LecturesPanel discussions

What makes a behemoth remain nimble? Lessons from Google

Read Time:15 Minute

Dr Suresh Ramanathan, Dean & Principal, Great Lakes
Institute of Management, in conversation with
Mr T T Ramgopal – Global Head, Android Partner Engineering, Google.

Dr Suresh Ramanathan: Companies like RCA, Kodak and RadioShack were behemoths in their own right, at one point in time. Yet they seem to have lost their way and some have even receded into the annals of history. What gives Google the longevity that it enjoys right now?


The companies that you mentioned were around for 100 years. They have been through World Wars, multiple depressions and so on. Google is 22 or 23 years old and it is just a young adult. So I cannot compare Google with those companies. However, if you look at Google in the technology timeframe, I would say it has done really well. We have an approach of launch and iterate.
Why did we create Gmail, when it was just another email product? A few engineers said they found it difficult to understand conversations and how they tied to each other and that it would be nice if they were clustered together. The second issue was storage. We built this big cloud with a lot of storage, so you don’t have to constantly delete emails because you run out of space on your local machine. Some of these led to the creation of a tech centric product. When we launched it, it had a lot of errors and bugs in it and over the years, we constantly reiterated and made it a much better product. Similarly Chrome. The world didn’t need another browser. This is something that Sundar Pitchai had actively worked on and we have over a billion users today for Gmail and Chrome.

The strength of analytics
The other aspect is our analytics. We look at data very closely and see how products are performing and where the issues are. We have built a very robust analytics background into it. We had a product called Google video. It was a very early version, where people could load some videos and share them but it was very clunky and there was not much adoption at all.
We found a start-up ‘YouTube’ did well in this. We paid what seemed then a huge price to pay to acquire YouTube and now we have billions of views every day, with thousands of hours of videos being uploaded every minute. The analytics informed us that our own product was not working. So we had no fear in acquiring and replacing that.

Fail well
The third, we fail well and fail fast. When we sunset a product, there is this sunk cost fallacy. Teams and users get disrupted, yet we haven’t shied away from doing it. Picasa, the photo sharing app is no longer there but Google photos is there. It took a lot of efforts to transition users from one to the other. We put a lot of time and effort to make sure we don’t hurt the end users but we are not afraid of making some of the bold changes.

If you don’t cannibalize yourself, somebody else will. When I first came to head the Google ads business in India, we noticed there were a lot of advertisers who were not logging into their ads account for a year but were spending every month. This was not giving them any returns at all. So I thought why we should be taking money from them. We contacted them and told them they were spending this money without any benefit. From our side, it was a very risky move because most companies would not even touch their holy cow revenues coming in.

We bet on technical insights. Google’s overarching mission is, to organize the world’s information and make it universally accessible and useful.

Many of our advertisers were very happy that we told them about it and shut the accounts down. Others were very happy that they learned about it and spent more time, optimizing their ads campaign. So we ended up at the same level or better.

Think 10X
The last one, we encourage everybody to think 10X. Think big, think of big problems, think of radical solutions and think about breakthrough technology.

We have a self-driving car. Google is not a car company but there are millions of people who are unable to drive because they are either physically challenged or really old and very constrained. We wanted to solve this problem. We have machine learning and we have a great vision. We asked: can we do this in a scale? Now, the Google cars have driven millions of miles. We have solved some of the hard problems through technology and some of it through analytics. We studied and found out that 53% accidents happen when people make a left turn (in India, when you make a right turn). We had to spend a lot of time on how to optimize that user journey. So we constantly cannibalise, check, iterate and launch.

Companies like Netflix, for example, were prepared to disrupt themselves before someone else came in and disrupted them. What is Google’s approach about innovation? Large companies typically tend to innovate by listening to its existing customers or potential customers. Invariably, the type of innovation that happens tends to be sustaining or incremental, trying to improve an existing product to such a point that it exceeds customer needs. In Google, do you listen to your customers or to your engineers?

Google is first and foremost an engineering driven company. It has a highly engineering culture. The first several hundred hires at Google were all engineers. The DNA of the company is very much engineering driven. A lot of our investments go into R&D. Fundamental engineering work takes up a lot of the investment—be it in the cloud, building the open-source Android platform or creating text-to-speech or speech-to-text technology or AI/ML technology or Quantum Computing. We invest heavily in them. However, there are companies like Amazon that are highly customer focussed. For a company like Google to work, we have a few fundamental things. We invest in building scalable platforms. We bet on technical insights. Google’s overarching mission is, to organize the world’s information and make it universally accessible and useful. It’s a fairly straightforward mission. But that means a lot of things to a lot of people.

