If You Want to Innovate, You Must First Take Care of the Plumbing

Mar 09, 2021 - 6 minutes read time

If You Want to Innovate, You Must First Take Care of the Plumbing

Backsliding. Several executives told me last week that their innovation momentum has been slowing down. They were referring to the initial productivity gain when people started remote work. They felt empowered. Decisions were delegated. And they got a lot done. 

Then things started backsliding. Somehow red tape has been growing back like ivy vines. But this isn’t a cultural issue. It’s a design issue. Specifically, it’s the uncontrolled growth of invisible work that is choking off creativity. Let me explain. 

There are two types of work we do every day. The first is the execution of a task, like writing a PowerPoint for a sales pitch. The second is managing workflow. You might need to run through some figures included your PowerPoint with your colleagues in the finance department. The trouble is, a task is visible. But the overall workflow is often hidden and ad hoc. And workflow is what structures our collective efforts.    

When it’s time to innovate, managers look at what products and services to offer. After all, those are what customers can see. Except in the narrow fields of manufacturing and service delivery, we rarely interrogate how we produce the desired result collectively—how we integrate product design, marketing and sales, legal, finance, production and customer support. 

This is unfortunate. Because the actual mechanics of how work is assigned, executed, and reviewed can make or break a company. It goes beyond better leadership. It’s not about setting clearer objectives. It’s not even about team collaboration. It’s operational science. It’s hard stuff. 

How Google Wins By Making Things Simple

You know how Google won the smartphone battle. Its Android operating system is free to hardware manufacturers like Samsung and Huawei. It quickly attracts third-party app developers like Facebook and Twitter. But guess what: Nokia had the same strategy. 
Nokia’s Symbian platform also deployed open-source software. Nokia knew that it had to woo app developers. The company harbored no arrogance. The board knew that only the paranoid survive. And the company was on a burning platform

But no one could do anything. Here’s an insider view by a young board member named Risto Siilasmaa. He would later become Nokia’s chairman. It was Siilasmaa who eventually led the company out from its near-death experience. He ultimately turned Nokia into the world’s third-largest telecom equipment supplier. 

Siilasmaa had been perplexed by the tardiness of the Symbian team. It had spent hundreds of millions of dollars hiring new programmers. Despite the extreme sense of urgency, software was always delivered late, and it was often buggy.    

Then he learned that it took 48 hours for a software programmer to compile the source code.  

Compilation is the translation of what programmers are writing into machine language that a computer can execute. “It was the equivalent of a movie director on a set having to wait 48 hours to view the latest take before deciding whether it needed to be redone,” said Siilasmaa. “Even a 24-hour compilation time was appalling. Waiting 48 hours to test how well a program works is an eternity.” The software team at Nokia wouldn’t be able to build what they needed. 

To Siilasmaa’s horror, he further learned that the overall build time—the time needed to gather and compile code from different teams—was up to two weeks. If it took this long for anyone to determine whether a change worked or would have to be redone, the company was doomed. “This was a recipe for catastrophe, and a catastrophe was exactly what we had staring straight into our eyes,” Siilasmaa said.

At Google, the same process took less than 20 minutes, and people still wanted it faster.

How You Should Think About Lead Time, Not Work Time 

The principle that Google came to embrace is no secret. There’s a whole community of DevOps professionals expounding on the benefits of letting teams make decisions without having communicate and coordinate with people outside. What you want is to avoid having to get approvals from distant authorities or committees far removed from the local work, because those are people who have no relevant basis to make good decisions. 

This is obviously not merely a cultural issue. It’s about how you structure an organization so that monitoring can be done automatically. Information can then be searched easily. And knowledge can be shared through a centralized repository. 

As I’ve argued here before, every tech giant has gone through a near-death experience like Nokia’s. But they have taken the medicine just in time to radically rebuild their organizations from the ground up. Usually, that means the input and output of each department gets standardized. Different teams learn to communicate through APIs (application programming interfaces). This means less email and fewer manual Excel spreadsheets. Then comes the next step: automating the routine co-ordinations. 

That’s why at Google Android, an app developer can run production-like environments on their own laptop. People can create, on demand, the testing environment they need via self-services. This is sustainable advantage. Now, you may ask, what does this all mean for my daily work?

