Comparison Search

I’ve been comparison shopping searching! I thought it would be fun to search for the same generic keyword “hotels” on Google, Yahoo, Ask and Live Search and compare the first 20 results of each search engine. This is what I found out:

Number of URLs in SERPs that were unique to that search engine:

  • Google: 13 (65%)
  • Yahoo: 7 (35%)
  • Ask: 14 (70%)
  • Live: 10 (50%)

Unique domains (including subdomains) present in the top-20 results for each engine:

  • Google: 20
  • Yahoo: 18
  • Ask: 20
  • Live: 15

Diversity in result type
I categorized results in booking sites (expedia, travelocity, orbitz, venere etc); hotel chain sites (best western, hilton, choice etc.); comparison sites (kayak, sidestep etc), news sites (travel section of major newspaper); trade sites (hoteliers info, jobs etc); wikipedia (wikipedia :)); reviews (tripadvisor)

Google

  • booking sites: 11 (55%)
  • chain sites: 3 (15%)
  • comparison sites: 1 (5%)
  • news sites: 1 (5%)
  • trade sites: 1 (5%)
  • wikipedia: 1 (5%)
  • reviews: 1 (5%)

Yahoo

  • booking sites: 9 (45%)
  • chain sites: 7 (35%)
  • comparison sites: 1 (5%)
  • news sites: 0 (0%)
  • trade sites: 1 (5%)
  • wikipedia: 0 (0%)
  • reviews: 2 (10%) both results from the same domain

Ask

  • booking sites: 2 (10%)
  • chain sites: 17 (85%)
  • comparison sites: 0 (0%)
  • news sites: 0 (0%)
  • trade sites: 0 (0%)
  • wikipedia: 1 (5%)
  • reviews: 0 (0%)

Live

  • booking sites: 9 (45%) from 5 unique domains
  • chain sites: 2 (10%)
  • comparison sites: 1 (5%)
  • news sites: 0 (0%)
  • trade sites: 2 (10%)
  • wikipedia: 0 (0%)
  • reviews: 0 (0%)

What about the missing percentage?

  • The remaining 5% (1 result) for Google was an actual spam-suspect / made-for-adsense / low-quality site
  • Yahoo adds up to 100%. Well done! Except for the near duplicate from the same domain :)
  • Ask also adds up, but results are rather limited to hotel chains
  • Live’s missing 25% (5 results) are for 3 web sites that limited their hotel offer to one specific, limiting location (e.g. only hotels in Vietnam) and 2 web sites that were exact duplicates with different URLs

Do URL’s matter?

Occurrence of actual query (either “hotels” or “hotel”) in the domain name:

  • Google: 8 (40%)
  • Yahoo: 8 (40%)
  • Ask: 3 (15%)
  • Live: 11 (55%)

Occurrence of actual query (either “hotels” or “hotel”) elsewhere in the URL:

  • Google: 7 (35%)
  • Yahoo: 5 (25%)
  • Ask: 1 (5%)
  • Live: 4 (20%)

Domain extension:

  • Google: 15 .com (75%), other extensions .nl; .org; .co.uk; .be; com.au
  • Yahoo: 19 (95%), 1 .be result
  • Ask: 19 (95%), 1 .org result
  • Live: 19 (95%), 1 .biz result (wait a minute, wasn’t that a spammy extension?)

Further comments:

  • First position was the same for all 4 engines (hotels.com). This was the only exact co-occurrence of ranking.
  • Google’s rather successful attempts to capture user intent is reflected in the diversity of results. Live also made a (less successful in aggregate) attempt to diversify results for the very generic query “hotels”.
  • Co-occurrence of the same page in the SERPs was strongest between Google and Yahoo (10 out of 20), then Google and Live (6 out of 20). Google was less in agreement with Ask (only 4 out of 20)
  • Same pages in the SERPs with combinations of 3 search engines agreeing was Google, Yahoo & Live with 5 out of 20, and Google, Ask & Live a close second with 4 out of 20.
  • The only co-occurrence that didn’t involve Google was between Ask and Yahoo (1 result)
  • Home page results (no trailing slashes) were 11 for Google (55%), 15 for Yahoo (75%), 19 for Ask (95%) and 11 for Live (55%)
  • Wikipedia lovers turn out to be Google (the usual suspect) and Ask

4 comments ↓

#1 Fabbrì on 06.21.07 at 11:11 am

As usual a post full of intresting data correlations. Zu keep the good work up.

#2 bart on 06.21.07 at 12:48 pm

yes, amazing! can you do the same for a generic keyword like ‘mobile diphthongs’? i think it would be of great interest to the world! :-)

#3 Quali saranno i migliori risultati di un motore? (parte1) on 08.25.07 at 3:43 pm

[…] letto già da qualche giorno il post sulla comparazione delle SERP sul blog di […]

#4 Hotel ischia on 02.18.08 at 1:19 pm

We see the excessive power of google

Leave a Comment