Nine Months of Search Engine Data of My GoogleNumber

It takes nine months for us to grow inside the womb and apparently it takes nine months to really analyze how search engines crawl and get results.

Every week for the last nine months I have done a search for “Auren Hoffman” in eight major search engines (Google, Lycos, All the Web, Teoma, Gigablast, Wisenut, Yahoo, and A9). I recorded only the raw number of results as my objective was to understand the reach of each search engine rather than the accuracy. Since, as far as I know, I’m the only “Auren Hofman”, the results are fairly finite and easy to define. (a chart and the raw data is below)

This experiment was born of my fascination with other people’s fascination about their “Google Number” (see What Is Your Google Number). When I started, my Google Number was only 794. It climbed to 2880 in March. Today, it is inexplicitely only 2220 — which means that either:
(a) about a six hundred pages with my name on it have dropped of the face of the earth in the last five months
(b) Google changed the way it crawls and stopped caching certain sites
(c) google misrepresents its raw search results (we’ll get to that later)

My observations:

1. Fluctuation in weekly results
Some search engines have results that fluctuate widely every week and others are very consistent (see graph below). Google fluctuates widely each week even though it is obvious (from just looking at the search results) that Google spiders new pages very quickly. Search engines like Gigablast and Wisenut fluctuate very little — and it also seems they rarely add newly crawled information to the search engines.

2. Numbers that end in zero
I noticed that for results above 1000, all Google and Yahoo results end in a zero. Maybe they are rounding to the nearest ten or maybe they are just making an approximation (which might be the reason that results vary so widely). If you follow every result, they both seem to overestimate the number of pages.

3. There is a huge discrepency between the search engines
Even if the results from Google and Yahoo are a bit inflated, they are still vastly higher than the other engines (though All the Web) indexes a good deal of pages too. That means that purely on who indexes the most pages, there is a big difference between the engines.

And indexing the most pages is very important. Now everyone always seems to talk about relevance. And relevance is very important. But so is mass. Especially when you are looking for something more archane.

What is your Google number??

Search Results of "Auren Hoffman"

See Larger graph here

Google Lycos All the Web Teoma Gigablast Wisenut Yahoo A9
12-Nov 794 649 636 582
11/16/2003 808 402 636 582
21-Nov 866 652 636 582
12/7/2003 949 419 406 540 540 956
12/13/2003 1020 418 404 537 267 984
12/18/2003 1070 411 404 539 268 1001
12/23/2003 1060 411 408 539 269 1001
3-Jan 994 409 402 539 269 1001
10-Jan 993 418 409 539 269 307
24-Jan 2390 693 699 537 272 1001
31-Jan 2700 677 693 557 273 1001
8-Feb 1810 696 680 557 282 1001
14-Feb 1570 693 703 509 285 1001
27-Feb 1890 719 703 509 280 1001 2070
5-Mar 1920 668 655 509 278 1001 2190
13-Mar 2770 690 674 509 278 1035 2140
3/20/2004 2880 649 483 568 278 1035 2190
3/26/2004 2970 636 307 568 278 1035 2400
4/6/2004 2160 275 361 568 278 1035 2180
4/10/2004 2010 278 421 567 278 1035 2100
4/18/2004 2200 290 335 643 278 1035 1580 295
4/23/2004 2330 274 290 643 278 1035 1790 297
4/29/2004 2380 268 293 643 271 1035 1730 320
5/9/2004 2390 271 276 680 293 1035 303
5/14/2004 2250 255 258 680 300 1035 1590 304
5/21/2004 1890 246 217 684 306 1035 1410 297
5/28/2004 2010 252 238 683 314 869 1390 303
6/6/2004 2080 261 335 706 317 869 1410 292
6/11/2004 1910 254 334 712 325 869 1350 299
6/18/2004 1670 250 439 712 336 869 1350 295
6/24/2004 1400 248 451 728 361 869 1340 284
7/2/2004 1430 256 484 728 363 664 1310 288
7/8/2004 1370 254 484 728 360 664 1290 278
7/16/2004 1390 253 1220 741 361 664 1250 305
7/23/2004 1510 251 1190 741 361 664 1230 309
7/29/2004 1550 251 1200 742 361 664 1240 321
8/6/2004 871 251 1210 758 361 664 1230 341
8/13/2004 1640 366 1170 756 361 664 1310 345
8/18/2004 1720 378 1180 133 361 664 1330 328
8/25/2004 1950 367 1230 765 361 664 1330 302
9/10/2004 2220 332 1210 765 361 660 1240 300

3 thoughts on “Nine Months of Search Engine Data of My GoogleNumber

  1. peter caputa

    you should correlate the new links/ month (use technorati and bloglines)linking to you. I believe that a link has less relevancy as time goes on, in google’s world. So, if a lot of people linked to your weblog during the months where you had “higher” numbers, that could explain why the #’s went down afterwards.

    Reply
  2. Get Real

    Auren Hoffman on GoogleNumbers

    Auren Hoffman has been tracking the various search engine results based his own name, and has found a wide disparity in results:Auren Hoffman[from Summation] Every week for the last nine months I have done a search for “Auren Hoffman” in…

    Reply
  3. Marc's Voice

    How many mentions do YOU get?

    Auren Hoffman just spent nine months analyzing the acuuracy and results of all the major search engines. He notices lots of descrepancies and inaccuracies. Me too. I’m currently at 46,900. There’s one other Marc Canter out there running around. He’s th…

    Reply

Leave a Reply