Por: Bernard Girard

wakefulchardSoftware and s/w Development

Feb 17, 2014 (7 years and 5 months ago)


El modelo Google

Por: Bernard Girard

Google’s management model

Bernard Girard

Google is a very special

Because of its economic and financial performances

Because of its market share

Because of its very fast growth

no employees in 1998

5000 in 2005

20 000 today

Because of its prospects and chances to keep on

Markets left to penetrate

Search on mobile phones…

Because of its resilience

But also because of its very original methods

A lot of growth to come

A company as a model

The management theory uses innovative companies as models,
examples to copy, imitate

Ford, 1910 : the mass production

General Motors, 1930 : The M organization

Toyota, 1970 : quality and mass production

Citibank, 1970, the «

one bank holding company


The theory is just the conceptualisation of methods and solutions
discovered by innovative managers

Google is the last of these models

Not because of its successes that come from

The context

The situation

But because of the methods it uses in the management of its
people, its products, its markets

The context, the situation

The situation : the Silicon Valley

Its universities

Its financial tools

Its financial competences (they know how to select the
best projects)

Its legal rules

It’s very difficult to impose non
compete clauses

Engineers and ideas flow freely from one company to
the other

The context

of the late nineties

The internet bubble

Engineers are cheap

The rapid growth of internet

What makes Google so

Its technologies

The algorithm

The «



Its economic model



Its governance

Its management methods


In product development

In marketing

In managing innovation

An algorithm for the search

Search on computers used two methods

Indexation : people read the documents, abstract and index them

Efficient but expensive : a lot of handwork, delays…

Automatic search : the computer selects the documents with the
words selected by the searcher

Fast, inexpensive but inefficient when too many documents

If the machine brings back 1 million documents, what do I

Page & Brin find a way to rank pages that makes automatic search

Imitated from the scientific method of ranking papers

The most cited ones should be the best

The documents with most links should be the more

If you want your site to be on the first page, one solution :
make it the best, so that others create links to it

And it works

Innovations in the «



A powerful «


» : 1 million computers in 60 «



When they started, Brin & Page were not rich

They had to build their computer farm with PCs and
cheap used computers

They chooed Open Source softwares that are free and
give a better control on home
made applications

Everyone knows PCs don’t last very long, are not very

They has to build an ad hoc system

Very redundant

That duplicates data several times

With a software that copies automatically the data
from one PC to the other

They have built a reliable system that grows incrementally

An economic model

When Page & Brin started Google, the search engines where financed
by advertising

Free search for everyone, money comes from advertisers

But too much advertising is a «

pain in the neck

», that slows
search and harms the reputation of the engines

They tried to find other ways of financing before finding their own way

Advertisements that are useful and almost invisible

The way these ads are priced

A price per click

A price defined by a double envelope auction

An automatic customer relationship

No need to sell the ads

No need to negociate the prices

The long tail : a lot of small advertisers that make a lot of
revenues for the company

Useful and invisible ads

Price : a price per click

The advertiser pays only when users click on their

He pays only useful impressions of his ads
whereas in the press (and on the other search
engines) he pays all impressions

Which could help Google grow in a recession :
advertisers just want ads that bring sales

If ads don’t bring sales, it’s not Google’s fault, it’s
because the advertiser does not transform clicks
into sales

In a way, Googe brings customers in the shop

An automatic sales system

When they started Brin & Page had not ressources to
hire a salesforce, they invented a system where sales
are made without salesmen

The buyer does everything

Chooses the keywords of his ads

Writes the text of the ad

Decides the price per click on a keyword

Chooses his daily budget

Google gives advertisers tools to improve their ads

Statistic tools (Analytics) to convert ads in clicks

The best ads bring the more sale to the advertiser
and the more revenues to Google

This automatic sales system
brings Google new
customers : small shops… that didn’t do advertising

The long tail

Price : the double envelope

But how Google chooses the ads it prints ont its pages?

