English subtitrări

← 2012 Korea Webmaster Conference

Obține codul încorporat
13 Languages

Showing Revision 1 created 08/16/2012 by Amara Bot.

  2. Good morning.
  3. It's very good to see
    everybody today.
  4. I'm really excited to talk to
    you a little bit about Search,
  5. and talk to you a little bit
    about Google, the evolution of
  6. Search, how Search works today,
    and the future of
  7. Search, and then also talk
    a little bit about
  8. openness and the web.
  9. So who am I?
  10. Why am I here?
  11. Why am I in front of you?
  12. I started out working on Search,
    as you heard earlier
  13. today, 12 years ago tomorrow.
  14. So I have one more day
    left before I'll be
  15. around for 12 years.
  16. I get to answer lot of questions
    from webmasters,
  17. people who want to know how
    search engines work, how to do
  18. better in search engines.
  19. And we've made a total of over
    400 different videos that have
  20. been viewed over 6.6
    million times.
  21. So, a lot of people want to know
    how search engines work,
  22. how to do better on search
    engines, and we're excited to
  23. talk about that today.
  24. I also have a lot of
    ties with Korea.
  25. For example, the phone that I
    carry with me every single day
  26. is Galaxy Nexus.
  27. It's a fantastic phone.
  28. I actually have two Samsung
    phones in my pocket right now.
  29. One's a the local phone and
    one's for the United States.
  30. And in fact, the very first car
    that I ever owned was a
  31. Hyundai Excel, a
    blue hatchback.
  32. I drove it all the way
    through college.
  33. It was a incredibly car.
  34. I drove it all the way
    into grad school.
  35. I have extremely fond memories
    of my Hyundai Excel.
  36. Now I know that Hyundai has made
    huge strides and has made
  37. even better--
  38. you know, like the Equus.
  39. Right?
  40. Now, Hyundai is a
    luxury brand.
  41. But I have been extremely
    excited and proud to use
  42. Korean products for years
    and years and years.
  43. So what am I going to
    talk about today?
  44. I'd like to talk about the
    evolution of Search, that is,
  45. the past. How Search got
    to the way it is.
  46. I also want to talk about
    Search today.
  47. It's important to know how
    search engines work.
  48. If people are interested in
    search engine optimization or
  49. knowing the process that Google
    goes through in order
  50. to decide whether to launch
    a change or not.
  51. And then I'd like to talk
    a little bit about
  52. the future of Search.
  53. What would Search look like in
    a perfect world, if you could
  54. control everything about it and
    make sure that you had the
  55. ideal perfect search engine?
  56. And then I wanted to close out
    by talking a little bit about
  57. the importance of the open web
    standards and how important is
  58. to make sure that the web is
    open so that everybody can
  59. benefit as a result.
  60. So let's start with the early
    days of search engines.
  61. One of the very first well-known
    search engines
  62. globally was Yahoo.
  63. And it's almost a little bit of
    a strangeness, a little bit
  64. unusual to call it a search
    engine, because Yahoo started
  65. out as a hand-compiled
    list of links.
  66. So an individual person would
    decide what category
  67. to put things in.
  68. They would decide whether it
    deserved to be a certain
  69. category or not.
  70. The problem with that is that
    it doesn't scale very well.
  71. You need to find a search engine
    that can work across
  72. the breadth of the entire web,
    or else it isn't going to be
  73. as useful for every kind of
    query that people get.
  74. And so the next generation of
    search engines looked at the
  75. content on the page, that is,
    the actual words that existed
  76. on the page.
  77. Now, whenever I joined Google,
    we were a start-up, so there
  78. was less than 100 people,
    whenever I joined Google.
  79. And at the time, I was
    worried that we would
  80. be crushed by Altavista.
  81. Google was a little
    tiny company,
  82. Altavista was a huge company.
  83. But Google has something that
    the other search engines at
  84. that time did not do.
  85. We looked at the links pointing
    to web pages.
  86. So, not just what a web page
    said about itself, because,
  87. you might meet someone and
    they say that they're
  88. fantastic, they say
    they're great.
  89. But if you meet someone and
    their friend says that they're
  90. fantastic, or you know someone
    who objectively says that
  91. they're great or fantastic, that
    means even more than if
  92. you say it yourself.
  93. So looking at links was a vital
    new way to discover the
  94. reputation of web pages and
    which web pages should count.
  95. Google also started to look
    at, not just pay for
  96. inclusion, but trying to make
    sure the very best pages rank
  97. in the best possible order.
  98. Back whenever Google started,
    it was relatively common for
  99. people to pay to be indexed
    in their search engine.
  100. And if you think about it,
    that's not a great idea.
  101. A search engine should
    try to be as
  102. comprehensive as possible.
