|
|
|
|
Effective Internet Search:
Search Help
by Judith Gill
Search engines are becoming the card
catalog to the Web. - CyberAtlas, March
2003 [1]
Online searching is one of the most
important and valued activities on the Internet
— and search engines are the gateways to access its
information. So improving your ability to
improve your online search skills
— to effectively use search engines — is more important
than ever before. Here are some improvements you can
immediately put into practice:
A 2004 Enquiro report [2]
describes search engine results pages as a "navigation menu" — "a navigation aid in negotiating
the online research interaction, as people continually
refer back to it and launch another online exploration
from this starting point." |
- Use more
than 1 search engine.
- Enter at least 3 search terms.
- Refine
your queries.
- Try
out
advanced search features.
- Look
beyond the first page of search results.
- Find out
how search engines work.
- Improve your overall search skills.
-
Use more
than 1 search engine
Today's search engines may be capturing
as little as 1 percent of the Web, largely because of how
they find and index online resources. - Associated Press, March 2004 [4]
Whether search engines capture one percent — or ten
percent of Web content, as some research claims — it's
self-evident you need to use more than one search engine.
Regardless, most people still employ the same tried and
true search engines time and again. Two recent iProspect surveys
illustrate this point.
Search Engine Loyalty |
|
User Behavior |
2002 |
2004 |
|
Always use the same search engine |
52 |
57 |
|
Alternate among several favorites |
35 |
30 |
| Deliberately use different
search engines for various types of searches |
13 |
13 |
| Source:
iProspect 2002 Search Engine Branding Survey [3]
and iProspect 2004 Search Engine User Attitudes [4] |
But it's a
mistake to rely on a single search engine for all your
online searching. Because if there's one big lesson for you in our book,
it's that no one search engine can do it all for
you — whether you are a rank amateur beginner, an
experienced user with sophisticated needs, or you're
somewhere in between.
Why? One major reason is
that search
results vary greatly depending on the search
engine used. A case in point: Expert Greg Notess has
long studied search engine database overlap, and has
consistently found little overlap in the search results
they produce — even among their top ten results and for
the most popular search terms used. Witness the chart
below.
In 2002, for instance, four small searches were run
on ten different search engines — AlltheWeb, AltaVista,
Direct Hit, Google, HotBot, iWon, MSN, NLResearch, Teoma
and Wisenut. Of the 141 unique webpages found:
- 50%
were only discovered by one search engine, and
not always the same one.
- 72%
were found by only two search engines.
- 79%
were located
by only three.
In other words, there was very little
overlap in search results produced by the various search
engines studied, and this has been consistently so
throughout the years. In fact, one recent study, at
searchenginewatch.com/searchday/article.php/3524411
shows that there is more divergence than ever.
Search Engine
Overlap
|
| Year |
Number of
Search Engines Studied |
% Webpages
found by 1 Search Engine |
% Webpages
found by 2 Search Engines |
% Webpages
found by 3 Search Engines |
| 2002 |
10 |
50 |
72 |
79 |
| 2000 |
14 |
37 |
63 |
77 |
| 1999 |
13 |
46 |
68 |
83 |
| 1998 |
5 |
71 |
87 |
96 |
| 1997 |
4 |
48 |
74 |
95 |
| Source:
Greg R. Notess, Database Overlap, Search Engine
Showdown, 1997-2002 [5] |
Suggested
strategy: Subject directories, meta search engines,
specialized and deep web search engines can all deliver
the results you want.
They all have the potential to deliver
superior search results under certain conditions, as
well as complement each other in the search process. The trick is to experiment with a number of search
engines:
- Try some out.
- Learn how to maximize their abilities by reading
search engine Help.
- Experiment to determine how to use them most
appropriately to meet your search needs.
-
Enter
at least 3 search terms
The vast majority of Web searchers use
approximately two terms in a query, have two queries
per session, do not use complex query syntax, and
typically view no more than ten documents from the
results list. -
Journal of the American
Society of Information and Technology, 2000 [6]
Even though the above study was published in 2000, this
would still be an accurate description of user
search behavior today. Mondosoft's research [8]
is more specific, finding:
- 52% of
searchers only
use a single search term.
- About 30% use two words.
- Only 2% use more than four words.
Suggested
strategy: You can increase the chances of quickly locating your target
content by adding
more keywords to your search. Specifically:
- Zero
in on important terms likely used in the title
of the document, which are given the highest priority in
a search engine's logic.
- Include expected words in a general description
of the document, such as its keywords or major
subject area.
- Use
important words included in the body text.
- Try entering the name of the author(s), or sponsoring
organization or company, or the name of the newspaper,
journal or publisher.
