back to article Inside Google, data drives 'everything'

Newly-appointed Google evangelist Don Dodge has confirmed that inside the Mountain View Chocolate Factory, decisions - big and small - are driven almost entirely by data analysis. "Everything is data driven," reads a recent blog post from Dodge, the former Microsoft Startup Whisperer who entered the Chocolate Factory last …


This topic is closed for new posts.
  1. Steven Knox

    Or not.

    "Earlier this month, Google said it had allowed ad blockers onto its Chrome browser because of an idealistic belief that most netizens will end up convincing themselves that online advertising is nothing but a good thing. But you can bet the decision was based on reams of data indicating that most people are too lazy to take the time or the effort needed to install an ad blocker."

    Or, they calculated that the amount of data they'll get from Chrome users using ad blockers is more valuable than the ad revenue they won't lose anyway because those users would be using ad blockers on whatever browser they currently prefer to Chrome because of the ability to use ad blockers.

    "I think 2010 will be the year that enterprises of all sizes start their transition to Gmail and Google Apps, and take their first steps towards the vision of the future. The move towards Cloud Computing is obvious," he says.

    Not until I get a Cloud Computing that I trust is able and willing to 1) give me 100% accessibility to my data, and 2) give everyone else 0% accessibility to my data, thank you very much. Google fails on both counts.

  2. Martin Owens


    >> Compare that to buying software licenses and maintenance from the old style software giants, and add the costs of server hardware, and IT managers to run them."

    Or you could just run some Enterprise FOSS and hire or train internal staff to take care of it. Depends on the scale of the operation and depends on weather you trust this particular company. I'd be much more willing to believe them if any small company could start their own Apps cloud where data from one could be migrated to another, should unforeseen things happen.

  3. Anonymous Coward

    In soviet Russia... drives you!

    Had to be said, sorry.

  4. marklittlewood

    Ex-smokers can be the most anti-smoking

    Don Dodge has always struck me as a pretty decent guy. and a huge advocate for everything Microsoft. I was sad for him when he left Microsoft and delighted for him when he ended up at Google.The transformation from Microsofty to Googlite was swift. I think it is great that he is such an incredible advocate for whatever business he is with.

    I was amused that he was so excited about Cloud Computing though:

    1. Crazy Operations Guy


      I find these to be the WORST people to have in a company. They will see everything though rose-colored glasses, and will always cause whatever they are in control of to stagnate. They are typically afraid to make a decision at all, unless that decision is to suck up to their superiors. Companies need someone that has the guts to say "We suck at this, maybe we should get someone that knows how to do this?", to have the brain power to identify problems, not just ignore them and continue to assume that your employer's bowel movements don't stink.

      People like him are the reason that Sun failed, their thought process tends to go like this "Hey, we are losing market share, in order to fix this we should do exactly like we did last year."

      Or he may be the kind of guy that will follow every stupid idea that comes out of the feeble minds of upper management, the kind of guy tat will agree with the boss that it would be a good idea to replace every desktop with laptops because the CEO loves his laptop...

  5. Anonymous Coward
    Thumb Up

    Don Dodge

    I just wanted to say he has an incredible name.

  6. Brett Brennan 1

    Google has discovered the Holy Grail

    As a data warehouse guy for the past 20+ years, what Don is preaching is something I have personally observed in a number of companies that have built enterprise data warehouses and drive the use of the data in their decisions. Not all of them have remained true to the "cause": mid- and upper-level managers usually derail the data-driven process when they see it as a threat to their own viability.

    And it's not JUST the data. Interpretation and context are crucial to using data - especially detail data - to drive valid decisions. It's tough to build and maintain an accurate business model that actually uses data correctly. The maintenance is really the tough part: as your business changes, you need to revise or replace large chunks of the enterprise model and the way the departmental components integrate into the whole. This effort is very large, complex, and can become unfocused due to the intrinsic disbelief of MBA types - both in not believing that the data is telling the (counter-intuitive) truth, but also because the cost of this dedicated maintenance is not cheap. But cutting corners corrupts the entire process...and there are corpses of companies along the EDW road that failed to do their analysis or cut too many corners.

    I wish Google the greatest of luck with this process. I hope they can keep it working for as many years as IBM.

    1. Anonymous Coward

      Holy Grail - You are missing the point

      Google data analysis is _NOT_ your granny's data analysis. It is not data warehousing, enormous queries, etc. It is probability and stats and specifically conditional probability. It may work with enormous amounts of data, but it remains grounded in mathematical reality.

      This is a big difference to the average Co where the average product person is running mindboggling queries over the warehouse while being barely able to add, subtract and divide. That is actually the less common case. The more common case is when the product person cannot be bothered to either run queries or add subtract and divide. So they chose to invoke the "dark forces of brand awareness and marketing" instead.

      It is also a big difference to most of Google direct competitors which do not grok probability and everything has to be engineered in a deterministic fashion which in turn simply cannot use the Googletastic amounts of data.

