Google has released a new Google Desktop Search tool today that allows people to scan their computers for information in the same way they use Google to search the web.
"This is what I think is one of our our more exciting announcements this fall," said Marissa Mayer, Google's director of consumer web products. "Our users have been asking for this. They say, 'Google is great, but why can't I search my computer the same way'."
The tool is remarkable for its power yet simplicity. Rather than create a standalone application, Google Desktop Search seamlessly blends into Google itself. Those using the tool see a new "Desktop" link on the Google home page and search results page. Selecting this link brings back results found on their own computers.
In particular, the tool indexes the full text of:
- Email within Outlook or Outlook Express (notes, contacts, journal and to do list items are not included, nor are emails in the Deleted Items folder)
- Microsoft Word, Excel & PowerPoint files
- AOL Instant Messenger chats
- Web pages viewed online in Internet Explorer or any HTML file saved to your computer
- Plain text files
The tool also indexes the text within the file names of image in JPEG or GIF formats, giving it rudimentary image search capabilities. File names of Adobe Acrobat PDF content and names of some other file types are also indexed. Full text indexing of information in these files is NOT done.
Unlike with Gmail or regular Google searches, ads are not shown with desktop search results or content viewed through desktop search.
Google Desktop Search is offered for free, is an incredibly small download size of just 450K and importantly provides such useful search memory-style features to Google that anyone using the search engine will want to install it to enhance their experience.
As a result, Google should quickly gain new desktop search converts in this latest battleground of the search engine wars -- and on territory that some had assumed Microsoft would win on. Some have predicted Microsoft would integrate some powerful desktop/web search feature into the next version of the Windows operating system that would be a Google-killer.
As it turns out, the next version of Windows has been delayed until 2006. However, Microsoft still plans to release a standalone desktop search application later this year, if not earlier. Ask Jeeves also plans to release one later this year. Yahoo is rumored to be developing its own tool.
AOL this week may release a desktop search tool of its own. In fact, the AOL tool sounds so similar to Google's tool that it suggests AOL may be licensing Google's or partnering with Google to produce its own. So far, no news from Google on whether this is the case.
Beyond the major search players, Google also has to compete with other desktop search players already out there with existing products.
Copernic Desktop Search is probably the most serious contender, in that it's free, easy to use and actually more powerful that Google's tool in some ways, such as in indexing more content types, better ability to restrict queries and easier implementation of restricting what exactly gets indexed. Blinkx is also a new free tool, and some paid tools include x1, 80-20, and dtsearch.
Why won't these tools trump Google? Those that cost money have an obvious disadvantage. The longer download time -- all are 2MB or more -- will also put off non-broadband users. However, the key issue is that none of the tools has the integration into Google itself that I suspect will be compelling to its millions of loyal users.
Does Google view the move onto the desktop as crucial in defending against Microsoft? Google didn't answer that directly, but indirectly, it's clear that on any search front, they want Google first and foremost in the user's mind.
"Doing really well with web search does have merit in and of itself. That said, if we do have interesting new search technology that can be applied to the user experience, it makes sense to do that. It makes sense that our users, when they think of search, they think about using Google or a Google tool," Mayer said.
The new tool also sheds new light on the speculation over whether Google would launch its own browser. Yesterday, Google cofounder Sergey Brin commented that Google wanted to somehow improve and enhance the experience of existing browsers. Now the desktop search implementation suggests that Google may be browser neutral -- it may instead seek ways to improve any existing tool to bring people ultimately back to Google.
Google Desktop Search is only for Windows XP or Windows 2000 users -- no news on a Mac version from Google, sorry. Once installed, the application starts indexing information on your computer in the file types it understands. At the moment, only files on your primary hard drive (the C: drive for most people) are indexed. Those on additional hard drives won't be searchable.
Indexing is fast and only happens when your computer is idle for 30 seconds or longer. Once the index is built, it is continually updated with changes on the fly. Get a new email? Visit a new web page? All this information is automatically recorded and made searchable within seconds.
"Our goal for the application is to have it behave as if you had photographic memory of what's on your computer," Mayer said.
How to search? A little swirly Google Desktop Search icon is shown in your Windows taskbar. Double-click on this (or right-click, then choose Search), and an Internet Explorer window will open with the Google home page on it -- or at least, what looks like the Google home page.
In reality, you're getting the Google Desktop Search home page. This is a page hosted on your computer, the home page of a local web server created to dish up what Google Desktop Search has indexed and found.
Do a search on this page, and by default, you'll search the contents of your desktop. A combined list of everything found will be shown, with little icons indicating if something is a web page, an email and so on. You'll also be shown a count in the reverse bar under the search box indicating the total matches, the number of email matches, file matches, chat matches and web history matches.