  103. You shouldn't need to pay for
    the search engine to give you
  104. good information.
  105. So Google has also evolved.
  106. From going from ten blue links,
    as we sometimes call
  107. it, the ten blue links referred
    to just the ten
  108. search results and the links
    that would come out of that.
  109. Google has gotten better
    and better.
  110. So now, at least in the United
    States, we're able to return
  111. all of the information blended
    together according to what is
  112. the most relevant.
  113. Now, I know that in Korea we're
    more likely to have
  114. segmented search, and so
    you'll have things in
  115. different sections.
  116. But it can be very useful at
    times to be able to say,
  117. first, let's see a web result.
  118. Then, let's see a video, because
    I've done the search
  119. Korean pop.
  120. And then, maybe a few
    more web results.
  121. So that sort of intelligence to
    figure out when you want to
  122. have a video result or an image
    or something that's a
  123. web result can be extremely
  124. So I'd like to take just a
    second and talk about what the
  125. world would look like
    without Google.
  126. Whenever Google started out,
    search engine quality was not
  127. as good as it today.
  128. So Google was one of the very
    first search engines that
  129. clearly marked advertisements.
  130. I remember whenever I started
    out at Google, I went and I
  131. talked to another company.
  132. And they had a list of results
    that they called featured, and
  133. they had a list that they
    called partnered.
  134. And I said, what's the
    difference between a featured
  135. result and a partner result?
  136. And the company said, there's
    no difference at all.
  137. Everything is paid for.
  138. And that didn't seem
    fair at the time.
  139. So Google did a very good job of
    trying to mark what clearly
  140. what were the ads and what were
    the organic, what were
  141. the editorial results.
  142. And you can't pay to get a
    higher ranking on Google.
  143. I'm proud that even to this day,
    you can't pay to get a
  144. higher ranking on Google.
  145. If there hadn't been Google,
    I think you would also have
  146. found more spam on the web.
  147. I don't know how far this
    wireless mic will work.
  148. Let's see how we can go.
  149. Even if you can read English,
    this might look like nonsense,
  150. like gibberish.
  151. And the fact is,
    it is nonsense.
  152. It is gibberish.
  153. This is something that
    my team works on,
  154. called spam or web spam.
  155. If you see, free sound effects,
  156. Puget Sound naval shipyard.
  157. Maybe someone would type
    in Santa Anita
  158. VIP baseball sounds.
  159. A spammer would make these sorts
    of pages with the hope
  160. that if someone were to type in
    two or three random words
  161. from this page, a person would
    land on this page.
  162. And then you could show ads, or
    you could include malware.
  163. You could basically do something
    that would infect
  164. the user's computer or
    otherwise harm them.
  165. And so spam is a really
    bad thing.
  166. And before Google, there was
    a lot more spam on the web.
  167. Google changed the conversation
    in a lot of ways
  168. to think about the quality of
    the search results and not
  169. having spam.
  170. So I think we would have had
    more of it if Google hadn't
  171. taken strong action on spam.
  172. We definitely would have
    seen a lot more viruses
  173. and malware as well.
  174. I remember one year
    at Christmas, I
  175. went to visit my relatives.
  176. And has anyone had the
    experience where the computer
  177. was running a little bit slowly
    because it got infected
  178. or it had malware,
    spiders, viruses?
  179. Anyone had that sort of
  180. experience with their computers?
  181. Maybe their parents'
  182. Anything along those lines?
  183. A few people.
  184. Absolutely, Well at the time, I
    had just spent an entire day
  185. cleaning out my mother-in-law's
  186. So, out of the vacation--
  187. I had about five days off--
    one entire day was spent
  188. getting her computer
    into good shape.
  189. And I realized that her computer
    was infected with a
  190. company that had been
    a partner of Google.
  191. So I went back to Google after
    the Christmas break and I
  192. said, we have to stop this.
  193. We never, ever want to partner
    with any sort of malware or
  194. negative provider or
    anything like that.
  195. And I'm very proud that Google
    has taken a strong stand.
  196. We try not to show ads.
  197. We try not to work with anyone
    who might be infecting
  198. someone's computer.
  199. We also make Chrome, which is
    not only fast, but protects
  200. users' computers by flagging
    malware and potential spyware
  201. and viruses, and we even mark
    potential hack sites and
  202. spyware in our search results.
  203. One last thing is that, if we
    didn't have Google, I think
  204. people would be a little
    bit slower.
  205. So, I've been in Korea
    for a few days now.
  206. I've gotten to take a tour of
    some palaces, museums, really
  207. got to see a lot
    of the sights.
  208. And I realized that there's
    this culture of
  209. pali pali, of speed.
