Google Launches Catalog Search

In December, Google rolled out a completely unexpected offering: Google Catalogs. The new service allows you to search through the contents of catalogs from over 600 companies.

Here's the real twist. These are print catalogs, not online ones. To create this search index, Google scanned these catalogs and converted their printed pages into machine-readable text, using optical character recognition software.

Do a search, and you'll be shown catalogs with pages that contain your search terms. Each listing will show the cover of the catalog, an image of the catalog page where the terms were found, and a close-up of the section of the page containing the terms. Google even highlights your search terms within the close-up image of the catalog page.

The term highlighting was no easy feat. Remember, the highlighting is happening in images, rather than in HTML copy. That means Google couldn't scan for matching words to know what to highlight. Instead, Google has a map telling it the pixel coordinates of every word appearing on every page in the catalog index. When you search, it finds the matching pages, then displays an image of those pages and highlights the pixels within it that contain your search terms.

When viewing a catalog page, a helpful navigation bar appears at the top of the screen. It lets you easily "browse" through the catalog page by page, as if you were flipping the pages of a real catalog. In addition, you can also keyword search within a particular catalog or jump to specific pages within it, using the navigation bar. The bar also shows contact and other information about the catalog merchant.

My first thought upon seeing the service was "Why?" Of all the things Google could choose to make searchable, why target mail order catalogs? Shopping-oriented searches do make up a significant chunk of any search engine's queries, so launching some type of shopping search service does make sense. But why not instead target online merchant stores?

The answer from Google is that mail order catalogs provide more comprehensive product listings than can be found online and that making catalogs searchable was something it could do easily.

"A number of people [at Google” thought offline catalogs are much better than online shopping sites," said Urs Hvlzle, Google Fellow and member of Google's executive management team.

Explaining further, Hvlzle said a major drawback to print catalogs is that you generally only have a few in your home and there's no way to easily "search" within them. Google Catalogs solves this by letting you search through a virtual library of catalogs.

As for the work involved, the project started several months ago with less than two percent of Google's engineering team working on it. There was also the help of some high school students drafted in to complete some of the scanning. Overall, Hvlzle calls the creation of the catalog "a relatively small effort" to make something useful and unique.

In contrast, "[online” shopping search would be a larger effort, and there are lots of good places out there that do shopping search," Hvlzle said. "Certainly it's something thats already covered."


As I played with Google Catalogs, I realized it was more useful than I had initially imagined. I hadn't considered myself a big catalog shopper until I really thought about the various ones we have in the house. For example, there's our battered 1999 Ikea catalog. We still keep it around because every time we go to Ikea, they're always out of the current ones. Problem solved -- now I can head to Google and have access to the latest catalog.

Why not just go to the merchant's site? You can certainly do that. The vast majority of the catalog companies listed will indeed have web sites. The problem is, they may not have all their products listed. Or, even if they do, you might not be able to locate what you want. The quality of search and navigation on merchant sites is often poor.

For instance, there's been many times when I've browsed through my MicroWarehouse catalog to find some computer supplies, only to discover when visiting the company's web site that the same item in the print catalog doesn't seem to exist online. These are times when I'm thankful to have that trusty paper catalog floating around.

Clearly, Google Catalogs is a great shopping reference tool. However, it's a poor solution for anyone wanting to actually purchase online. For example, consider a search for "remote control blimps," which are helium-filled mylar balloons that you can fly around rooms.

Google brings up plenty of good matches, with the first being from Hammacher Schlemmer. Great -- I'm ready to buy, and I'd like to do it online. To do so, I have to find the product code in the image of the catalog page displayed. Then, I have to either write it down or remember it, so that I can enter it into a box at the bottom of the page that says, "To find an item on the Hammacher Schlemmer website, just type in the product code and press Go!" Doing so delivers me to the right page in the Hammacher Schlemmer web site, where I can purchase the blimp.

That's not too bad, but a direct link would be better. For example, over at, a long-standing site that features mail order and online catalogs, the same search also brings up a match from Hammacher Schlemmer. Clicking on that match immediately brings me to an order-enabled page. There's no need to enter any product numbers by hand. Of course, only listed Hammacher Schlemmer. In contrast, Google was far more comprehensive in its matches.