Each count is also a link, and clicking on them will narrow your search to one type of data. In other words, click on the count for emails found, and you'll be shown only matching email messages.
Phrase searching with quotes and term exclusion using the - sign, as on Google itself, is supported.
Results are sorted by date by default. The "Sort by relevance" link lets you rerank results so those deemed best to match your search terms come first.
How's relevancy determined? Different factors are used, but Google won't disclose them. From what I can see, the proximity of search terms near each other is the key factor -- content that contains all your search terms in the order entered seems favored.
Any search engine marketing types hoping the tool will give insight into better rankings on Google are likely to be disappointed, however. Off the page factors such as link analysis are crucial with Google's web page rankings. They can't be measured using this tool. In addition, the on-the-page analysis used by the tool is likely different in some significant ways compared to web page analysis.
Dates shown vary based on the content involved. For email and IM chat, it's when the email was sent or the IM chat happened. For web pages, it whenever a page was last viewed. For files, it's when that file was created or last viewed.
Last viewed? You mean Google will know when you last viewed a file, even if the file isn't modified? Yes, and this record of what you've seen is one of the tool's coolest features.
Any item listed will initially have a "1 cached" link after its file name. Similar to the Google page cache feature, this lets you see a copy of the file as Google has indexed it, without actually opening the file itself. So if you have a spreadsheet file, you see a copy of the spreadsheet without having to open Excel.
Each time you view something, a snapshot of what you've seen is created. Did you visit the same web page several times in a month? A copy of the page each time you visited is made. The "1 cached" link will change to reflect the number of copies recorded.
This is a fantastic way to keep a record of exactly what you've seen on the web and how you saw it, over time. On many occasions, I've wanted to go back and see how a web page may have looked a few days ago, a few weeks ago and so on. Tools like the Internet Archive have sometimes helped, but not always. The new tool Seruku is another solution, but at a small cost.
Now Google Desktop Search makes it easy to painlessly preserve your own archive of what you've seen and for free. It becomes, as Gary Price wished for last week, a "Tivo for the web."
In addition, the cached copies of your local files provides some automatic backup insurance. Make a change to a file, then wish you hadn't? Visiting your cached copies may help you get back some of what was modified. The data won't be in the original document format -- with spreadsheets, it can especially look weird, but some of what you grab may help.
Searching your desktop can be done by opening the special Google Desktop home page, as described above. However, I suspect many people will simply end up searching via the regular Google home page. I've certainly been doing that, in my testing so far.
That's part of the elegance of the tool. Once installed, the Google home page will show a new "Desktop" link. This effectively integrates your desktop into Google itself.
Do a web search, and any matching content from your own computer will be shown above the regular search results, in a OneBox display, similar to how news, product, local and book search results are shown. Any one of these OneBox display may also still appear below any desktop search results.
Should you dislike this integration, you can hide desktop results on a one-time basis by clicking on the small Hide link within the display. Using the Desktop Preferences option, you can also shut integration off permanently.
The integration means you can easily spot any of your own emails or data files that might also match something you seek from across the web. I haven't found that too helpful so far. But the ability to have relevant web pages you've previously viewed be bundled as part of Google web search results is fantastic. Its helps you find new things you want to view plus recover things you've seen before.
I've written recently that search memory features like this at a9 have gotten me to use that service more. Similar features released recently by Ask Jeeves and Yahoo are also compelling. Overall, I was beginning to seriously dislike Google for not having them.
Now Google's gained some search memory of its own. a9, Ask Jeeves and Yahoo's tools are more mature and feature-rich, but Google Desktop Search is a good stopgap for Google. It makes my searches there more personal, more useful and importantly, helps tie me in more to the service. In addition, I get the ability to scan for files and email on my computer.
The integration feels so comfortable that it makes me think desktop search tools from Google's competitors will have to have similar integration. A desktop search tab may become de rigueur.
Google Desktop Search is a further move toward what I speculated might happen in my Welcome To The Google Desktop? article back when Gmail was released.
In it, I suggested that Google might cause us to reconsider what we consider to be our desktop. Rather than it being tied to a physical computer, our desktop could go virtual, with files located on Google (or competitors), accessible to use wherever we are.
Google Desktop Search doesn't physically get us there. Our files still reside on our computers. But metaphorically, those running it now have their desktops moved to Google. There it is -- a little link right above the search box. Your desktop, on Google.
Forget the idea of Google as operating system replacement. This isn't a move that locks you into a particular platform for running programs and applications, as an operating system does. This goes beyond that. It locks you into something more important, your data -- and perhaps prepares you for trusting Google (or others) more with that.