The DNA of the company is very much engineering driven. A lot of our investments go into R&D.

The user-led innovation can come from anywhere. There are 400 million people who are hearing impaired. It is a very big disadvantage for them. Not more than about 20% of them can afford to have training, go to a university, learn sign language and get hearing aids. Among the rest of the population, very few people know how to use the sign language and communicate with them.

User-led innovation
We have an engineer by name Dmitri who went deaf at an early age. He worked in the speech research team. We have the Android phone which is available all over the world. We have speech-to-text engine and multi-language support. A group of folks figured out how they can build a product that can help Dmitri. It was not a top-down. We don’t have an innovation officer. A bunch of people organically created and built a prototype. Once it was validated, that team got investments to make this a legitimate product. It’s called Live Transcribe and it’s freely available in 50 languages. It gives deaf people an opportunity to be more empowered.

Now there is a battle for standards; in other words, one who drives the dominant standard in the whole technology ecosystem matters a lot. For example, Blu-ray versus HD DVD—Blu-ray became the dominant player. We have Q-LED from Samsung vs. OLED from LG. Over time, one standard becomes the dominant standard around the world. Yet we see this very interesting divide in the world, where in the rest of the world Android probably dominates 72% and iOS has 28%. In the US, it has flipped around with iOS being 59% and Android about 41%. How does Android manage to fight this battle of standards?

The question isn’t so much about standards because iOS / Apple is a closed ecosystem. Everything is vertically integrated from the chips that they make, all the way to the iOS software. Then we have the Android ecosystem that has 3.5 billion users. We had amazing phones from Palm and Nokia. The Symbian Nokia phone was well ahead of an Android phone. By the way, iOS produces some amazing products. I also use an iPhone though I’m an Android Developer. So I am well aware of both the products.

The fundamental thing we have in Android is we want to make it easy for developers who can write one application and it will run on any phone from any manufacturer, whether it’s a 35$ phone or 1300$ phone. Of course, the performance may be different across those devices.

We had the Symbian and Microsoft Office, and each platform was different. The developers had to build an app for each of those platforms and maintain them. They had to pay for the license. With Android, we have it as open source. Anyone who wants to use it can use it. We have also put Google applications on it. The developers follow API compatibility. This also allows innovation to happen. Many OEMs are developing products based on Android. Thus, Android became really big.

Supporting the OEMs
There is also a healthy competition between all these OEMs because each of them wants to have a better camera or a better screen. Android supports that. The OEMs add their own secret sauce on top of it. So that’s how Android has succeeded. There’s a lot more of work that goes behind it. My team has to work with over 400 OEMs building devices and also chip makers like Qualcomm and Mediatek. It’s a huge ecosystem and it is very complicated trying to get everybody to move in one direction. iPhone has it easy compared to us because of their one unified approach in one thing alone. Every year, we try to corral this ecosystem and make it move forward. It has been a very challenging but interesting journey.

At least four years before the iPad was launched, Microsoft had the technology for a device on which you could write with your fingers and flip pages. I’ve taught a Harvard case based on that. Microsoft chose to market that device as a laptop plus an additional feature, as they had existing relations with Compaq, Fujitsu and others. Apple was not constrained by any such things. Today your grandmother can play Candy Crush on the iPad. How does Google manage to get the cell phone manufacturers and others to think about innovation at the cutting edge rather than simply doing incremental stuff?

In Android ecosystem, innovation has been happening not from within but from outside. The cell phone market is highly commoditised. The margins are very low. So to distinguish their products, the phone makers have a better camera, better user experience or better sharing. It is either hardware or software improvement. A year before, there was a company that had a rollable product. It wasn’t a foldable one. You can roll it in and out. We had to do a lot of changes to the framework and to the UI in order to smoothly transition the home screen as it expanded. It was very difficult and they have big demands on the Android engineering team to support them.

The cell phone market is highly commoditised. The margins are very low. So to distinguish their products, the phone makers have a better camera, better user experience or better sharing.