Another problem at Google Catalog Search is that less than 10 percent of the catalogs it lists even have an option allowing you to enter product numbers and reach an online order page. For the rest, if you don't see a box, you have two choices. First, you can visit the catalog's web site via the URL listed at the top of the catalog page, then try to find the product online. Second, and the option Google recommends, is to pick up the telephone and call in your order.

All this may improve, especially given that this is a "beta" launch by Google. It may especially improve quickly if Google follows-up with some advertising options, discussed further below. But so far, users don't seem to find the difficulty in fulfilling orders to be a problem, Google says.

"We havent heard from users, 'Why cant I shop with this,'" said Google spokesperson David Krane. Instead, Krane said the predominant feedback is that users seem to enjoy being able to search and browse for ideas, using Google Catalog Search as a research tool rather than an online shopping mall.


Another reason to target print catalogs rather than an online shopping sites involves money. The print catalog industry is more stable and lucrative than the online shopping industry. By targeting it, Google may have thought there was more long-term money to be made through advertising and partnership deals.

In fact, going after print catalogs in this way was such an unexpected move that it make you think there's someone at Google who knows the industry well and has contacts to get advertising deals rolling. However, Google says there was no particular "champion" behind the idea.

"This was more people thinking, 'Gee, if we had this, it would be useful,' rather than someone being an ex-catalog person, Krane said.

Nevertheless, Google clearly would like to make money off the new service and looks to be far more aggressive about that than when it rolled out its other "vertical" search services. For example, Google Groups and Google Images were launched without any sign that advertising options would soon follow on them. In contrast, an "Information for Catalog Vendors" page suggests ways that Google might aim to work with catalog vendors, in return for payment, such as through direct links to products.

These paid-for direct links don't yet exist, but should they actually be made available, it will be a fundamental change for Google. For the first time, cash would have a role within Google's editorial content, something the company has suggested is wrong when other search engines have done this.

Specifically, Google is the only major crawler-based search engine that does not have a "paid inclusion" program, and it is proud of this fact. In paid inclusion, sites pay to be included in editorial listings. They generally are NOT guaranteed to rank well for particular terms, as part of the deal. That remains left to the usual ranking algorithms. Instead, the paid inclusion programs are designed so that sites that pay have more pages indexed than might ordinarily happen when a crawler follows purely automated processes.

Google dislikes paid inclusion, because it involves money in the editorial listing process. In turn, that may cause people to question the quality of the listings they receive. To avoid this, Google keeps things clear cut. Money can get you a top ranking at Google, but only in the clearly marked "Sponsored Links" areas of the results page, which are just above and to the right of editorial results. Money has no impact on how the editorial results rank or appear.

Direct links in Catalog Search wouldn't change this policy. Google has no plans to let money influence which catalog results would get a top listing. Instead, direct links would simply help a user make that final connection between a product they are interested in and an online ordering option or more information directly from a vendor.

Would this be bad? Not at all. Indeed, it could even be good. Direct links would also help users, who want to click through to online ordering options, when these are available. Revenue from direct links also helps users another way, by keeping Google in business. Overall, done with care, money can play a helpful role, even in editorial areas.

Of course, that's exactly what Google's competitors, Inktomi, AltaVista and FAST, have been saying about their paid inclusion programs. And while direct links in Google Catalogs wouldn't be paid inclusion, the fundamental mix of some cash within editorial content remains the same. Implementing such a program will make things a little less "clear cut" at Google, a significant philosophical change.

It's important to note that this isn't yet happening nor may it actually happen. "The main goal of the vendor page is to stimulate conversation with major catalog publishing companies," Krane said.


One of the other ideas being floated is "Reporting Information," where catalog owners could learn how people located their products at Google.

"What are the keywords users enter to find your catalog? How many pages do they typically examine? What kind of clickthrough rates and sellthrough do online catalogs generate? Google can provide you with information about how your customers use your catalog in ways no other research tool can, all the while adhering to the strictest standards of individual user privacy," the information page says.

Hey -- that's a great idea! It's also one that web site owners have long been asking for, especially in light of Google's oft-repeated warnings for nearly a year now against using rank checking tools. Web site owners -- just like catalog merchants -- want to know how people are finding them. So why isn't this type of service being suggested to them?

