The Little Engine That Could - Part II

[Mike Grehan]

Originally Posted by rustybrick

I was going to wait, but I hear Mike Grehan has a new article coming out in the next few says with Jim Lanzone and the man behind the technology, Apostolos Gerasoulis. I am pretty excited to read it.

Barry Brick,

I have so much going on at the moment but I'm trying desperately (make that "I am bullying Christine") to get the periodical out tomorrow.

Both Jim and Apostolos were great and they did address many of the concerns mentioned here. Plus some very, very interesting take aways for the industry.

Bear with me... phooh, phooh, blows the smoke off his keyboard

[Robert_Charlton]

Originally Posted by rustybrick

Lets just assume they only put three sponsored (paid) ads at the top.

Now, would that change everyone's opinion?

No, at least not mine.... I just ran a bunch of commercial searches on areas I watch closely on Google, Yahoo, and MSN... as well as a utilitarian search I ran on Google this afternoon before a trip to the hardware store.

For me, the Teoma/AJ results were bizarre across the board. (Note that I use Teoma instead of AJ. Results are the same, and I can't stand the framing on Ask. Also, the Teoma interface shows me more things that interest me.)

First, the utilitarian search... which was for fluorescent tube flickering. I had a flickering tube problem and wanted to fix it. A bunch of the top ten Google results told me what I needed to know. I didn't even want to waste my time clicking on any of the Teoma results.

Here's Google...
http://www.google.com/search?hl=en&q...o ogle+Search

Here's Teoma.
http://s.teoma.com/search?q=fluoresc...=39&submit.y=9

Note that the Teoma results return a lot of Asian fluorescent manufacturers and suppliers. Now, Teoma talks about Subject-Specific Popularity. Well, fluorescent lighting is virtually the standard lighting in Asia, and it's clear from these results that the "communities" on this subject are skewed in that direction. So are the serps, in a rather strange way.

I looked at some commercial searches in the travel area, and I have to say that the sites returned often verged on sleezy (I mean, more so than I'm used to for travel ). Looking at the Teoma page, I think I see why.

If the "link collections" returned on the right are illustrative of the communities Teoma/AJ references, and they seem to be, that's the problem. These are mainly just huge collections... lots of links, mostly low quality pages. Apart from DMOZ, if these are Teoma's authority sites, they're bringing the whole thing down. There need to be some extra quality factors in the algo to screen the members of the communities in the first place.

This is admittedly a very quick and superficial look at a very complex algo, but I've seen pretty much the same thing... with some occasional surprising exceptions... much too often when I've searched on Teoma.

Add to this all the rest... stale index, can't handle 301s, too many ads and paid results, Ask's frame, very slow response, small index size, and I feel they've still got an uphill battle. I'd love to see them make it. When they first launched, they were impressive.

[orbbital]

In Firefox I get 3 sponsored results after the product ad.
In IE I get 5 sponsored results, which combined with the product ad takes up my entire screen area. Generic results require scrolling.

<rant>
I reckon that they have cottoned onto the idea that it is generally going to be experienced users that use firefox, and will generally disregard sponsored results (I do most of the time). But take your Joe Average home user. He does a search for Ipod on his standard Windows & IE based pc, and gets a page of results. He clicks on the one of the first 3-5 links, puts his 2 cents in Jeeves pocket, and is on his way.

Its really just a way for Jeeves to cash in on the uneducated users out there.
</rant>

[I, Brian]

Originally Posted by orbbital

<rant>
I reckon that they have cottoned onto the idea that it is generally going to be experienced users that use firefox, and will generally disregard sponsored results (I do most of the time). But take your Joe Average home user. He does a search for Ipod on his standard Windows & IE based pc, and gets a page of results. He clicks on the one of the first 3-5 links, puts his 2 cents in Jeeves pocket, and is on his way.

Its really just a way for Jeeves to cash in on the uneducated users out there.
</rant>

That would just make the TV ads just over-expensive PPC campaigns. Which is precisely why they need to tone down on sponsored results if they want people to take the organic search seriously.