  210. I was in my hotel and I needed
    to get this shirt cleaned so
  211. that it would be nice
    for everyone so
  212. it wouldn't be dirty.
  213. And the phone in the hotel
    has a button that
  214. says, instant service.
  215. So you just pick it up.
  216. And it was 10:00 at night.
  217. And I said, OK, I need to
    get this shirt clean.
  218. And they were very apologetic.
  219. They said, I'm sorry,
    it's going to take
  220. until tomorrow morning.
  221. And I was like, OK,
    that's fantastic.
  222. I wasn't expected it in
    just a few hours.
  223. South Korea is fantastic
    for speed.
  224. Things get done very quickly.
  225. We care about speed
    at Google, too.
  226. If we can't return the search
    results to you in under 500
  227. milliseconds--
  228. in under half of a second--
  229. we consider that a
    failed search.
  230. In addition, I worked in the
    Ads group for over a year.
  231. Now at Google, typically the
    engineers who work on Search
  232. Quality never talk to the
    engineers who work on Ads.
  233. They even sit in different
  234. It's as if they're almost
    in different companies.
  235. But the Ads group had a
    really good practice.
  236. They said, unless the ads are
    ready, we're not going to make
  237. the users wait to get
    the search results.
  238. So let me say that again.
  239. If your search results have been
    finished, they're ready
  240. to go, but we don't know what
    the ads should be, we're never
  241. going to make you wait
    for the ads.
  242. We go ahead and show the
    search results to you.
  243. And I think that's a really
    good practice.
  244. People are in a hurry.
  245. If you're in a subway on your
    phone and you're doing a
  246. search, you don't want
    to wait for the ads.
  247. You want to get the information
  248. You want to move on.
  249. And so I think that emphasis
    on speed has been something
  250. that we've care about a very
    much amount at Google.
  251. So that's a little bit about the
    history of Search, about
  252. the evolution of Search.
  253. Let me talk about how Search is
    today, how it works, and a
  254. few things that you might want
    to know about search engine
  255. optimization.
  256. The first thing that you need
    to know is that search is
  257. actually very hard.
  258. We have seen well over one
    trillion URLs on the web.
  259. One trillion, so one thousand
    billion different URLs.
  260. The web is huge.
  261. Finding the right information
    is like finding a grain of
  262. sand on the beach.
  263. It's extremely difficult to find
    the right grain of sand.
  264. In addition, over one billion
    searches a day come to Google,
  265. every single day.
  266. If it's a slow day it's still
    well over a billion.
  267. If it's a fast day, it can be
    even much, much higher.
  268. But no matter what, we have so
    many searches coming in that
  269. we have to be able to do it
    as quickly as possible.
  270. And finally, there are people
    who try to cheat.
  271. There are people who try to rank
    higher than they should,
  272. or who try to abuse users'
    trust. And those people make
  273. over a million spam
    pages every hour.
  274. Now the net result of all of
    this is that we have to do as
  275. much as we can with computers.
  276. If you remember, I talked about
    Yahoo, and I talked
  277. about how they compiled a
    list of links by hand.
  278. That approach can never
    scale 100% all the
  279. way up to the web.
  280. So what Google tries to do is,
    it tries to figure out how it
  281. can handle as much of its
    searches with computers.
  282. Because computers
    can work 24/7.
  283. Computers don't get tired.
  284. You don't have to give computers
    the day off.
  285. Computers will run the same
    program the same way every
  286. time, and they don't
    get biased.
  287. They don't have a particular
    point of view.
  288. So, I wanted to talk a little
    bit, given that we do use
  289. computers, about how we
    change the computer
  290. programs that we use.
  291. It turns out it's a pretty
    involved process.
  292. We do a lot of work to try to
    make sure that we return the
  293. best possible search results.
  294. It all starts out
    with an idea.
  295. An engineer has an idea.
  296. Over 20,000 ideas
    last year alone.
  297. Some of them are pretty
    good ideas.
  298. Sometimes they don't
    make sense.
  299. But the next thing you're going
    to do when you have an
  300. idea is you start out
    and you implement
  301. it in a test sandbox.
  302. That is, you test it
    out before you
  303. try it on real users.
  304. When that happens, if it looks
    pretty good, we have an entire
  305. group of thousands of readers.
  306. And we showed them the search
    results before the change and
  307. after the change, and they don't
    know which one is new.
  308. So it's like a blind taste
    test to decide whether a
  309. change is good or not.
  310. If the change still looks good--
    because people tend to
  311. like the new results, even
    though they don't know which
  312. one is the new one, but they
    tend to pick the set of search
  313. results that they like--
  314. then we actually send
    that out to a small
  315. percentage of real users.
