Biznology Blog: January 2006

« December 2005 | Main | February 2006 »

January 31, 2006

More on Audio Search

Just a few days after I blogged about Podzinger, John Battelle has a story about Nexidia, an interesting search engine because it uses a different approach. While Podzinger uses speech recognition technology to translate audio into words and then indexing them using standard text retrieval techniques, Nexidia skips the speech recognition step—storing each sound as a phoneme (a unique speech sound) and indexing all phonemes.

I know a lot about text retrieval technology, but not much about audio information. So, I need to make some educated guesses about the advantages of Nexidia's approach.

My guess is that by translating the search words entered as text to phonemes, the matching process is far more reliable because translating text to phonemes is far more accurate than translating phonemes to text (speech recognition). Why is this true? One reason, especially important in search applications, is proper names are frequently missing from speech recognition dictionaries and so are often recognized incorrectly or not at all. Any word that is not correctly recognized won't be matched at search time to its correct counterpart.

For applications where searchers are typing their search queries as text, I would think that a drawback to Nexidia's approach would be that speech recognition would still be needed to display the title and text of the audio file in the search results page. If Nexidia truly finds the best matches, however, searchers would still be better off, because speech recognition approaches have the same problem. Basically, a speech recognition approach has errors in both matching and snippets while a phonemic approach has those errors only for snippets.

Now, the basis for the claims of superiority for a phonemic approach is that it is more accurate than speech recognition, which also may be changing. David Nahamoo, a colleague at IBM, is demonstrating far more accurate speech recognition than has been seen before. As with most competing technical approaches, it's hard to tell which approach will work better, but there certainly seems to be a lot of innovation around audio search, which ends up good for searchers and for search marketers looking to attract visitors to their podcasts and other audio content.

Posted by mikemoran at 9:28 PM | Comments (0) | TrackBack

January 27, 2006

A New Way to Find Podcasts

If you're using podcasts to get your message to prospective customers, you should know about a whole new way for people to find your content. A new podcast search engine offers a breakthrough that may make podcasts easier to find than ever.

Up until now, podcast search was no better than image search, which depends on searchers using the words around the image file, or the name of the file itself. The search engine does not know that it's a picture of a zebra—it is counting on the image being named "zebra.gif" (for example). Podcasts, until now, were similarly second-class content, compared to first class Web pages (and blog entries) that can be found by entering any word found on that page.

A new search engine, PODZINGER, takes a new approach, using speech recognition technology to convert podcasts (or any audio) to text. Speech recognition software can "listen" to audio and identify most of the words spoken, converting it to plain computer text, the same as any Web page. Once converted, any text search engine can find that audio file based on any word spoken.

PODZINGER is a neat tool that can help you find podcasts, but search marketers are waiting for mainstream search engines to offer speech recognition—when Google and Yahoo! can perform this trick, the average searcher will start to find podcasts. The development will cause a huge increase in podcast usage. You may even see it happen this year.

Posted by mikemoran at 9:12 PM | Comments (4) | TrackBack

January 26, 2006

Replay of Yesterday's Interview on wsRadio.com

If you missed the live broadcast, you can still listen to a replay of my interview yesterday on wsRadio—you'll hear all about the basics of search marketing along with special tips for the search marketing impact of blogs, podcasts, and RSS feeds.

First segment

Second segment

Third segment

Posted by mikemoran at 1:35 PM | Comments (0) | TrackBack

January 25, 2006

Interview in BtoB Magazine

Here's an interview that I did a few months ago but that was not posted to the public Web until recently. BtoB Magazine asked me about the future of search marketing, the biggest mistake most search marketers make, and more.

Posted by mikemoran at 10:36 AM | Comments (0) | TrackBack

January 24, 2006

Tune in Tomorrow

Tomorrow at 10 am Pacific Time, I'll be the guest of "RSS Ray" on wsRadio, on the eCommerce RSS Radio Show. I'll discuss the basics of search marketing, and how RSS feeds are particularly helpful to your search marketing efforts. I'll answer callers' questions, too, so let's hear from you. After the show airs, I'll post a link for you to download a replay in case you miss the live broadcast.

Posted by mikemoran at 7:12 AM | Comments (0) | TrackBack

January 23, 2006

Column for Revenue Magazine

Starting in May, you'll see my words in a new place—a column in Revenue magazine called, "The Searchers." Revenue magazine, which is targeted to Internet and especially affiliate marketers, distributes 125,000 copies of each issue and has just increased its publishing frequency to bi-monthly. I am excited to be part of this growing publication.

Check out the official word in the Revenue blog.

Posted by mikemoran at 11:36 PM | Comments (0) | TrackBack

January 21, 2006

Creating Your Content Improvement Scorecard

The content on your Web site may be the most important factor in improving your results in organic search marketing (and your Web site search, too). And many aspects of your content must be monitored--keyword density, titles, descriptions, and more. But it's one thing to set standards in your organization and to train the people responsible, another to actually check that they comply.

Everyone tries to set some kinds of standards for their content—even if you just hang up a list of tips on the wall for you to remember when you write for your small Web site. For a larger site, your standards are more formal—you may even have committees to ratify changes for especially large sites.

Large or small, the standards say a lot of the same things when it comes to improving your content for search. Use your keywords in titles, descriptions, and body text. But how do you enforce those rules?

One way is to use a content reporter—software that creates a content scorecard that shows compliance of each page against your site's standards. Your scorecard lists which pages contain titles and descriptions and which don't—missing titles and descriptions are common searchability errors. With a little more work, you can check that each page has a unique title and description (not one used on other pages). With just a bit more work, you can check keyword density (the number of keywords compared to total words) for each page.