What happens when several advertisers want the same

Their solution : an auction system

The sealed
bid second price auction

All bidders give their best price in sealed envelopes

The winner is the highest bidder

He pays the second highest price (and not his own price)

The bidder is encouraged to give his best price vs the english ou
dutch auctions)

The price is fair (it’s the market’s price)

The buyer can trust the transaction

Pricing is automatic thanks to the auction system

Everything can be automatized

No need for negociations, transactions costs are reduced to
almost nothing

Innovations in management

Google innovated in


Human Resources

The hiring process

The 20% rule

Reputation as a motivation driver

A very special working environment

Customer relationship

Product management


Page, Brin, Schmidt


Usually, there is one head, the boss who knows better and decides for everyone
(or tries to)

At Google, they have a tripod : Schmidt, Page, Brin… three leaders who work

Which opens up the cognitive capacities of the leaders

Instead of one brain, three brains…

Which improves the quality of the information they use to take decisions

Which gives them the opportunity to look at the business from different
points of view and not only from the financial one

Which gives them more freedom

vis the shareholders

vis the high management

This governance mode helps create a company with less layers of management
and bureaucracy

It’s not totally new (many start
ups have the same structure) but they choosed to
keep it active

Human Resources

The hiring process

Reputation as a motivation driver

The 20% rule

A very special working environment

Hiring the best

All companies say they want to to hire the best

Most don’t really do that, Google really tries

It affects a lot of resources to the recruitment process

It’s hiring methods are non conventional

They don’t wait for the best people to come to them, they
look for them

They reverse the usual way of hiring

Diplomas are not a proof of your technical abilities
but a proof or your personal abilities

You prefer future to present (you invested in
education and stayed long years in school)

You are able to solve difficult problems

Tests and meetings are used to discover your
technical capabilities

Reputation as a motivational

How to motivate workers is a question all managers have to deal with

Two classical answers

The external drivers : salary (commercial companies), prestige

The internal drivers : pleasure to do, to achieve an objective…

In the US, the usual answer is money, index salaries to performance

Google taught there is another driver : reputation

What motivates people is the desire to acquire and keep a good
reputation among co

It’s a bit like in the academic world : scientists don’t run for the
money but for the reputation their work can bring them

This motivation friver is embedded in Google in two ways

The selection of the brightest

Working with the brightest gives the impetus to work a lot : one
wants to be considered bright and capable by those one
considers so well

The peer reviews : rather than being judged by hierarchy, projects
are judged by colleagues

The 20% rule

Each engineer can spend 20% of his working time on personal projects

An unconventional but efficient way of managing people

When one has an idea, he has time to work on it, even if it’s not in
the company’s priorities

In traditional companies, these ideas are abandoned or
their inventors go elsewhere to develop them

To find these 20% time you mus rush on your official job : it’s good
for productivity

Procrastination is no longer a loss of time which is also good for

Several Google products come from this 20% rule

Googlenews, cloud computing…

Free food…

A very special working

The popular press emphasized the massages, the free
food, the swimming pool…

It’s important

To build the good image of Google : it’s a place
where it’s nice to work

But, it is also a place where you can find the best
conditions to work

The techshops : any problem with your network? Your
computer? Specialists solve them quickly

Engineers don’t have, like in other companies, to
work on these trivia!

They can concentrate on satisfying users

A new personage, the user

In all companies, the


customer is the king

»… but is he

On the web, the user is the main personage of the play

He is the one who creates the contents search engine

Without users and their contents, no… interesting results

And Google understood that better than others


Take care of the users all else will follow

» (L. Page)


First make sure our users are satisfied, we’ll always
find ways to monetize our invention

» (LP)

The user helps the company in a kind of renewed potlatch

A modern gift economy

The user benefits from Google’s free products, but, in

He brings contents on his websites, on his blogs…

He gives Google informations

On his behaviors, his preferences that statistical
analysis can analyze

Metrology is a marketing tool

He tests the products delivered in early phases of
development (beta phase)

He promotes them through blogs and forums…

He develops them when he participates in a mashup…

An application born in a mashup

From product management to innovation

Product management : how to
solve the complexity problem

tech companies
meet the complexity

At Microsoft : several
years between two
major releases of a

Which slows innovation

Google solved that
problem with the swiss
knife approach

The Swiss Knife

With the swiss knife every
innovation, every tool is
independent vs the


» à la Microsoft

Google can easily add new
products, new tools : it’s not
limited by the obligation to
respect the past applications

Any of these products can
change quickly without any
impact on the other tools

It’s of course possible to mix
these tools to build new ones

In these cases innovation
comes from integration of
different applications

Release early, release often…

This approach allows to launch innovations as they come

Once a product is «


», it is released

Very high rate if innovation : almost not a week without
something new

The method gives a very quick feed
back from the users,
it forces to listen to the users

But the products are «

work in progress

», they are not
perfect, they are… perfectible

Users are often disappointed but they keep quiet : these
products if mediocre are… free and they know they will

This solution is possible only because the applications are
free and there are no distribution costs