- Make
sure you place the most important words first.
- Avoid small or unimportant words (e.g., "a", "the", "to") whenever possible.
-
Refine
your queries
The majority of us type in a few top of mind
words and hit the search button without giving a lot
of thought about how to construct our search query.
We tend to take an iterative approach to searching,
refining our search based on the results that are
returned to us. -
Enquiro, April 2004[7]
The above is how
most people search [7], [9]. 2001-02 research indicated when
target content was not located:
- The
majority of users abandon their search altogether
after the first or second attempt [8],
[10],
[11],
[12],
[13].
- Even if
they search more than once, most do not add or delete many terms to their subsequent queries [10].
- Only 7.5% refine
their search with additional keywords when they
couldn't obtain satisfactory results [3].
But iProspect's
2004 Survey [4]
revealed a significant change in user search behavior:
- When respondents
were dissatisfied with the first three pages of
results after an initial search, 91% modified their
original queries and retried them using the same
search engine, before giving up on that engine to
return satisfactory results.
This was an
increase of 20% over 2002 survey
results.
- iProspect interpreted
this to mean that users had more
confidence in their selected search engine to formulate
queries than in their own abilities.
Suggested
strategy: If you don't get the search results you
want:
- Don't
give up — don't abandon your search.
- Try
refining it:
- Add more search terms;
- Change some of your search words;
- Use
a thesaurus to suggest additional search
terms.
-
Try out
advanced search features
Most search queries are very simple and generic
in nature. Few people take advantage of extended
Boolean search capabilities or other advanced search
features. - Enquiro, April 2004 [7]
Advanced search features typically include the use
of Boolean operators, phrased searching, and stemming,
for example. They permit the formation of more complex
queries, which can lead to improved search results,
more quickly.
Research
indicates most searchers do not know that search engines
offer advanced search features. Therefore, not
surprisingly, few use it [6],
[7],[9],
[10].
For instance, a recent Enquiro study [7]
indicated about three-quarters of respondents
rarely or never used them, 5% not
even knowing what they were. Only 25% reported often or
almost always using them.
Advanced Search Feature Use
|
|
User Responses |
Percentage |
|
Don't know what advanced search
is |
5% |
|
Never use it |
21% |
| Rarely use it |
48% |
| Often use it |
21% |
| Almost always-always use it |
4% |
|
Source: Enquiro, April 2004 [7] |
Suggested
strategy: As you acquire more knowledge about using
search engines and develop more confidence in your
search abilities:
- Start
experimenting with advanced search features.
- For
explanations on advanced search features and how to
use them:
- Consult the Help section of your selected
search engines;
- Use the links found on their general search
interfaces;
- Read
our book which deals with advanced search
features in great depth.
Finding that needle in a
haystack
 |
-
Look
beyond the first page of search results
If none of the first ten are any good, what are
the chances that the next ten will be any
better? - Mondosoft, July 2002 [8]
Finding the content you want is only half the battle.
Getting 20 pages of search results is not helpful either
— because the average user won't look at most of it.
Research tells us the majority of web users do not
browse beyond the first or second page of results, a
common trait shared in both Europe and North America [15],
[3],
[4],
[7], [8],
[9],
[10],
[14].
The average search engine user scrolled through
1.8 result pages during a typical search.
- PEW, August 2004 [15] |
For example, in 2003, Consumer WebWatch reported that 88% of result links
selected by participants were located on the first page
[14].
And when searching for health information, it's even
higher. A 2002 British Medical Journal
study found that 97% of
the time consumers chose a search result ranked among
the top 10; in 71% of cases, they selected a link from
the top five results
[16].
To summarize iProspect's 2004
Survey [4]
results below:
- 23% of search engine users will only look at the first few search results
before trying another search.
- 41% expect to find the answer to their query on the first page of search
matches.
- 67% won't look past the first 2 pages of results.
- 82% will not
read beyond the third page of
search results.
Search Engine User
Behavior
|
|
Search Results Viewed |
% of Users |
% Cumulative
Total |
|
First few entries |
22.6% |
22.6% |
|
1st page of results |
18.6% |
41.2% |
| 2nd page of results |
25.8% |
67% |
| 3rd page of results |
14.7% |
81.7% |
| More than 3 pages |
10.8% |
92.5% |
| Entire list, if not too long |
7.4% |
99.9% |
|
Source:
iProspect 2004 Search Engine User Attitudes Survey [4] |
Suggested
strategy: The most relevant search results may not
be necessarily listed first. It all depends on how your
selected search engine gathers data, then indexes, ranks
and prioritizes it. Also be aware of paid listings which may or may
not be clearly identified, but are generally placed at
the top of your search results pages.
