They might run out of a particular size or particular color,

but it’s unlikely that they’ll run out of an entire category, like jeans.

But we know, across industries,

companies carry less of that backup inventory than they used to.

And we have immense product variety.

But much of what we buy

is made in highly specialized or automated factories,

and that makes it harder to repurpose that capacity when demand changes.

That's why during COVID-19,

those toilet-paper manufacturers had enough capacity,

but they didn't have the right kind of capacity.

Commercial toilet paper is very different from what we consumers use at home.

They could have met demand,

but it would have meant shoppers lugging home rolls that are nine inches wide,

and clearly designed for cost, not for comfort.

(Laughter)

Streamlining supply chains

is a big reason why we consumers have incredible choices at low prices.

And for many companies,

warehousing extra raw materials

or keeping idle equipment on the factory floor

is simply too expensive.

Competition is steep as it is.

But buffers matter in a crisis.

So how do we get similar benefits, but in a different way?

We can share or pool risk.

It's much like a well-established industry that we all know well --

insurance.

It's unlikely that you'll ever get into a major car accident,

but if you do, it will be horribly expensive.

That's why, for low-probability, high-impact events,

we share risk,

and in some instances, like car insurance,

we even require it.

We could do the same thing in supply chains.

Industry players could team up to share the cost of keeping extra raw materials,

key components or even machinery,

in exchange for an ongoing shared fee,

and only to be used in times of a crisis.

For example,

many countries stockpile important medicines,

but relatively few keep their active ingredients, or "APIs."

Pharmaceutical companies could share the cost of storing extra APIs

during normal times.

Then, in a crisis,

drug manufacturers could dip into that supply

to avoid running out of these important medicines.

Again, it's just like insurance,

except instead of pooling money, we're pooling a physical resource.

And importantly, it would be paid for, and so not seen as pure waste.

But in order to do this, we need to know who shares the same risks.

And in order to know that,

we need a radical improvement in supply-chain transparency.

Right now, seemingly everything we want to buy is delayed

because of a lack of microprocessors,

even cars.

But how could the auto industry --

arguably the industry that invented supplier collaboration

and supply-chain visibility --

be caught by surprise?

Microprocessors power the computers

that, increasingly, make our cars work the way they're supposed to.

And early in the COVID-19 pandemic,

auto manufacturers canceled orders for microprocessors

out of fear that car sales would plummet.

And around the same time,

demand for those microprocessors in consumer electronics went way up.

Parents bought tablets so their kids could learn from home,

newly remote workers bought laptops.

Then, a fire at a chip plant

meant fewer of these microprocessors were even being made.

And by the time car sales began to recover,

microprocessors were on back order.

In other words,

the reason why automotive companies can't get the microprocessors they need

has little, if anything, to do with the automotive industry.

And this is a good example of a broader problem.

In supply chain,

your risks are not tied only to your customers or to your competitors,

but also to those companies who are using the same inputs.

I doubt Ford considers the PlayStation 5 to be a competitor to the F-150,

but in this case, they could have been competing for the same, scarce resource.

And that's why we need a radical improvement in supply-chain transparency.

It's not enough to know who your suppliers are.

You have to know who your supplier's suppliers are,

where those supplier’s suppliers get their raw materials,

who else is buying from those suppliers,

and who else is competing for those raw materials.

Often what looks like several unrelated product lines

can be traced back to a single source.

We think we have diversified supply chains, but we don't.

We have a supply web.

And if key players in that web fail,

then many supply chains are in big trouble.

If we want more resilient supply chains,

then we need accurate, up-to-date maps of key inputs

and where they come from,

in any given industry.

Supply-chain managers are, fundamentally, planners.

They analyze and interpret information,

and they revise their plans to efficiently meet expected demand

with a steady stream of supply.

Supply-chain control towers bring all of that information into one place,

and better data should help better decision-making.

But too much data can simply be overwhelming,

and that's why we need technology to help us.

The good news is that advances in data mining,

artificial intelligence and machine learning

increasingly mean that computers can help us

analyze thousands or millions of points of data,

predict problems before they arise, notify managers

and even recommend actions to take.

Right now, some types of plastics are in short supply.

Plastic resin manufacturers decreased production,

dealing with COVID-19 outbreaks.

But by the time they began to recover,

they were hit with a string of natural disasters,

from Hurricane Laura on the Gulf Coast in 2020

to a brutal winter storm in the South in 2021.

No amount of data-sharing could have told us

what companies would be affected,

what companies don't use plastics.

But a computer could have helped spot the problem among the data points.

Decreased production and COVID-19 lockdowns,

combined with severe weather,

a surge in demand for plastic packaging as grocery sales grew,

and a spike in trucking rates

would have been enough to trigger some warnings in specific areas,

and then notify managers and recommend what they might do about it.

"Hey, this supplier is looking risky.

Do you have a chance to order from somebody else?

Maybe start calling other trucking companies

to make sure you get your supplies on time,

and we should notify these customers that their orders might be delayed."

This is similar to how the airline industry

manages passenger delays and missed connections.

Computers analyze data on available flights, weather conditions,

passenger locations and passenger destinations,

to reroute hundreds or even thousands of people in a day.

And we could do the same thing in a supply chain.

I don’t want to sound

like the things supply chain managers have done haven’t helped,

but we can do better.

When the virus struck,

we faced a dire need for N95 respirators and personal protective equipment.

Government stockpiles of N95s were depleted,

hospital orders were backlogged two to three years,

and health-care workers reused single-use protective equipment.

But again, I'm not talking about COVID-19.

I'm talking about the 2009 outbreak of the H1N1 flu.

And we have to stop repeating the same mistakes.

So let's be imaginative

about how we challenge businesses and government

to use shared risk,

supply-chain transparency and automated recommendations

to make our supply chains more resilient.

Because if they do,

we can be more resilient too.

Thank you.

(Cheers and applause)