      There is no contest in marketing vs math and engineering vs math. Math wins hands down in both cases. This is especially the case for anything Internet. The Internet is a naturally probabilistic environment. Google operates on it natively, while its competitors continue trying to engineer the probability out of it (and fail miserably in the process).

      The Math side of Google is an advantage that other companies in the market cannot match. Not like they are trying to either...

    2. Ken Hagan Gold badge

      Re: the Holy Grail

      When you say "discovered the Holy Grail", are you simply observing that Google has finally caught up with something that the scientifically inclined have known for several hundred years, and what well-managed businesses have known for decades, or are you crediting Google with some actual achievement?

      You're certainly correct that it is not just the data. You have to decide which data to collect, and if you get that wrong then your conclusions will be wrong. (Google in, Google out.) I was tempted to quote Donald Rumsfeld here about unknown unknowns, but I think it is better to splice a couple of Richard Feynmann's: "Science is the process of learning not to fool ourselves, for Nature cannot be fooled.".

  7. RW

    Those massages....

    Do they ever lead to a happy ending?

  8. ForthIsNotDead
    Thumb Down

    Oh dear...

    "I think 2010 will be the year that enterprises of all sizes start their transition to Gmail and Google Apps, and take their first steps towards the vision of the future. The move towards Cloud Computing is obvious," he says.

    Well, he would say that, wouldn't he, working for Google and all! Yeah right, like it was obvious 10 years ago that Thin Clients were the 'future'.

    Can we have a tumbleweed icon?

  9. Ian Michael Gumby

    No duh!

    "He does not point out that Gmail and Google Apps are also a means of collecting your data."

    Wow, I guess it is important to state the obvious. Google is a capitalist company. TANSTAAFL so if they offer a "free" service, then they are getting something from it.

    Your e-mails are theirs to use as they like. ;-)

    "Do no evil" is starting to sound like Bill Clinton when he was under oath...

  10. Neal 5

    Yes, but,

    All the data says is that fish can breathe in water, ergo fish being a lower life form, there is absolutely no reason at all why either Dodge or Norvig, or preferably both of them at the same time, can't.

    Seeing as no model, not even a subjective one is needed, I fully expect Google's Norvig and Dodge to peither prove this one way or another, then I might just believe all data and discount all models, subjective or otherwise.

    Christmas day would be soon enough for me, I need to plan ahead, and can't go through the better part of next year waiting for a result to that very simple theory, so Google, either prove it true or (fuck I've been warned about expletives I can't finish this without offended some Google droid or government official, oh bugger)

    Note to self for new year, must learn some idiot tolerance.

  11. Tim Roberts 1

    it had to be said ....

    Says Don Dodge "Google really likes data"

    "any less is just a non starta"

    "I'm now way to lofty"

    "to be microsofty"

    "thank Christ my desk's not near a farta"

  12. Mage Silver badge

    Costs? 500 a year? No.

    My server for email, files, printing is Running Windows 2000, MDaemon, Apache, IIS, MYSQL, SQL7.0. Since install it has cost nothing other than electricity.

    It serves 6 people.

    The PCs and laptops run a mix of Ubuntu, Win98 (programming set-boxes/mobile radio), Win2k, XP, MS Office Apps, Open Office as well as all our specialist applications.

    2001 was the last time we bought MS apps.

    5 PCs with XP, two dual boot with Ubuntu

    2 Netbooks (Linux)

    1 Laptop multiboot DOS/Win3.11, Win98, Win2K and Ubuntu

    3 laptop XP

    1 Laptop Ubuntu

    1 Laptop Dual Boot XP and Ubuntu

    2 Routers / WiFi (Linux)

    1 Win2K server

    Instead of 3,000 (6 users) or 6,500 (13 computers) the annual SW budget is probably mostly on Games for Nintendto DS, Sony PSP and PS2.

    About €60 p.a. on Domain Names and < €100 on hosting to run the websites (with MySQL etc).

    Unless you have redundant fibre and no cap you can't consider Cloud. Even then it's complementary not a replacement.

    I only use gmail and Google Docs for those other online people that insist on collaborating via it. I can do all of that on my own secured hosting without Google reading it.

    For really private collaboration we have VPN and our own server.

  13. Anonymous Coward
    Anonymous Coward

    Almost entirely?

    "decisions - big and small - are driven almost entirely by data analysis"

    Hopefully the rest are driven by legal requirements and a code of ethics. Or we're stuffed.

  14. Anonymous Coward

    Little detail..

    AFAIK the economic crisis was also entirely built on data models. Unless you very accurately know what you're doing with data you can spectacularly screw up as well - small mistakes make big differences in large models (cascade failures, my favourite, but then I always liked domino).

    Sample: Google has missed the fact that privacy is slowly making a comeback now the incentive to keep people at war has disappeared (the money has gone and political leadership is changing).