Consider that down the line, Google might offer to mirror its searchable copy of your desktop data on its own site. That would be useful. If you're away from home, recovering all your data would be as easy as getting to a browser and searching on Google. Your desktop could become wherever Google is -- or its competitors, if they follow suit.
While Google may seem in the lead on this, others may not be far behind. The Yahoo desktop search product that's rumored might involve storage of files with Yahoo itself. LookSmart's Furl service recently expanded web page storage to 5 gigabytes and envisions allowing file storage and searching. Lycos UK last month launched an online drive service for file storage. And Microsoft itself has a long-standing Stuff I've Seen research project that could potentially expand this way or be bundled into the desktop tool planned for release later this year.
Perhaps a virtual desktop via Google is something that will come. In the meantime, access to searching your own data remains available via Google, as long as you are at your own computer. And that data, that searchable index, is all bundled up into something called the Google Desktop Cache.
It's not a very elegant name. A better one might be PersonalWeb, which John Battelle coined earlier this year as something that stores everything you've searched for and seen. Or you could go with search workspace, another term that Jeremy Zawodny coined recently.
Whatever the name, the point is there's a big giant repository of data on your computer with copies of your data, from email to web pages you've viewed. It lives independently of them. Delete an email, delete all your temporarily cached internet files, and that data will still live on in the Google Desktop Cache -- or the GDC, for short.
This is both reassuring (hurray, I've got a backup!) and scary at the same time (uh oh, that was supposed to be deleted forever!).
On the deletion front, you'll need to remember that anything you remove from your computer still stays in the GDC. If you want it gone from the cache, you'll need to search for the item, then use the "Remove items" link next to the search box to permanently delete it.
Unfortunately, there's no easy way to wipe out lots of data at the same time. Even if you try things like searching for the directory path containing many sensitive files, you still can only remove ten items at a time.
Deletion control needs to improve. One nice feature might be that any time you delete something from your computer, Google Desktop might prompt you to delete the item from your cache at the same time. Google says it's considering something like this.
How about backing up? After all, if the GDC goes, your web history, IM chat and cached versions pretty much will go with it. They won't be recreated when your hard drive is reindexed, even if you have all the same core data.
You could try making copies of the cache from time to time. The data is located within the Documents and Settings folder, then your Windows user name, then Local Settings, Application Data, Google and finally Google Desktop Search. That last folder seems to have everything.
If your computer crashed, potentially you could reinstall all your programs, data and the Google Desktop, then restore the cache to the right place before indexing. Once there, switch indexing on, and you MAY be able to get back both your history and go forward with new data recording. MAY, I say -- I haven't tested this!
I've not used instant messaging much until about a week ago. Without broadband until recently, it just wasn't something I found useful. Consequently, the need to record and index my instant messaging conversations didn't seem that crucial.
Well, plenty do want it, if only from some of the tools I found designed to let you log and search conversations. Those using AOL Instant Messenger now have a solution via Google Desktop Search.
AIM allows you to save conversations. If you've done this in the past, then these will be indexed as part of a chat search. Going forward, saving isn't required. Google Desktop will automatically capture your conversation and make it searchable.
What about other instant messaging clients? Google says it went after AIM first because it has the largest network and most activity processed daily. Others might be added in the future, based on market share. And while Google did notify AOL that it was working on AIM indexing, the technical work was all done by Google itself.
As for email, one of the cool things about Gmail is how it will present email correspondence in related threads, when that makes sense. Now Google Desktop Search brings the same thing to your Outlook or Outlook Express email.
How about other email clients? As with IM, market share will dictate what other programs may be supported.
Web mail systems, like Yahoo Mail and Hotmail, are supposed to be supported to the degree that anything you read in your browser is recorded. However, the email messages will be stored as web history records, rather than as email records, since they were viewed on the web. In addition, anything you don't actually view will remain unindexed.
Ironically, Gmail content is NOT being recorded even as web history records. Google says it is working on this. Potentially, Google could even make it possible to search Gmail as if it was part of your physical email client. However, the company doesn't appear to have immediate plans to do this.
How about the increasingly popular Firefox browser? If that's your default browser (Tools, Options, General then check the Default Browser box), then the Google Desktop icon will open up a Firefox window. You'll also see the Desktop link if you visit Google.com. However, the web history of what you view within Firefox doesn't appear to get recorded.
Making Money & Development Background
Where are the missing ads? After all, doesn't Google place them wherever it can these days? For the moment, Google Desktop Search is free of them.
"We don't have ads on Desktop yet. Largely we're looking at this as a distribution model. If they use Google Desktop, they'll use Google web search more," said Mayer. "First and foremost, we're addressing the user need."