If you have less than 100 pages on the site, your scorecard can be a simple list of all pages with columns calling out whether titles and descriptions are present and are unique. With thousands of pages, you'll want reports that roll up totals organizationally—"93% of the pages in the Consumer Products Division comply with standards, but 7% do not."

A content reporter has three major parts:

Spider. If you have a site search engine, it likely has a spider (crawler) that discovers all pages on your site. If you don't have such a spider, you can try to customize a spider (such as Xenu) or you can build your own. Developing a spider is not an easy task, but you can check out Spidering Hacks to get some tips on how it's done.

Analyzer. After each page is crawled, you need software to pore over every tag and every word to test compliance to your standards. Your analyzer can make sure that the title and description tags have text. It can store the text from these tags in a database so that it can check that no other page contains the same text for those tags. If you carefully encode your top keywords (in order) in your keywords tag, your analyzer can check that those words are present in titles and descriptions—and can check the keyword density in the body text. Most site search engines have a content pipeline that allows you to insert your custom analyzer. One site search engine, IBM's OmniFind, has an open framework called Unstructured Information Management Architecture (UIMA), that makes it easier to insert your analyzer.

Scorecard. Once the analyzer has checked each page, you can use Crystal Reports or a similar program to produce your scorecard. Then, use "management by embarrassment" to go after the managers and writers of each group of pages that exhibits low compliance. Over time, they'll correct their content if only to get good grades on the scorecard.

If you have a very large site, it might make sense for your to purchase and end-to-end system, such as the one from Watchfire—it's pricey but it can be worth it for a site with thousands of pages. Regardless of whether you build or buy your content reporter, it's a critical tool to improve content for searchability across your entire site.

Posted by mikemoran at 11:44 PM | Comments (0) | TrackBack

January 20, 2006

New Book Review

We're gladdened to receive another strong review of our book, Search Engine Marketing, Inc.—this review is by Andreas Ramos of Adwords-User-Group.com. Remember, the price was just dropped to $29.49 on Amazon. (Subtle hint, huh?)

Posted by mikemoran at 8:19 PM | Comments (0) | TrackBack

January 14, 2006

Another Amazon Sale

If you missed the discount on our book, Search Engine Marketing, Inc., a few weeks ago, don't miss Amazon dropping the price under $30 once again. It's now selling for $29.49 with free shipping.

Posted by mikemoran at 9:16 PM | Comments (0) | TrackBack

January 12, 2006

How to Embarrass Your Company in One Easy Step

Just in case you don't have enough reasons to stay away from unethical spam practices, here's another one: you can embarrass your company. You can bring shame on your company no matter who you work for with these black hat techniques, but large companies, in particular, are vulnerable to having their spam exposed—what big companies do is newsworthy.

Spam is hard work—even harder work if you intend to keep varying your techniques to stay ahead of the search engines—but the bad publicity you can bring on your firm is a factor that few consider.

Posted by mikemoran at 9:45 PM | Comments (1) | TrackBack

January 9, 2006

Upcoming Radio Interview

Just a quick note to let you know that I will be appearing on wsRadio, on the eCommerce RSS Radio Show. If you are interested in how RSS feeds can help you with search marketing, tune in to wsRadio at 10 am Pacific time on January 25.

Posted by mikemoran at 7:20 AM | Comments (0) | TrackBack

January 8, 2006

Web Site Search Just Needs Smarter Users

If you manage the search facility for your Web site, no one has to tell you how many complaints you get from frustrated searchers who can't find what they are looking for. You're frustrated too as you struggle to improve things, but don't turn on your own customers. January's Biznology newsletter asks you whether you believe that "Web Site Search Just Needs Smarter Users."

Posted by mikemoran at 6:56 PM | Comments (0) | TrackBack

January 2, 2006

The Mobile Search in Our Future

As a frequent conference speaker, the same questions from the audience tend to come up repeatedly—one of them is, "What are the future trends in search marketing?" One of the trends that I like to mention is how much more ubiquitous search will become in the next few years. We only think that we search a lot now. What will it be like when we use mobile devices that are always connected to the network?

As important as search marketing is to your business today, it will be even more important soon, if for no other reason than more people will be searching more often. Not only will people without computers today gain access to computers over time, but people who already have computers will be connected to the network more frequently than they are now, using portable devices such as cell phones, PDAs, MP3 players, and many more.

One of the things that will drive search usage from these devices is what designers call the "form factor"—how large the device is. You can imagine that browsing and navigating the large Web pages that exist today will never work as well on a cell phone's tiny screen as it does on a computer monitor. And while some researchers believe that information can be projected in front of your eye or on the table at Starbucks, that's farther away than people using the small screens that we already have. Those small screens just need constant connection to a network to become the search platform of tomorrow, and everyone from cell phone providers to wireless computer networks are figuring out how to make the network pervasive at a low cost. So, one simple reason that search will explode in the next few years is that mobile users will get connected and no way of accessing information will work better than search.

I came across a neat demo that gets us one step closer to more ubiquitous search—one that uses voice recognition to empower mobile users with the kind of information power that today you need to be seated in front of a computer to wield. XM Satellite Radio teamed up with VoiceBox to show how voice recognition can be paired with satellite radio to provide easy information access while driving—hands free and eyes free.

Imagine how this technology could work with a next-generation search engine that could answer questions, rather than just listing Web pages. Ask a question, get an answer—just as naturally as asking another person. Now it will be some time before search technology works as smoothly as it looks in that demo—the demo is confined to some well-understood subjects, rather than an unconstrained search in Google—but this demo shows you what will become possible.

Hat tip to Orbitcast for this nifty demo.

Posted by mikemoran at 11:48 AM | Comments (1) | TrackBack