"One fundamental difference is that with the catalog search, the sole purpose of catalogs is to sell products," Hvlzle said. "Nobody is going to go in and do a non-commercial search whereas web search is different. It is primarily non-commercial."

Hvlzle went on to further explain that Google is committed to making a "separation of church and state" when it comes to web search, to ensure that no one considers results there to be influenced by advertisers or partners.

The reality is that web search is plenty commercial. One only needs to look at some of the terms in Google Zeitgeist, which shows top searches on Google, to see this. Queries such as "windows xp," "lord of the rings" and "beatles" all represent at least some people who are after commercial content. Yes, there are certainly plenty who want non-commercial content, as well -- but to say web search is primarily non-commercial doesn't ring true.

That's especially the case given that Google makes about half its income by selling advertisements on web search results pages. How could this be, if most people want non-commercial information? Clearly advertisers are finding plenty of customers from Google who come there looking for commercial listings.

Ultimately, Google is considering web site owners to be second-class citizens, if it denies them reporting information while providing it to catalog owners. Certainly Google has faced spamming from some web site owners, while catalog owners will have far less opportunity to do this, within Google Catalogs. Such spamming has made Google concerned that adding a reporting option for webmasters would just make matters worse. However, given the demand that's out there, it would behoove Google to at least try a beta program and get some experience with it.


Google Catalogs currently includes catalogs from over 600 companies. Including back issues, the total number of catalogs indexed is 1,500. The company wants to raise this number further, soliciting catalogs from its users, from catalog vendors and even its own employees.

"There are donation boxes all around the Googleplex," Krane said. "It's amazing how many you can collect."

Google is not charging catalog owners to be included in the index and says it will include any catalog it receives by mail, within days. The only exception are catalogs that focus on liquor, tobacco, firearms, or similar products. Google didn't explain why it bans these type of catalogs, only to say that such restrictions match existing restrictions it has on ads about those products.

If many catalogs start flooding in, that "within days" promises on the Google web site might get longer, or Google might become more selective about what it includes, in the near term. Given this, here's some advice that I've run past Google, to help ensure your catalog is easy to process.

Obviously, send Google your catalog, but carefully remove the pages from the binding. That will make it easier for them to quickly scan it, if the catalog is selected. Also, include a short cover letter. BRIEFLY explain what your catalog covers and why it is unique. Perhaps millions of people read it, or perhaps only a few thousand do -- but you have products no one else has. Either point might help stress your importance.

What if you don't want to be included? You can ask Google to drop you by sending them an email. And has Google violated copyright by scanning all these catalogs -- some of which there is a charge for -- without permission?

"We believe that our approach is consistent with copyright laws," Google spokesperson Eileen Rodriguez said.

Google failed to explain further on how making wholescale copies of catalogs without prior permission is consistent with copyright laws. My guess is that the company is relying on the fact that those few catalog owners who don't want to be included will be satisfied by the ability to opt-out, rather than pursuing a court action.

Google is aiming to update its catalog index on a daily basis, so new catalogs may appear at any time. The company didn't say how many per day are being added.

It's also important to note that so far, the catalogs are for US-based companies. There are no immediate plans to add catalogs for countries outside the United States, but this is being evaluated, based on feedback from users.

Google Catalogs

Google Catalogs Help

Review this page to learn more about how the navigation bar works at Google Catalogs.

Information for Catalog Vendors

Explains advertising options being considered and where to mail your catalog to be included. Also use the email address on this page if you want to be excluded.

Suggest A New Catalog

Form allowing users to suggest their favorite catalogs.

Google Catalog List

Full list of catalogs that Google Catalogs includes.

Google Catalogs Advanced Search

By default, a search on Google Catalogs only looks through current catalogs. If you want back issues, use the Advanced Search page. Set the date option to "Search all catalogs, including past catalogs," in order to also include back issues. Want to see back issue from a particular vendor? Set the date option as shown, then search for the catalog, by name.

Google targets catalog business
MSNBC, Dec. 17, 2001

Nice stats here comparing the earnings of the print catalog world to online shopping.


AltaVista did have its own in-house online shopping service, but last November, that shifted over to being powered by DealTime. The deal also handed DealTime the domain that AltaVista previously owned. DealTime also provides shopping search services to Lycos and AOL.