Former Tesla president offers innovation tips

How should innovation be done in healthcare? Speakers at the recent HIMSS conference in Las Vegas addressed this topic, repeatedly pointing out that other industries have surged ahead on productivity by applying technology, while healthcare efficiency hasn’t improved. In fact, in healthcare, technology often makes things worse.

Just look at the application of EMRs, slowing down care by demanding too many clicks and time-consuming documentation.

Sometimes it helps to look at the experience of another industry, and the keynote address of former Tesla president John McNeill was enlightening.

“What we’re going to talk about is how you can systematically guide your organization to become more innovative,” said McNeill in a provocative and stimulating presentation. “You can become both intentional and repeatable with innovation.”

He focused on a few core principles:

• Define your problem and set bold goals.
• Question everything as a way of eliminating poor processes.
• Attack complexity and make things simpler.
• Often, innovation is a process of making things simpler.
• McNeill said the principles of innovation can be applied to any organization – including healthcare providers.

But importantly, when innovating, you should apply technologies like AI last, said McNeil. It may seem counterintuitive, but it forces you to improve your processes before you implement a complicated technology like AI. He explained this further after giving a bit of background.

When McNeill took on the president’s job at Tesla in 2015, he was faced with a crisis: Tesla was on the brink of bankruptcy. “We had 90 days of cash on the balance sheet, but it was worse than that,” he averred.

“We had 70 days of payables, so essentially, we had about three weeks worth of cash.” Wall Street figured this out, and they were predicting Tesla’s demise, he said.

So, the challenge was not only to turn the business around, but to survive.

McNeill described how Tesla had its chain of storefront dealerships. But it also had an online business, where people could buy a $100,000 car online, sight unseen.

After brainstorming, the team came up with a clear and aggressive goal: increase online sales by 20X.

To do this, they first looked at the barriers people faced when trying to purchase a Tesla on the web. The major issue was that it took 64 clicks to get from start to finish – people who wanted to buy a car simply gave up through exhaustion.

By contrast, they saw it took only 10 clicks to order a pizza from Domino’s.

“Elon [Musk, the CEO] said, ‘Alright, that’s the goal. We’re going to get clicks down from 64 to 10.”

How did they do it? They realized that to buy a car online, there were too many options – the interior, the motors, rear or all wheel drive and more.

But in analyzing buying patterns, they saw that customers ordered just three different cars – basic, mid-range, and top-tier, with all the options.

Moreover, 44 of the 64 clicks concerned financing – and people were getting hung up on the complexity of financing.

So, the management team did two things. They told the engineering department they only needed to produce three different options – basic, mid-range and souped-up.

“To my surprise, they said, “We have been waiting for someone to clear up this mess.”

Engineering told McNeill and Musk, “You don’t understand how complex our supply chain is, and how hard it is for our production systems to accommodate an infinity of configurations.”

Reducing the number of configurations made sense – and it was quickly done.

“Then we were down to the 44 clicks in financing,” said McNeill.

With a little more research, they found out that most of those clicks weren’t needed. A myriad of loan and lease clauses were there because, over the years, the lawyers of car companies wanted to protect their clients from liability.

However, most of the terms and conditions weren’t legally required.

“I said, I don’t believe this, but a 12-page loan document comes down to one paragraph, it comes down to how much are you going to pay for a car – what the interest rate is, over what period of time, and what the monthly payment is,” noted McNeill.

They then found a bank, US Bank in Minnesota, that would work with Tesla and process the loans.
“I applaud that institution, because they were the first to say, we’ll give you a one-click loan and one-click lease.”

This questioning of the way things were done resulted in a major innovation. “We got from 64 clicks down to 13. We didn’t get to 10, but we got to 13. And Tesla’s online sales went up by 20 percent.”

Thus, to innovate, Tesla had questioned the assumptions of lease and loan practices “that all our peers had overlooked.”

Meanwhile, Tesla had 1,200 cars sitting in a lot because the doors didn’t fit. No one in the organization could find out why not.

It turned out that the engineering team hadn’t factored in the weight of the wheels – the last step in assembly. The extra weight made the frame buckle slightly, and the doors wouldn’t close properly.

McNeill explained that originally, no one wanted to question the engineers on whether they had included the wheels in their calculations – it would be too embarrassing to ask about such a simple thing.

But the management team finally did it. Engineering checked and to their chagrin, discovered the error.

After the engineers fessed up, it was an easy matter to add a piece to reinforce the skeleton of the car.

