| Having been in this industry for as long as I | | | | person searching from San |
| have, I often forget some of the basics. | | | | |
| Well, it's not that Iforget, it's just that I | | | | Francisco will likely be served their search |
| assume that everyone else in the industry has | | | | results from a data center near them, such as |
| the same knowledge level as I do. | | | | an Oregon or |
| | | | |
| So when I get a "newbie" asking a question | | | | California data center. |
| like "Why are my Google results different | | | | |
| than my clients" I have totake a step back | | | | It is because of the differences in these |
| and realize that we aren't all equal. | | | | data centers that someone searching in New |
| | | | York will see differentresults than someone |
| Therefore in this article I'm going to answer | | | | in Los Angeles. |
| this one seemingly simple question. While it | | | | |
| may be simple tothose of us "in the know" | | | | It is reasonable to think that each data |
| this isn't always the case with others. | | | | center acts somewhat independently of the |
| | | | others. That means that theirupdate schedules |
| One of the scariest things you can do as a | | | | are different as are their crawling |
| new SEO is talk to a client. That's because | | | | schedules. One can even assume that the |
| you are always worriedthat they are going to | | | | algorithm changeswhich affect the index |
| ask you a question that's over your head. | | | | happen at different times as well. |
| And while you can fake your way through | | | | |
| somequestions when you are new, there are | | | | This accounts for why there are differences |
| some that just stump you. | | | | in search results. Because of Google's |
| | | | perpetually updating index,the results you |
| Of course even the greenest of most SEO's | | | | see near you may be similar to results in |
| usually know more then their clients so you | | | | other data centers but ultimately different. |
| don't often have to fakean answer. Of course | | | | |
| I would never advocate faking it. When I | | | | This could be due to how the crawlers |
| don't know, I've found the best thing you | | | | retrieve sites - a crawler may find a site |
| cansay is "I don't know, let me find out and | | | | closer to it more easily andtherefore add it |
| I'll get back to you." | | | | to the index sooner than a geographically far |
| | | | away crawler would find the same site. |
| And that's just what happened the other day. | | | | |
| A new person came to me and said "I had a | | | | For example, a site in Vancouver, Canada may |
| client who saw different | | | | appear in the California data center days |
| | | | before the eastern datacenters. Because this |
| Google results than I did, and I didn't know | | | | site would be added to the west data center |
| what to tell them." | | | | sooner, it will have an impact on the |
| | | | searchresults returned sooner. |
| So let me give you the answer I gave him. | | | | |
| | | | Consider it like the ripples you see in the |
| Google, like all the other engines, is | | | | water when you drop a rock into a pond. If |
| comprised of literally tens of thousands of | | | | you drop just one rock,you see the ripples |
| servers. Each server is partof a larger | | | | move out from where you dropped it. However, |
| cluster of computers. Each cluster forms | | | | if you drop 2 rocks close to each other and |
| part of a datacenter. Each datacenter then | | | | atslightly different times you see how the |
| acts as anindependent branch of Google. | | | | ripples interact with each other when they |
| | | | meet. |
| These datacenters are found all over the US. | | | | |
| For the longest time Google only had about 13 | | | | The index changes reflect this type of |
| data centers thatserved all the results to | | | | interaction. One site can have a subtle but |
| the world. Now the number is estimated at | | | | noticeable effect on theindex. Yet the |
| around 80. | | | | effects aren't noticed across all data |
| | | | centers at the same time. We can also see |
| While some of these data centers are used for | | | | the changes inthe index grow over time, so |
| pre-testing results (for example, testing a | | | | that one Vancouver site's effect grows over |
| new algorithm outbefore moving it to the main | | | | time, but the effect is differentacross the |
| data centers) most are used just to deal with | | | | data centers because changes happening with |
| the load that Google receives everyday. | | | | other sites also has an impact. |
| | | | |
| These data centers are dispersed throughout | | | | As you can see, this is why you will see |
| the US in geographically specific areas. | | | | different results across the data centers. |
| They have done this sothat queries are served | | | | It's not necessarilybecause of one single |
| to the data center nearest to the user. | | | | event. Like SEO itself, it's a culmination |
| | | | of smaller events which causes the |
| For example, while there are a large number | | | | noticeabledifferences. |
| of data centers on the eastern coast, a | | | | |