You always want to look at your assignments from two perspectives: the work execution itself and streamlining workflow. 

When you are assigned a task, check where the task handoff is. Is it seeking approval from someone out of habit? Is the approver or reviewer adding real value? You want to eliminate review as much as possible. Why? Because even if it only takes three minutes for the other person to check your email, your email will be looked at after that person has finished their own work. That might mean it will take up to three days for you to get a reply. And if that person, in turn, needs to clarify something else, the complexity of the situation will be compounded. Your lead time will grow exponentially. 

Over the medium term, you want to agree with other departments on what you and your team can decide for yourselves. Trust me, most people have enough on their plates. They would be glad to have this conversation with you. 

Then you must make the invisible visible. Spend time with your team discussing the overall workflow. Knowledge work is tricky because the workflow cannot be seen easily. In a factory, if something isn’t working, you will see inventories and work-in-progress piling up. For knowledge workers, inefficiency is less observable. So unless you map out the workflow, you won’t know how bad things really are. 

Finally, at management retreats, you need to think big. Innovation is not about the front end. Launching a new app is easy. A hackathon can become an innovation theater. What you want is to apply the same passion to streamlining and simplifying. 

Smart companies pay down their complexity debt every day. They know that if they don’t, complexity will strangle their work. So it’s good take care of the plumbing. 


Stay healthy,

Howard Yu signature

P.S., Have you come across organizations that are disciplined in ridding themselves of complexity? How did they do so? What enabled them to stay focused? Join the discussion. Let us know what you think. 



Daniel Kearney Mar 11, 2021 at 3:52am

The "2 pizza team" concept and API approach to programming is the bedrock on which the Amazon innovation ecosystem is built. All companies are innovative (to varying degrees) but innovation at speed is more rare. Building up muscle within the organization to increase the speed of change is key; indeed it starts with the plumbing.

Martin Venzky-Stalling Mar 10, 2021 at 1:19am

I wonder if you could create an internal "open innovation" competition, a bit like Kaizen, listening to employees and teams on how to make things faster within the organization faster and foster an atmosphere where ideas can be proposed.

Perhaps you could even run a competition on hard wins, not quick wins, but how you could tackle high impact hard wins.

Luca Rigazzi Mar 09, 2021 at 8:29pm

“Simplify” fights every day with the old fashioned way of thinking: “We always made in this way, why change?”

Willy Sussland Mar 09, 2021 at 12:58pm

Thank you, as always very insightful ! Innovation does not sprout any old place, it needs the right setting. Look the top innovators on S+P or on the NASDAC. They are agile and innovative enterprises. Agile management and business ecosystems provide the platform of principles and practices for the right setting. The former advocates that teams keep things smart, simple, and swift. The latter highlights synergies. Both tackle complexity effectively, both eliminate bloated bureaucracy.

Carlie Mathis Mar 09, 2021 at 9:19am

Every company needed this. There's so many excellent must-take points here that can help future-proof innovation capabilities and strengthen business growth strategies. Great read.

Sally Jefferson Mar 09, 2021 at 8:20am

Well written, you always produce high quality blogs! I'm always learning from every article you posted.

Latest Thinking

What Is More Exciting Than Coinbase’s IPO in the Crypto World?

Apr 15, 2021

visual representation of the digital Cryptocurrency Bitcoin and a dollar bill

Gold is valuable. Pawnshops are not. When a new investment opportunity bursts onto the scene, people sometimes get confused between the value of an asset and the value of the…

Remote Work is Here to Stay. Here’s How to Restructure Office Interactions

Apr 13, 2021

People holding empty colourful speech bubbles

The interesting aspect of the post pandemic office won’t be who’s returning and who’s working remotely. It’s how we’ll change the way we interact. Certain work practices that we used…

Three Charts That Explain Investors Are Not Short Term, They Are Just Skeptical

Apr 09, 2021

Freelance broker trading through laptop at desk while working from home.

Short-term pressure—that’s what everyone complains about. In my discussions with senior executives at large companies, they speak openly about the market pressure for short-term performance. Specifically, it’s the pressure from…