  316. So if you have used Google in
    the past, maybe to do deep
  317. research, there is at least a
    chance that we've looked at
  318. the clicks on the search results
    to help us make Google
  319. Search quality better.
  320. Now we try out over 20,000
    Search experiments every year.
  321. But what we end up doing gets
    compiled into a report.
  322. And every week we evaluate
    that report, and then we
  323. decide how many to launch.
  324. Let me just very briefly
    show you some stats.
  325. So these are numbers from 2009,
    but the proportions, the
  326. rough percentages, are
    about the same.
  327. We would try out anywhere from
    10,000 to 20,000 ideas.
  328. Of that, many more thousand,
    8,549, we would send to these
  329. blind side-by-sides.
  330. And then a smaller fraction of
    that actually get sent out to
  331. real users and to see whether
    users tend to click on the
  332. newer results or tend to click
    on the older results.
  333. And the final number changes
    that we launched
  334. last year was 585.
  335. That means we change how
    Google ranks its search
  336. results more than once a day.
  337. Every single day, we're looking
    for ways to improve
  338. how we rank our search results,
    and we never stop
  339. that process.
  340. It's always going on.
  341. Once you work at Google, you
    start to notice anytime you do
  342. a search and you don't find the
    exact result that you're
  343. looking for.
  344. And then you end up
    filing that away.
  345. And you're going to send an
    email and ask how can we make
  346. that search better?
  347. So I wanted to cover in just
    three slides, if you wanted to
  348. rank better in global search
    engines, here are the tips and
  349. the tricks that you would
    want to know.
  350. There's basically only three
    slides of material here, but
  351. if you are comprehending of this
    knowledge, you will know
  352. more than probably 80% of people
    who make websites.
  353. On this slide, I just want
    to cover the basics.
  354. You want to make your site
    crawlable by search engines.
  355. Now, how do you do that?
  356. The first step is just to make
    sure that you give permission
  357. to the search engines.
  358. So there's a very well-known
    standard known as robots.txt.
  359. And if you allow search engines
    to crawl using this
  360. robots.txt, then they
    can enter your site.
  361. Otherwise, it's as if you put up
    a sign that says you're not
  362. allowed to enter.
  363. So the very first step once you
    decide you'd like to be
  364. crawled is to allow search
    engines to crawl in
  365. robots.txt.
  366. The next thing to do you can
    actually do at home or at work
  367. using any web browser.
  368. And the idea is just to click
    on links and make sure that
  369. you can find all the different
    pages of your site.
  370. For example, if you have to do a
    search in a web form, search
  371. engines might not know how
    to do that search.
  372. So the way to get around that
    is to offer links at the
  373. bottom that show you where you
    can click to reach each of the
  374. amounts of information.
  375. So simply by testing things out,
    by clicking links, you
  376. can find all of the pages
    on your site.
  377. That means your site should be
  378. discoverable in search engines.
  379. It should be able
    to be crawled.
  380. The next thing to bear
    in mind is to try to
  381. use standard HTML.
  382. So you saw in the last
    presentation by Junyoung that
  383. if you use Flash, then if you
    have an iPhone, you just can't
  384. see that site, because the
    iPhone doesn't know how to
  385. interpret Flash.
  386. If you can use standard HTML,
    that makes it much easier for
  387. search engines to interpret
    your site.
  388. So for example, most search
    engines won't know what to do
  389. with ActiveX fast or
    asynchronous JavaScript also
  390. known as AJAX.
  391. So if you can use standard HTML
    technology, rather than
  392. Flash or AJAX or ActiveX, or
    at least use less of that.
  393. Use it for pictures in the
    middle, but make sure that the
  394. navigation is standard,
    static HTML links.
  395. That way, every person
    and search engine can
  396. discover your site.
  397. So those are the basics.
  398. Here is just one or two more
    slides about search engine
  399. optimization.
  400. This slide is about the text
    that you put on your site.
  401. You would not believe how many
    pages have a title that just
  402. says Untitled.
  403. Or they don't have
    any title at all.
  404. Or the title of the page is
    exactly the same on every
  405. single page of the site.
  406. Whenever anyone does a search
    and they see the snippet for
  407. your web search result, the
    snippet that they say will be
  408. determined by your title and,
    in many cases, your meta
  409. description.
  410. So in many ways, it's like if
    you were walking by a shop,
  411. and the shop had an inviting
    glass display with lots of
  412. things you can see, versus
    whether it was boarded up.
  413. Now, if you see something that's
    very inviting, like if
  414. users see a title that looks
    interesting and useful to
  415. them, or if they see a meta
    description that says, this is
  416. going to be the information that
    you need, then they're
  417. more likely to click
    through and find
  418. information on your website.
  419. So it makes a really big
    difference to make sure that
  420. you have page titles
    and descriptions.
