What It Takes to Keep the Internet Online

Your next video will start in

Recommended Videos

  • Info

  • Comments


Jan. 23 (Bloomberg) -- Michael "Mike" Abbott, general partner at Kleiner Perkins Caufield and Byers, discusses the tech industry with Emily Chang on Bloomberg Television's "Bloomberg West." (Source: Bloomberg)

Country's internet goes down, what happens on the back end?

What are engineers actually doing?

First of all, thanks for having me, emily.

I think the first thing is depending on who the engineers are, they may be freaking out and saying, what is actually going on?


I know actually when i first joined twitter the site would go down and we didn't always have a lot of visibility and asked why is it down?

How would you find out why it's down?

You need to provide the right instrumentation and people that collect the data to understand what's going on.

What's the overall statistics on the environment?

In this case, i don't have enough details to understand exactly what they have or not but typically going out and really seeing, looking at what kind of visibility can we get into the kind of current state of the system so we can root cause, what's actually gone, how do we get out of that state and furthermore, how do we do a postmortem so it doesn't happen again.

Put it in context a bit, when you were at twitter, what did you do to get out of the farewell?

It was bringing in a lot of great people and augmenting the team that was there and would be the first and foremost, getting folks from google and other sites ahead, seeing that type of scale and many of the engineers as well as the engineers over there focused on getting the visibility and instrumentation into the overall infrastructure to understand what was going on.

Basically getting eyes and ears out there.

And also sharing information between teams.

I know when i joined and the site was down there was only maybe a handful of engineers who understood the complexities of the system.

How can we decompose that information and get it to others so we can have a collective team to root cause the issue and solve it versus heroics by a handful of folks.

We all know engineers are in demand and that there are not enough good engineers.

What makes a good engineer?

What can a good engineer do that a mediocre engineer cannot do?

It's a great question.

And it's hard to answer that question very specifically because it depends on what you're trying to build.

But i think great engineers want to work with other great engineers, he or she, they want to make an impact on whatever they're working on.

And they are working on an interesting problem and having a great impact and working with great people, they want to work at your company.

Do good engineers drive innovation in the sense that, you know, are they given a problem to solve or do good engineers come up with the problem to solve?

Do you get my point?

I do.

I think great engineers come up with solutions to address a problem.

And i think in many great product companies, the problems, if you will, are outlined by the design team or maybe a product manager.

I think oftentimes, at least in today's world, it's the design team.

The design team may have a vision for a particular interaction and particular product and it's up to engineering to think through, how do i come up with realizing that vision or dream?

Now, there has to be a path, too, for engineers to be able to provide input to that innovation.

One of the things we started at twitter to drive that was basically how do we carve out time for engineers to come up with ideas, not of which actually made it into the product but had an opportunity, had the knowledge for that innovation.

I think great engineers love computer science and love deep problems and want to be able to have that quiet in silence, be able to work on the problems without dealing with a lot of other -- speaking of the relationship between engineers, product, design, at twitter just last week, the head of product left, we don't know exactly why but he said in a blog post-it was time to leave but there's been a lot of speculation that twitter has been a very difficult company to ship new product at, the twitter we all sort of love and know has been pretty much the same for many, many years and been difficult to make any changes for whatever reason, maybe users don't want it or maybe other forces in the company don't want it.

What's your take on what happened there?

Well, i don't have a lot of the details but what i can say in most companies there's a center of gravity that exists, either in engineering, design, or product, and depending on where that center of gravity resides, there's a lot of other implications for the company.

And i think with twitter the center of gravity has always been in product.

Certainly jack is a product person, ed is a product person, biz and jason goldman.

It always sat there and it's difficult when you have a product so focused on simplicity when you have kind of this push to add new things.

So we had this balance, i think a constant tension of how do we keep the service really simple, yet think thoughtfully how to extend it in ways that are meaningful for users.

As an investors, you're working very closely with your portfolio companies.

Where should the center of gravity be and is it different in a startup versus a company as big as twitter now that has to make money?

I don't think there's one answer where it should sit.

I think oftentimes in consumer companies, the -- really needs to sit probably closer to the design and product because that's really what the main consumers love.

Then you have people -- the revenue team saying but we need to make money.

I'm a big believer if you've got a great product and great service there's a lot of different ways to extract economics from the service.

And i mean, the same way at twitter we rolled out and promoted tweets.

That was done in a very deliberate, slow way.

A lot of that technology was done way in advance of it actually rolling out but we wanted to be really careful about with not disturbing the user experience at the expense of just generating some revenue.

Having that discipline is really important.

I feel like on the enterprise side, ooven times you look at early stage companies, that gravity is in engineering because they're targeting an oftentimes known, not always known, but problem in the enterprise and they're coming up with a different type of solution.

And i think innovation varies between consumer enterprise.

In enterprise you know the problem and the project manager in the enterprise company is representing the customer

This text has been automatically generated. It may not be 100% accurate.


BTV Channel Finder


ZIP is required for U.S. locations

Bloomberg Television in   change