This survey has been on the SearchTools
site since December 1998. The following report summarizes the responses
rating search tools with many responses (6 or more entries) as of July
31, 2002. It is a self-selected group of respondents, and probably not
statistically significant, but the results are interesting.
This page summarizes the ratings for search engines: for information
on web site size, audience, server location, etc., see the Survey
Search engines should be fast, accurate and easy to install, they have
to index all published file types on a site. Some situations require
a simple admin interface, preferably via browser, while others need
an SDK to integrate with other applications. But universally, search
engines must be responsive, have good relevance ranking and provide
basic customization, to make their customers happy.
This survey is not at all rigorous: it is a self-selected group of
respondents, and not statistically significant, but the results are
interesting. While the ratings vary for each engine, many respondent
provided enlightening comments. When reading the disadvantages section,
please check the date entered: many of these search engines have been
In the survey, we included a section about the site search tools that
web administrators are using. The questions asked about the name of
the product and allowed them to rate their tool, using this measure:
We also asked the people how long they have been using the engine --
those who are new to a product may not be the most reliable at evaluation.
And we provided fields to describe what they liked and disliked about
Disclosure: The SearchTools.com site is provided as a free service to the web
development community and is not sponsored by any advertisers. However, Search
Tools Consulting has provided analysis and information to search engine devlopers
including Atomz, AltaVista, Siderean, Google, Inktomi, Maxum (Phantom) and Mondosoft
(MondoSearch). We do not give them site visitor or survey personal information
or allow our relationships with any vendors to change any product review or analysis.
The following tables list the responses rating the 16
most popular search tools (6 or more entries) as of July 31, 2002.
Comments from users indicate that the highly-rated products provide
solid indexing and search functions, are flexible and easy to administer.
Those with lower ratings tend to be older packages which don't have
configurable indexing, adjustable results rankings, or good administration
- New in 2001/2002, Microsoft's Site Server makes a respectable showing,
while Logika's FusionBot service has at least 23 happy customers,
who were asked to respond to the survey.
- In 2000/2001, the FreeFind remote search service gets very good
marks, while the Muscat product and service (now owned by SmartLogic
and/or BrightStation) is not well regarded by its users.
These search tools had five or fewer responses, so we didn't even try
to average the ratings. However, we did report all the ratings and comments
for these search engines:
Alkaline, AltaVista, Apple e.g., Ask Jeeves, ASTAware SearchKey
PRO, Autonomy, Beseen, Bravenet search, cfm.index, CNIDR Isearch, Database,
DBtextworks Intranet Spider, Delphes, DioWeb, divine Participant server,
Domino, exploit7, Fast Site Search, Fluid Dynamics, Freesite, Fulcrum,
google, Harvest, HitBox Search, HomePageSearch applet, Hotlinks, Hyperseek,
IBM NetQuestion, iCat, ICE, iFilter, Inquery 5.0, Intelligent Miner
for Text, Lycos Site Spider 1.0, Namo Web Search, NetIS, nsearch, OfficeVision/400,
Oneworld Search engine, OpenFTS for Postgres, opentec, Oracle ConText,
Oracle InterMedia, Perlfect, Phantom, PicoSearch, Pinpoint, PLWeb, Qfinder,
ResourceMiner, Sambar, Scout, Search Light, Search Maestro, search maker
Pro, search.pl, SearchCSV, SFE Search, Sinky Search Engine, SiteMiner,
Sun i-Planet search, Surf Map, web-search/dataeasy, WebBoard, WebIndex
/ WebFind, WebSTAR, WhatUSeek, yahoo store default search.
While each individual engine had few responses, the comments
in the responses are quite enlightening.
This report covers custom-written and homemade search engines. Of these
entries, only a few people were satisfied with their search engines.
Complaints include problems with coverage, duplicates, relevance ranking
and speed. Given these problems, and the number of solid search engines
available, whether as commercial products, remote services or open-source
code, we recommend against custom search engines.