  421. It's amazing what even that can
    do as far as making sure
  422. that users click through
    to your site.
  423. Junyoung also mentioned that a
    lot of people use pictures of
  424. text rather than the
    text itself.
  425. We've seen this over and over
    again to the point where
  426. Google has joked internally
    that we should try to run
  427. optical character recognition on
    images to try to figure out
  428. what the text is on
    various pages.
  429. But the fact is, if you include
    the text yourself
  430. rather than just pictures or
    pictures of text, then we
  431. actually have words to index.
  432. The last bit of information
    about what text you should put
  433. on your page is to think about
    users and what they will type.
  434. So for example, suppose I wanted
    to know how high is
  435. Namsan Tower.
  436. What would I type?
  437. I would type, how high
    is Namsan Tower.
  438. But somebody else who is making
    the web page might say
  439. Namsan Tower is this high.
  440. And so users will often type
    different things than what the
  441. webmaster will put on
    their web page.
  442. We try to help users out, so
    if you type automobile, and
  443. the web page has car, we'll try
    to return that web page
  444. because we know about synonyms.
    But if you could put
  445. the words on your page directly,
    that makes it work
  446. much better.
  447. Now, I'd like to try
    an experiment.
  448. I don't know if this
    will work.
  449. And I haven't warned the
    translators that this was
  450. going to happen.
  451. So I don't know how well
    this will go off.
  452. But let's give it a try.
  453. In my pocket I have a device.
  454. You guys have probably all
    seen one of these before.
  455. Right?
  456. You put it into your computer.
  457. You can store things on it,
    images, pictures, files.
  458. What would you call
    this device?
  459. Lots of different answers.
  460. I heard USB stick.
  461. What did someone else say?
  462. Flash drive.
  463. Anything else?
  464. Memory stick.
  465. Thumb drive.
  466. Two gigabytes.
  467. There's so many different
    words that people
  468. could use for this.
  469. Now, I don't know how
    it works in Korean.
  470. It might be you all have the
    exact same word for this.
  471. But in English, people could
    type five or ten different
  472. things if they were searching
    for this one product.
  473. So it's a very simple analogy.
  474. If you had a web page and you
    were selling USB sticks, you'd
  475. want to use all of the
    words to describe it.
  476. You'd say, this is a premium
    quality flash drive.
  477. If you haven't seen this thumb
    stick before, it's a
  478. retractable tip, it has two
    gigabytes of storage, and when
  479. you use this flash drive, you'll
    find that it easily
  480. serves up the pictures
    and files.
  481. Now in just three sentences of
    text I've gotten four or five
  482. different synonyms for thumb
    drive, USB stick, flash drive.
  483. I haven't done in a way
    that's artificial.
  484. It's not spamming.
  485. It's helpful because it
    describes all the different
  486. ways of using this particular
  487. It's amazing to me how many
    people will use specialized,
  488. very technical terms, and they
    won't think about what a real
  489. user will type when they're
    looking for information.
  490. If you make something that's
    interesting or useful, people
  491. will want to talk about it.
  492. And so the best way to get
    links, in my experience, is to
  493. come up with something
  494. An interactive feature or some
    kind of research that people
  495. haven't seen before.
  496. Pictures.
  497. The kinds of things that people
  498. want to send to friends.
  499. They want to bookmark.
  500. They want to come back to.
  501. They want to tell
    people about.
  502. Any of that can be compelling
  503. It could be a video.
  504. But whatever it is, that's the
    kind of thing that can cause
  505. people to want to
    link your site.
  506. It's also pretty interesting
    that social media
  507. can amplify a message.
  508. There's a lot of people
    on Facebook.
  509. There's a lot of people on
    Twitter, Sidewall, right?
  510. So you can get the word out
    in lots of different ways
  511. whenever you have
    new material.
  512. It can be extremely useful to
    build up a following, because
  513. if you engage in conversations
    with the people who read your
  514. site, then they're more
    interested and they want to
  515. share your content with
    other people.
  516. So we have seen times where
    someone can do a single tweet,
  517. and from that tweet, tens of
    thousands of people will visit
  518. the site and view the site.
  519. So social media can be a
    wonderful way to help spread
  520. your message.
  521. I'll do a little bit of a plug
    and just mention that Google
  522. provides free tools
    for webmasters.
  523. Google.code.cr/webmasters I
    think we'll be showing a demo
  524. later today of Webmaster
  525. But it's very simple system that
    shows you how fast your
  526. site is, how responsive it is,
    if we know about errors on
  527. your site, if we ever detect
    that your site might have been
  528. hacked or have malware, we'll
    send you a message directly so
  529. that you can be alerted of it.