Chinese OEMs came out with different demands. You use the three-button navigation on your Android phone. But a Chinese OEM had the gesture navigation. You can the flip it back and forth and up and down. I told the OEM when I met them in Hong Kong that such demands would completely break the developer ecosystem. They didn’t listen. They went ahead and launched it in China. We came back with a lot of data that this was working. Now we have adopted this as part of the platform and you have a choice when you set up your phone whether you want to use gesture navigation or the three-button navigation.

Does enough innovation reach the bottom of the pyramid? Are underdeveloped worlds and underdeveloped economies getting the benefit of the developments? This is a question that needs to be addressed.

We have Android phones with a low cost to a very high cost. We have 2 GB RAM phone or a 16 GB RAM. The performance differs. We have invested a lot to make sure the operating system works well in the low end phones too. The other part is the apps. We have studied the use cases to make useful apps work in lower end phones also. For maps and YouTube, we have light weight apps called Maps Go and YouTube Go. We have done a lot of work to make the apps work in the local market. Like I said earlier, if you build for India, you can build for the rest of the world. Though data is cheap in India, we have connectivity issues here. So we have eliminated a lot of dependency on the cloud and created ‘on device-AI/ML’ model.

If you build for India, you can build for the rest of the world. Though data is cheap in India, we have connectivity issues here. So we have eliminated a lot of dependency on the cloud and created ‘on device-AI/ML’ model.

Many users have language and literacy problems and they have difficulty in operating the phone and understanding the messages. If we can use voice to operate, that will be very easier for such people. So we voice-enabled the entire phone. Today, smartphone has become a super computer and people use to make payments and do many other transactions. Security of transaction is therefore very important. So we have made sure with the OEMs that security updates are available to all the users at least once a quarter, so that privacy and security are not limited only to the rich but to everybody.
A bunch of passionate Googlers went out and spent several immersive weeks with the users of the phones to observe and understand how they used them. One of them went to a family who was a daily wage earner. The children were sent to school. The mother and father were not literate. They did not know what the teachers wanted their children to do. We created an app called ‘Bolo.’ When the child reads out the lesson, the app will detect if every word is read out correctly.

We also have an app called Google Lens for text recognition from a camera image. The phone can convert the text to voice. With this, a mother can scan the progress card of her child and the app will convert it to voice, so she can know about the marks scored by her child in every subject. People can use it for a variety of applications like reading a train time table or reading a bus number. We launched a phone called Jio Phone Next. In that, you can launch the camera and walk around the world and experience whatever is written by selecting your preferred language. The phone will read out to you whatever is written.

Tell us about the culture at Google. What makes it tick?

I am reminded of a Peter Drucker’s quote that culture eats strategy for breakfast. I have been with Google for 15+ years. When I joined, we had 8000 people. Now we have 160,000. The company has grown twenty fold. Obviously, culture changes and it evolves. Yet, fundamentally certain things remain the same.

  • We have a fairly unique and democratic hiring process. We don’t look at just deep technical knowledge. We also want to make sure that those who join us are adaptable, have team management and leadership skills. Can I sit with them and make an interesting conversation?
  • We have a panel that looks at these things quite objectively. The recommendation of the panel goes to a hiring committee. It slows down the process but we take the recruitment process very seriously.
  • We also look at D&I.
  • We set stretch goals to our engineering team also. What is important for us in not what you do but how you do it.
  • Team dynamics is very important.
  • By default, we are very open. Our software and promotions are all peer-reviewed. People can nominate themselves for a promotion. It goes through a committee that evaluates their performance. We also have peer feedback.
  • Take action and beg for forgiveness if something goes wrong. Last year, during Covid, one of the updates we had done to the Android had created huge problem and my inbox was flooded with emails. 300 million users could not use their phones. It was quite a disaster and we gracefully recovered within 24 hours. Instead of firing the few people who set the wrong configuration, we had a blameless post mortem to figure out what we learnt from it. We improved the processes. The person who messed it up had, in fact, helped us to fix it. We have open hierarchy.
  • Every week, I have certain non-office hours where any employee can discuss anything that he/she wants to discuss. We expose people to this culture from the time we hire them. We also listen to them as they come with different perspective. We even encourage them to tell us what is wrong with us.

All these make Google’s culture a different one.