“It took us a matter of weeks, but that saved the company,” said McNeill, because at $100,000 a car, the solution provided $120 million in quick revenue.

It was another case of questioning everything.

Online sales then soared, helping to turn around the company.

McNeill talked about one other challenge at Tesla, and it was a big one.

He approached the issue of technology and automation. “You’ll notice that I’m 10 minutes in and I haven’t mentioned AI or automation,” he noted.

He emphasized that automation should be the last step in innovation. It comes after redesigning your processes, the way you work.

“We automate last because we believe that if you automate a flawed process, you’re just getting to the bad answer faster,” said McNeill.

Moreover, if you invest a lot of money in technology at the beginning, it’s very hard to make changes later. Even if mistakes have been uncovered, the investment might have been so great that it’s nearly impossible to start over again.

He said the first step in solving a bad process is to map it out and identify all the existing steps in the process.” It’s the part where you put yellow stickies on the wall.

He advised putting a circle around all the steps that add value and eliminating the rest.

This process when Tesla set an enormous goal: reduce manufacturing costs by 50 percent.

McNeill described how cars were produced – there were two buildings, each a kilometer long. One turned out the chassis (what is also called the skeleton), where 300 robots put the underlying frame together, and another factory for general assembly. “That’s where human beings are hanging parts on the skeleton.”

In this second factory, everything was done by hand. While some would have added robots to automate and bring down costs, Tesla did something different.

It examined the whole process and realized that much of the hand-done work could be eliminated – the outer body could be moulded from aluminum and snapped onto the skeleton.

What they saw was that the body of the Tesla consisted of many aluminum pieces that were being bolted onto the skeleton one piece at a time.

The idea for doing this was inspired by something different, but related. A colleague of McNeill one day rolled a toy metal car across a table top. His kids played with the little vehicles – all which had molded metal bodies.

He asked whether the body of full-sized automobile could be moulded into two pieces and then secured onto the chassis – in this way eliminating many steps in the factory and a good deal of labour.

Everyone said it couldn’t be done – nobody had ever cast aluminum parts that big. The fear was that the body would crack.

But McNeill and his team started experimenting on their own. They smelted their own aluminum behind the factory. They started small, and little by little worked up to the large pieces that were needed.

“We figured out how to cast a whole side of the car,” said McNeill. “And so today, the factories that Tesla builds are half the size they used to be.”

He said this one design improvement “took 50 percent of the factory core out, and 50 percent of the costs.”

Moreover, Tesla has created a process that nobody else has. “We invented the machines that could do this. Now, you can imagine Toyota, Honda, GM, Ford, Stellantis – the rest of industry wants to do this. But you can’t buy the machines.”

It was a clear example of his innovation principles – think boldly, question everything, and make things simpler.

Finally, McNeill described how technology – used too early – can be the enemy of innovation.

In designing a factory to produce a new model, the company had used computer-aided design (CAD). However, the computerized model didn’t account for the amount of space needed between robots and machinery for repairs.

Thus, when the factory was built, and machines went down, it was impossible to repair them.

“We were stuck. We couldn’t produce a single car because we had automated it first,” said McNeill.

To rectify the problem, a Tesla employee led a team in a tent that would build the cars by hand.

In this way, they learned how to design an effective process for turning out cars – including enough space to get around. Not only was the original problem fixed, but the team pioneered a new way assembling cars.

Instead of just-in-time delivery of parts, which can go wrong when suppliers can’t deliver, the Tesla team stockpiled a cache of all the parts they needed outside the tent.

The parts could be used as needed, when needed – no waiting for just-in-time deliveries.

“So, if you look at our factory in Shanghai today, that breakthrough has caused one of the biggest changes in manufacturing,” said McNeill. “Almost every new car is designed this way.”

He stressed that once the process was mapped and improved in the tent, it could be automated and expanded to factories around the world. By tinkering with processes on a small scale, often with minimal technology, teams can improve ways of working before scaling up and adding technology.

Lastly, McNeill said managers and designers should be users of the products they’ve produced. In this way, they can experience the solutions and provide feedback to management.

McNeill is now doing work for General Motors. As part of his work, he drove an electric Cadillac on a four-day, 2,000 mile trip. “So I could experience firsthand what the charging experience was like,” he said.

Afterward, he wrote an extensive report for his boss, GM’s CEO Mary Barra.

“It’s a feedback loop that executives can use in their companies.”