  530. There's all sorts of great free
    information in Google's
  531. Webmaster Tools.
  532. And we try to make
    that available.
  533. So that's a little bit about
    the past of Search and a
  534. little bit about how
    Search is today.
  535. Let's talk a little bit
    about the future of
  536. Search and what to expect.
  537. Larry Page would like to say
    that the perfect search engine
  538. understands exactly what you
    mean and gives you back
  539. exactly what you want.
  540. So for example, I don't
    speak German.
  541. But what if I had a long layover
    in Germany on the way
  542. back from Korea?
  543. And as a result I wanted to
    go ride in the subway.
  544. Well, I don't speak
    any German.
  545. I don't recognize German.
  546. I don't read German.
  547. But wouldn't it be great if I
    could just point my phone at
  548. some German text and it
    would tell me what
  549. that says in English?
  550. Or what if I walk up to someone
    and I want to make
  551. sure that I'm getting on
    the right subway line?
  552. Wouldn't it be great if I could
    speak in English, my
  553. computer, which is my mobile
    phone, which I always have
  554. with me, could do voice
    recognition and then could
  555. translate that text from English
    into German or Korean
  556. or Spanish.
  557. And then it could synthesize
    that text.
  558. And so it could do speech
    synthesis so that I could talk
  559. to anyone in any language
    even if I
  560. don't speak that language.
  561. The fact is, we're not that
    far from having that.
  562. We're pretty close.
  563. And the idea that you could drop
    down in the middle of the
  564. world with anybody.
  565. I got to go to a country in
    Africa last year, Tanzania.
  566. Very few people in the
    world speak Swahili.
  567. But if your computer can speak
    Swahili for you, then you're
  568. able to talk to anybody
    in the world.
  569. So I think that we're making
    good progress on that.
  570. Google Translate is not perfect,
    but it's free and
  571. it's only going to get better
    and better and better.
  572. So if you want to know, what's
    the direction to the nearest
  573. subway station?
  574. You can type that in.
  575. You can even get it written
    phonetically, and then once
  576. it's translated you can have
    it say that aloud.
  577. And then if something doesn't
    translate correctly, you can
  578. hover over it to see alternate
  579. So a lot of people think
    about Google as
  580. just a search engine.
  581. But Google's mission is to
    organize the world's
  582. information and make
    it universally
  583. accessible and useful.
  584. Great information is available
    in every language.
  585. Great information is available
    in Korean.
  586. And so the more that Google
    can do to help surface, to
  587. highlight, to display and show
    all of the great information
  588. that exists in Korean, the
    better off everybody in the
  589. entire world will be.
  590. This is just a picture to
    demonstrate where we'll be in
  591. a few years.
  592. Mobile is already to the point
    where we pretty much have a
  593. net connection wherever we go.
  594. And if you could take a picture
    of a water bottle and
  595. figure out, OK, where does this
    water bottle come from?
  596. Is there any nutritional
    information that I need to
  597. worry about?
  598. Pretty soon you'll be able to
    do that in any language.
  599. And we're getting very close to
    being able to talk to your
  600. phone and have it talk back
    to you in any language.
  601. So the future of Search is not
    just going to a desktop
  602. computer and typing into
    a web browser.
  603. The future of Search is that you
    will always have a smart
  604. computer right next to you.
  605. And you've seen this
    with the incredible
  606. growth of mobile phones.
  607. Your phones will be able to help
    you because they'll know
  608. more about you, because you
    choose to give information so
  609. that you can get fantastic
    information back out.
  610. Now, one more area about the
    future of Search is that it's
  611. not just the exact same
    list of links for
  612. every single person.
  613. In an ideal world, if you were
    to ask about information from
  614. just a random person on the
    street, or from a friend, who
  615. would you trust more to
    get your information?
  616. You'd probably trust
    your friend more.
  617. So if I were going to use a
    home router, or if I were
  618. going to buy concert tickets,
    I would trust Inhyuk
  619. and I would ask him
    for advice.
  620. I wouldn't just necessarily
    talk to any random person.
  621. So I show this picture
    earlier today.
  622. And you might not have noticed
    but, at the very bottom of
  623. this page, I have Louis Gray
    shared this on Google+.
  624. Now, we've started out on
    Google+, but we're actually
  625. trying to pull in more
    information from
  626. all across the web.
  627. Quora, FriendFeed, Twitter,
    TypePad, WordPress, all the
  628. different places you can find
    good information on the web,
  629. we would like to highlight
    that information.
  630. So when I searched for Korean
    pop, I found a recommendation
  631. from my friend for
    a specific video.
  632. That makes it more likely that
    I will probably like that
  633. particular video.
  634. And I don't know whether we
    want to play it right now.
  635. We want to save time
    for questions.
  636. But I played it last night, and
    it's a fantastic video.