Products with an original life

They improve incrementally and stay for long periods in
beta phase

Which gives the users the opportunity to give the
company feedback on their uses

They have a long life. There is no obsolescence strategy as
at Microsoft (Vista vs XP) or Apple

They slowly diverge from the competition :

At first Googledocs was a terrible «

me too

» Office

It is, today, a suite of office automation applications
for the web and collaborative work

With special functions that you can only find on the

And appear to be, in the end, competitive

Today Microsoft imitates Googledocs :)

Search in a spreadsheet

Innovator, but not like the others

Microsoft, Apple, Sun, HP… are also innovators, but the
flavor at Google is different

At Microsoft, innovating is the best way to build

Which is the classical strategy : patents give their
owners a monopoly on their invention

Apple applies Hollywood’s theory of creating
blockbusters to industrial innovation :

A star (Steve Jobs),

Excellent products (although usually not disruptive)

Formidable advertising campaigns

Google is an explorer that applies the scientific
experimental method to industrial innovation (hence the
word «


» in Schmidt’s sentence)

The experimental method

In the experimental method, researchers propose
hypotheses and design ways and means to validate them

Eric Schmidt : «

We try this, we try that, we see what


This logic of «

trying and trying

» is embedded in the

In its management,

In an organisation that was designed to make plain use
of this method

In the management of its products

Two examples

The peer review and the management of innovations

The swiss knife and innovation

The peer review, innovation,
productivity and quality

Thanks to the peer reviews…

Ideas flow between teams

The colleagues that judge your projects work on
other topics

OCR : the fight against spam and Google Book

If you want your colleagues to judge your work you’d better
make it short and readable

Everyone speaks the same programming language

Source code documentation is a must

Projects are short

Small projects = small teams = less layers of management

Innovation by integration

Mobilize all creative ressources

But the experimental is not the only novel way of doing
things at Google, the company is also a master at
mobilizing creative ressources

wherever they are

Its engineers and their ideas (the 20% rule), its users


The Open Source Community and its developers

Google code and project hosting

The archives of IT

The Tesseract story and how it became OCRopus

Competitors : rather than reinventing their applications,
Google buys young companies

Users as creators of contents and as developers

Google gives them tools (API, Sketchup…)

Which help fight the «

Not Ivented Here Syndrom

» : Google
does not reinvent the wheel every morning as others do

Thanks to this method…

Google regularly discovers new domains that could become
very lucrative new businesses


Collaborative office automation

Cloud computing

But also new frontiers that could limit its markets and slow
its growth

Technical : translation

Good translations could open the web, but where are
the engines that produce them?

Legal : Copyright laws

Everyone would like freely available rich contents,
but copyright laws are an obstacle in video, music
and books

Economic : innovations must be monetized

And it’s not always easy to find a way to do it (Youtube)


No preview available…


How users helped overcome the
copyright obstacle

Because of copyright laws, big search engines
cannot offer movies or music

The solution : give the users the means to
create and publish their own content

Create our own contents? we can all do
that with our cameras

Publish? We can do that with Youtube and
we do it!

368 000 movies on Obama!

Innovation as a strategic tool

I am often asked : what is Google’s strategy?

The answer : innovation

Fast innovation stifles, asphyxiates the competition and
make sure all roads lead to… Google

Search, gmail, picasa, blogger, chrome, google books,
docs, notes… Google gives almost all the tools you
need on the web

Thanks to the beta model and the continuous
improvement of products, the competitors are uncertain
as what will be their next move,

They run after Google

This same model make allies of users and customers

They help develop new products, they are the best


», they give the best informations on what
the products should look like…

And they are, most importantly, loyal!

Questions on this model

Is it sustainable?

Is it for everyone?

Is it sustainable?

Is what is good for a small and young company
appropriate for a mature and bigger company?

How long can Google deliver solutions it cannot

Can this model be applied to products that are sold
rather than given?

Is the ad market big enough to finance all th web?

Is reputation so good a motivaton driver when stock
options dive

A model for every one?

This powerful model helped Google become what it is today

Following all Google’s rules is not a recipe to become a new Google

But most of these new ways of managing will in the years to come
enter the catalogue of business’s best practices

Somes companies already started emulating them

In the hight tech sector

Most start
up imitate Google’s ways

But also in the old industry

Chrysler and the tripod as a mode of governance

Even in very different businesses

The Restaurant and the 20% rule

The consulting company in Bucaramanga, Colombia

A model for you?

You tell me…