    If I then also see that Google is trying to predict human behaviour I suspect that someone needs to be told that (a) Hari Seldon was a FICTIONAL figure (Isaac Asimov Foundation series) and (b) he had BENIGN motives - Google just makes money and doesn't even do anything (and I mean ANYthing) charitable with it AFAIK (I'd be delighted to be corrected on the latter).

  15. Anonymous Coward
    Anonymous Coward

    Gmail ain't that good

    Just converted to Gmail and it is awful. The "threaded conversations" mean you lose mail all the time. You can't sort anything only search. No such thing as a nested mail list. People having to re-create lists with dozens or hundreds of names in it. No address book to look up, nothing that gives any organisation info. Overall a VERY limited experience and simply because they couldn't programme them. Ajax etc would support ALL of the above features - but they don't...

    Oh and I switched to Bing after Schwartz called me a miscreant.

  16. amanfromMars 1 Silver badge

    Flip the coin and on the other side ......

    "decisions - big and small - are driven almost entirely by data analysis" Feed in and/or analyse false data delivers false conclusions, with the likes of a Google being responsible for servering trash.

  17. Field Marshal Von Krakenfart
    Big Brother

    There, fixed it for you

    Inside Google, profit drives 'everything'

  18. Steven Jones

    Where data driven can get you...

    Of course it wasn't Google that invented data driven analysis. The investment banks got into this in a big, big way and the related mathematical modelling. Hence we ended up with a financial industry like a pack of cards joined together by exquisite mathematical formula that wholly missed the basic systemic faults in the sytem. This wonderful data analysis, and the vast amounts of data processed to produce guaranteed returns based on the management of risk turned out to have this fatal flaw - that there were underlying common factors which would bring down the whole edifice of commercial banking.

    I'm very data driven, but there are some things where an analysis of the numbers without understanding mechanisms can lead you to wholly erroneous results.

  19. MinionZero
    Big Brother

    @"Do no evil"

    This do no evil sales pitch phrase is nothing but a manipulative two faced sales pitch that has always sounded suspect. These days, the more I hear it, the worse it sounds. For me Google showed their true colors when Google CEO Eric Schmidt said only miscreants worry about net privacy. e.g.

    There has always been very good reasons for privacy. Not least to avoid the state having so much data on peoples lives that government meddling busybodies (and outright control freaks) can make peoples lives an utter nightmare with relentless exploitation of people and interference with their lives.

    Google's greed for information, in other words, their goal of "organizing the world's data" is indistinguishable from the creation of Big Brother. Google want to be on and in everything. For example, Google are already working GPS location based mobile phone searching that tracks our position via GPS. They will even know what our voice sounds like, from their Google voice recognition research. Both of these technologies are already working now. Google has already demonstrated both of them.

    So considering where we are going even short term, imagine how much further Google can go long term with moves towards ever smarter and more specific data mining, plus being tied into ever more of our lives.

    Big Brother is evil not because it knows everything, but because of the power it gives a minority of people at the expense of everyone else. Knowledge is power, but only a source of power for people who have access to the knowledge. Its the asymmetry of Big Brother total information access that gives so much power. Google are on course to become the most powerful company in history.

  20. mhenriday

    Data driven ?

    What would Mr Metz suggest instead - faith-based analysis ? Is it not the case that decisions ought to be based on the data available at the time they are taken, with reservations made based on an analysis of their reliability and applicability ? Contrary to Mr Metz, I find that Gmail's threaded conversations plus search greatly facilitate retrieving information, but perhaps the system doesn't work for everyone, in which case dissatisfied users can easily choose a alternative webmail provider. As far as I know, those who use Google as their default search engine are not require to use Gmail, nor vice-versa, unlike some Windows Live Mail users who have lately received messages purportedly from Microsoft informing them that if they don't use Bing to search, their WLM accounts will be terminated....


  21. disgruntled yank


    "Decisions are made based on data that has been analyzed, and going forward everything is monitored and measured based on the data. Opinions and 'gut feel' are considered too, but in the end, results are measured by data. "

    1. "on data that has been analyzed". Breathtaking concept. Better, I'll admit, than "data that has been synthesized", i.e. made up.

    2. If you say "going forward" shouldn't you say "at the end of the day" rather than "in the end"?

    3. How about

    "We base our decisions on rigorous analysis of data collected in quantities our competitors cannot match. We prefer to measure our results than to guess what they might be."

  22. Anonymous Coward

    Can't we all just Google

    "Why is Google the most important intelligence collection system ?"

    If quite a few people do this, their automated data-analysis will surely make them a red face. I think that's worth it.

  23. Anonymous Coward

    Why is Google the most important intelligence collection system ?

    Open source intelligence (OSINT) is a form of intelligence collection management that involves finding, selecting, and acquiring information from publicly available sources and analyzing it to produce actionable intelligence. In the intelligence community (IC), the term "open" refers to overt, publicly available sources (as opposed to covert or classified sources); it is not related to open-source software or public intelligence.

This topic is closed for new posts.

Other stories you might like