  637. It's exactly the kind of thing
    that I think a good
  638. introduction for me
    that could lead to
  639. more Korean pop videos.
  640. So the future of Search
    is not just far away.
  641. It's coming very close.
  642. It's mobile, it's being able to
    understand language better,
  643. and it's being able to
    understand your relationships
  644. and highlight things
    from your friends.
  645. One last area is that a lot of
    people think about search
  646. engines almost like they're
    a black box.
  647. They don't know how they work.
  648. I think it's important for all
    search engines, every global
  649. search engine, every major
    search engine, to talk about
  650. how they work and explain more
    about their policies.
  651. In an ideal world, search
    engines would be transparent.
  652. They wouldn't be a black box.
  653. So five or six years ago, the
    fact that my team worked on
  654. web spam was confidential.
  655. It was a secret.
  656. We didn't even really like to
    talk about the fact that we
  657. tackled web spam.
  658. But we realized that's
    not the best policy.
  659. The best policy is to explain
    how Google works so that
  660. people understand how
    search engines work.
  661. They know not to be
    afraid of them.
  662. They know their advantages.
  663. And they know how to manage
    public information on the web,
  664. so that they're not surprised
    when information shows up in
  665. search engines.
  666. And I think it's been real
    progress for people to know
  667. how search engines work.
  668. It can only help if everybody
    has an idea of the criteria
  669. and the different types of
    signals that search engines
  670. use in order to score results,
    the sorts of things that I
  671. talked about earlier in
    the presentation.
  672. The last area that I wanted to
    talk about is that I would
  673. like to close out a little
    bit by talking about the
  674. importance of the Open Web.
  675. Now, before I talk about the
    Open Web, let me just say, I
  676. think we will have a little
    bit of time for questions.
  677. And I think we will have a
    little bit of a small gift for
  678. whoever wants to ask the first
    one or two questions.
  679. So be thinking now if there's
    a particular question that
  680. you'd like to ask because the
    first person who's brave
  681. enough, I think we will have a
    little something for them.
  682. I want to give you plenty of
    time to think about it in case
  683. you have any questions about how
    Google works or anything.
  684. So let's talk about the
    importance of the Open Web.
  685. This is a really interesting
  686. Historically, Internet
    Explorer has done
  687. very well in Korea.
  688. But according to the most recent
    stats, and I pulled
  689. these statistics, literally,
    yesterday, in the last six
  690. months, the usage of Google
    Chrome has more than doubled.
  691. Six months ago, in July of 2011,
    Chrome was about 4.3%.
  692. And if you look at the slide
    now, you can see that in
  693. December of 2011, Chrome
    is at 11.38%.
  694. That's important because Chrome
    is a browser from
  695. Google, but it's also
    a very good browser.
  696. It's fast. It fits very
    well with the pali
  697. pali ideas of Korea.
  698. But it's also secure.
  699. It protects users.
  700. It makes sure that they
    don't get spyware
  701. and malware and viruses.
  702. Every year, we have a contest
    in which people trying to
  703. crack browsers and they see
    if they can hack them.
  704. And for the last three years,
    Google Chrome was the only
  705. major browser that
    did not have a
  706. security hole get found.
  707. So Google has donated money to
    people who find security holes
  708. so that we can make Google
    Chrome even more secure.
  709. Now there's another reason
    why this graph is really
  710. interesting and why it's
    really important.
  711. Look at the share of
    Internet Explorer.
  712. In the last six months, it's
    dropped to about 80%.
  713. Now that is still huge.
  714. But if you have a website that
    relies on ActiveX or some
  715. other technology that only works
    on Internet Explorer,
  716. you're excluding 20% of
    all of your visitors.
  717. So if you have some technology
    that people could only use
  718. Internet Explorer with, one
    out of every five users is
  719. getting annoying, or is not well
    served, or is somehow not
  720. able to access your website and
    they have to go and unload
  721. their browser and load up
    Internet Explorer to use it.
  722. So website standards
    are important.
  723. Your website should work in
    any browser, not just in
  724. Internet Explorer, not
    just in Chrome.
  725. But if you build it right, it
    will work well in Safari, it
  726. will work well in Firefox,
    and it's more
  727. likely to work on mobile.
  728. You guys already heard
    today mobile is an
  729. incredible upward path.
  730. And so if your website is
    going to be prepared for
  731. things like the iPhone, which
    doesn't have Flash, or other
  732. mobile browsers, you want to
    use standard technologies.
  733. So it's absolutely the case that
    with Chrome above 10%,
  734. websites should use
    open standards.
  735. And in fact, you should avoid
    standards locked to one
  736. particular browser.
  737. Now when I talk about the
    importance of the Open Web,
  738. I'm not just talking about
    websites and web servers.