If you don't get
the search results you want:
- Start
looking past the first page of search results.
- Preferably consult at least the first three pages.
- If you
still have no luck:
- Refine your queries;
- Add more or use different search terms;
- Try
out advanced search features.
-
Find out
how search engines work
Most participants had little understanding of
how search engines retrieve Web pages or how they
rank or prioritize links on a results page.
- Consumer WebWatch, June 2003 [14]
Every search
engine has its own individual method for collecting, cataloging, and retrieving information to answer your
queries. Nonetheless, to be a skilful searcher, you
must acquire some basic knowledge about how they work.
With over three hundred search engines to choose from, this
can be a daunting task.
Suggested
strategy: Everyone has the same goal when
searching — you want the most relevant and
authoritative information listed and prioritized to meet
your needs. To achieve this, you must have some
understanding about the inner workings of search
engines. To learn how their ranking and prioritizing
technologies work, how they make their money, and how
this influences the search results you get:
-
Improve your overall search skills
People don't want to be spending time searching and looking for things. They want to be spending
the time analyzing the information.
- Factiva, 2003 [17]
More than ever before, finding information
takes know-how. By increasing
your knowledge and improving your online search skills,
you can work more successfully with
your selected search engine(s) to accurately locate
your target content quickly and easily.
Suggested
strategy: The search suggestions listed above are
practical — and you can start implementing them today.
References & Links
-
CyberAtlas, Search Guiding More
Web Activity, Brian Morrissey, March 13 2003: www.clickz.com/stats/big_picture/traffic_patterns/article.php/5931_2109221
-
Enquiro,
Into the Mind of the Searcher, Gord Hotchkiss,
February 2004:
www.enquiro.com/research.asp
-
iProspect Search
Engine Branding Survey, May 2002: www.iprospect.com/web_site_promotion/press11142002.htm
-
iProspect
Search Engine User Attitudes Survey, March 2004:
www.iprospect.com/premiumPDFs/iProspectSurveyComplete.pdf
-
Search Engine
Showdown, Search Engines Statistics: Database Overlap
- Little overlap despite database growth, Greg R.
Notess, March 6, 2002:
www.searchengineshowdown.com/stats/overlap.shtml
-
Journal of the
American Society of Information and Technology, Web
User Studies: A Review and Framework for Future Work,
B.J. Jansen and U. Pooch, 52(3), p. 235-246, 2000:
jimjansen.tripod.com/academic/pubs/wus.html
-
Enquiro,
Search Engine Usage in North America, Gord
Hotchkiss, Marina Garrison and Steve Jensen, April
2004:
www.enquiro.com/research.asp
-
Mondosoft, Web
Site Usability Metrics: Search Behavior - Search
Trends, July 2002:
www.mondosoft.com/SearchBehaviorWP.pdf
- Florida
International University, Things you might not know
about how real people search, Marc L. Resnick and
Rebeca Lergier, Miami, June 28,
2002: www.searchtools.com/analysis/how-people-search.html
-
IEEE/CS Computer, From e-Sex to
e-Commerce: Web Search Changes, Amanda Spinks,
et al, March 2002:
ist.psu.edu/faculty_pages/jjansen/academic/pubs/ieee_computer.pdf
-
UIETips, People
Search Once, Maybe Twice, Jared Spool,
November 20, 2001:
world.std.com/~uieweb/Articles/search_once.htm
-
UIETips, Users
Don't Learn to Search Better, Jared Spool,
November 27, 2001:
world.std.com/~uieweb/Articles/not_learn_search.htm
-
Alertbox, Search:
Visible and Simple, Jakob Nielsen,
May 13, 2001:
www.useit.com/alertbox/20010513.html
-
Consumer
WebWatch, False Oracles: Consumer Reaction to
Learning the Truth About How Search Engines Work,
Leslie Marable, New York, June 30, 2003:
www.consumerwebwatch.org/news/searchengines/index.html
-
Pew Internet & American Life
Project, The popularity and importance of search engines: Pew Internet Project Data Memo,
August, 2004:
www.pewinternet.org/pdfs/PIP_Data_Memo_Searchengines.pdf
-
British Medical
Journal (BMJ), How do consumers search for and
appraise health information on the world wide web?
Qualitative study using focus groups, usability tests,
and in-depth interviews, Gunther Eysenbach and
Christian Kohler, March 9, 2002, Vol. 324:
bmj.bmjjournals.com/cgi/reprint/324/7337/573.pdf
- Factiva,
2003:
www.factiva.com/index.asp
|
|