This applies to what happens within the specific desktop search area, however. Do an ordinary web search, and desktop results are integrated, as mentioned above. Ads are also shown as part of that web search, just as they always have. They only go away when you drill-down inside your desktop matches from the web search page.
I've got no doubt we'll see the ads come down the line within desktop search, if Google feels they'll be acceptable and in appropriate ways. If contextual ads can be targeted against email in Gmail, there's no problem with making them show up within Google Desktop. Similarly, ads could be positioned to show up based on the content of web pages being viewed from your history. Keyword-targeted ads could also be integrated into results.
As for the Google Desktop project, it has been in development for the past year, Google said. The first testing version went out to some within Google in February. Further versions followed, with a release candidate sent to 2,000 Google employees and trusted friends and family.
When news leaked of the project back in May, it was said to be code named "Puffin." That was actually the name of the Google Deskbar applet released last year. Code name for the Google Desktop project? "Fluffy Bunny."
A Wish List
I've been running Google Desktop Search for nearly two days, and already it's proven itself a keeper to me. Having said this, and bearing in mind it's still a beta release, I've already got a wish list.
Most annoying so far is that you can only see 10 results at a time. Though the tool has a Desktop Preferences option, that doesn't yet include an option to increase the number of results seen at one time. Google said there's nothing to announce about potential changes to this yet.
Also missing is an advanced search page. On that, it would be nice to have drop down boxes as with Google itself letting you limit searches to particular file types, phrases or especially date ranges.
For example, getting back a long list of matching emails, then having to use the result page numbers at the bottom of the results list to browse to a particular date is a pain.
You can use some of Google's search syntax to get around this. To narrow to file type, use the filetype command. For example, this:
would bring back only email matches with the word cars mentioned. A list of known file types we've tested to date and found work are:
- Word: filetype:word or filetype:doc
- Excel: filetype:excel or filetype:xls
- PowerPoint: filetype:powerpoint or filetype:ppt
- Text: filetype:text or filetype:txt
- Email: filetype:email
- Chat: filetype:chat
- Web History/HTML Files: filetype:web or filetype:html
- Images: filetype:jpg or filetype:gif
- Acrobat: filetype:pdf
- Windows Media: filetype:wma or filetype:wmv
- MP3: filetype:mp3
For email, chat and web history, you can also narrow by clicking on the count numbers, as described above. As for images, Acrobat and other files, keep in mind that only text in the file names will be matched, not any meta data or actual text contained within the files.
Images, Acrobat, Windows Media and MP3 files are also not officially supported by the product as searchable content. While I did find it capturing some of this content on my computer, the bulk of it was not retrieved. Why some but not all of it was found is unclear, especially given that the index process appeared to have completed OK. But since this isn't even promised, I can't complain much.
Google, of course, purchased the Picasa photo indexing solution earlier this year. Perhaps it will be that Google Desktop Search will evolve some integration with that in the future.
Google Toolbar/Deskbar integration would be nice. At the moment, I can use the Google Toolbar to specifically search for just images, news, shopping, the web and so on. But desktop search isn't an option. Google says this may come for future versions.
A real personal wish is for indexing of content of compressed/zip files. I constantly zip up data -- right now, none of that gets indexed by the tool.
I desperately want the search memory features to mature. I want this tool -- or some other system -- to keep track of what I actually have searched for and track that in association with pages I've found. So far, Yahoo's implementation of these types of features are the best I've seen (but those at a9 and Ask Jeeves are great as well).
Overall, I love the tool. Gary Price, who worked with me closely on this article, loves it as well -- and Gary is as cynical than me about new things. Anyone who uses Google will want to install it, if only because it's so easy to do and will likely improve their web searching experience.
Certain parts need to improve, of course. I have two dedicated email search tools, my long-time favorite NEO and Microsoft's newly acquired free Lookout. I wouldn't give either of them up yet, because they can still do somethings better than Google Desktop Search. Nevertheless, Google Desktop has proved more than good enough for many of my queries.
That's going to be a key thing. For those who have no desktop search at all, this product is a great start down that road. It will be a major improvement for them -- and another thing that will tie some closer to Google.
Meet Your Favorite Search Engine Watch Contributors
Many of SEW's leading expert contributors will be at ClickZ Live, the new online and digital marketing event kicking off in New York (March 31-April 3). Hear from the likes of: Thom Craver, Josh Braaten, Lisa Barone, Simon Heseltine, Josh McCoy, Lisa Raehsler, Greg Jarboe, Dan Cristo, Joseph Kerschbaum, John Gagnon, Eric Enge and more!