  739. I'm also talking about being
    open to search engines.
  740. So this was a newspaper article
    that came out in
  741. December in Korea, and it noted
    that something like half
  742. of government websites were
    blocked from being crawled
  743. from search engines.
  744. That means that there's a lot of
    resources that people could
  745. discover that they weren't
    being able to discover.
  746. Now, I want to note that
    there's been a lot of
  747. progress, even since this
    article has come out.
  748. So many, many organizations
    have, since this article,
  749. unblocked in robots.txt so that
    any search engine can
  750. come in and find high
    quality information.
  751. So it's not as bad as this
    article makes it sound
  752. anymore, but there's still some
    progress that could be
  753. made on making sure that
    websites are discoverable and
  754. are good resources.
  755. If you haven't seen this,
    this is something
  756. called the Khan Academy.
  757. I know that in Korea, education
  758. is incredibly important.
  759. It's vital that people be able
    to learn from the highest
  760. quality resources.
  761. And the web is becoming one of
    the highest quality resources
  762. available in the entire world.
  763. The Khan Academy is a set of
    videos and a website that you
  764. can use to achieve nearly a
    college-level education for
  765. free from anywhere in
    the world from home.
  766. So this is one guy.
  767. His name is Salman Khan, and
    he was a hedge fund trader.
  768. He made money as a banker.
  769. And then he said, OK, I'd like
    to do something that feels a
  770. little more meaningful
    with my life.
  771. And he was teaching
    his cousins math.
  772. Because it was a lot of trouble,
    he made these videos
  773. so that they could review the
    videos anytime they wanted.
  774. And he would call them
    on the phone.
  775. And eventually his cousins
    said, you know what?
  776. I don't want to talk to
    you on the phone.
  777. There's a lot of pressure.
  778. Just make the videos and then
    I can watch the videos as
  779. often as I need to so I
    can practice the math.
  780. And what really surprised Sal
    Khan was a lot of other people
  781. started to watch the videos,
    and so he started to
  782. make more of them.
  783. And now, you can learn all
    kinds of mathematics,
  784. economics, science, many
    different subjects just from
  785. watching these videos.
  786. The fact is the web is
    a fantastic source of
  787. information, and Korea should be
    represented as one of those
  788. fantastic sources
    of reputation.
  789. Nobody wants to be an island.
  790. An island is isolating.
  791. An island means that not as much
    development is happening
  792. as could happen.
  793. And so I think that there is a
    message that I'd like to leave
  794. for Korea, which is, we have
    a saying in English.
  795. And the saying is, to punch
    above your weight.
  796. What does that mean?
  797. To punch above your weight
    means you better than you
  798. would expect given the
    size of something.
  799. So it's someone who is
    doing a better job
  800. than you would expect.
  801. I think Korea is one of those
    companies that punches above
  802. its weight.
  803. It represents itself so well
    on the international stage.
  804. The World Cup, the G20, the
    Olympic Games that already
  805. happened, the Olympic Games
    that will happen in 2018.
  806. Korea has so much
    to be proud of.
  807. Korea is number one in the
    world in broadband
  808. penetration.
  809. Korea is number one in the world
    in getting information
  810. to people via broadband.
  811. I think the only area
    where Korea doesn't
  812. punch above its weight--
  813. at least, not yet--
  814. is getting its information out
    to the rest of the world.
  815. Korea is fantastic at providing
    good broadband, at
  816. providing wonderful things that
    it should be proud of.
  817. But if Korea could do a little
    bit more, so that if people
  818. want to learn about Hangul, if
    people want to learn about
  819. writing resources, if people
    wanted to learn about the
  820. official Sunshine Policies and
    the successes of President
  821. Kim, there's much great stuff
    that Korea can tell the world.
  822. And so I think it will be
    fantastic if that information
  823. is available to the
    entire world.
  824. So I hope that everybody has
    found this talk a little bit
  825. useful, that people have enjoyed
    hearing a little bit
  826. about the past of search
    engines, how search engines
  827. work today, and how to do better
    in search engines with
  828. search engine optimization,
    what the future of Search
  829. might look like, and then how
    important is for Korea to
  830. represent itself on the
    international stage, so that
  831. anybody wants to find more
    information about how great
  832. Korea is, anybody can find it.
  833. With that, you can see--
  834. this is me.
  835. I got to do a little bit of
    tourism on the weekend, and it
  836. was a lot fun.
  837. With that, I would love to take
    any questions that people
  838. have. I think we've got just
    a few minutes if anybody's
  839. interested.
  840. I think we do have at least a
    couple small things for the
  841. first one or two or three
    people who wanted to ask
  842. questions if anybody's
    interested at all.