The user in the above example - if trying for the first time and not finding the result that he/she needs, has a good probability of moving on. Wasted TV cash.

You guys are so focused on on how many sponsered links are on Ask.com. Ask Jeeves also owns MyWay, MyWebSearch, MySearch, iWon, Excite, Bloglines and Maxonline. Success for Jeeves isn't just the number of sponsered links on Ask. They have multiple brands that all have to be successful and some have more users than Ask.com. Alexa ratings list MyWay and MyWebsearch with higher user numbers than Ask. And Bloglines is growing much faster than Ask in usage and iWon has more searches per month per user than Ask. Why is the number of sponsered links on Ask the whole issue on this thread?

I think Jeeves is the industry leader in 2 areas. RSS feed aggregators with Bloglines and search toolbars with MyWebSearch. I think these 2 areas are going to be very big growth areas in the future of search. Blog search is something that is a natural for Bloglines and Teoma. Downloadable search applications might end up to be the biggest growth area in search but is practically ignored on these forums. Search toolbars, desktop search, search branded browsers etc all could have a huge future in the industry.

These 2 areas - Blog search and downloadable apps are areas where Jeeves can take the search industry by storm and lead the industry. Thats where Jeeves focus should be if they want to take market share from the big players - Google Yahoo and MSN. They have a shot in these 2 areas.

[andrewgoodman]

Originally Posted by rustybrick

I'll try to summarize what I wrote at my blog entry named Ask Jeeves Rockin' These Days here.

Since the inception of this thread Ask Jeeves has done a lot, including:

(1) TV Commercials
(2) Discussions about Open Sourcing Search
(3) Acquiring Bloglines
(4) Increased communication through Ask Jeeves Blog and Jeeves Coming Out

Now I see that they have removed the 7 of the 10 top sponsored results. They have place only 3 top sponsored results at the top and then is showing those juicy natural results.

Check out a search on ipod for example.

I want to say this in the nicest possible way as I do know Jeeves is sincere about the search side as proven by some of these recent initiatives.

However to play devil's advocate specifically on the removal of 7 of 10 sponsored links, I think this example actually shows that ASKJ is monetizing their search just as visibly and "intrusively" as previously. The large banner from pricegrabber coupled with the spaced-out "top three" sponsored links seems to take up about as much screen real estate as the ten text links used to do.

Since the top three links generate the highest CPC's I doubt they are taking a revenue hit. Where few web results are above the fold, it still looks to me like the average user will have to hunt a bit to get some "search" with their search.

[andrewgoodman]

Originally Posted by MUSCLE13

You guys are so focused on on how many sponsered links are on Ask.com. Ask Jeeves also owns MyWay, MyWebSearch, MySearch, iWon, Excite, Bloglines and Maxonline. Success for Jeeves isn't just the number of sponsered links on Ask. They have multiple brands that all have to be successful and some have more users than Ask.com. Alexa ratings list MyWay and MyWebsearch with higher user numbers than Ask. And Bloglines is growing much faster than Ask in usage and iWon has more searches per month per user than Ask. Why is the number of sponsered links on Ask the whole issue on this thread?

So once again you explain to us that Ask Jeeves is a holding company. I am still scratching my head wondering why someone might get so excited about a holding company! Cheering some dot com stock ticker on because the whole bundle of its owned properties can boast x million page views is so 1999.

I mean, why not talk about, say, Thomson Corp.? They own about 1,000 diverse information publishing companies.

People here talk about Ask Jeeves and Teoma for what I think are obvious reasons. I can't see any obvious reason for anyone here to be interested in iWon - so they don't talk about it.

Originally Posted by andrewgoodman

So once again you explain to us that Ask Jeeves is a holding company. I am still scratching my head wondering why someone might get so excited about a holding company! Cheering some dot com stock ticker on because the whole bundle of its owned properties can boast x million page views is so 1999.

Andrew - Tell me how I am cheering on Jeeves by saying they have a huge challenge competing against industry giants in Google Yahoo and MSN who all have huge dollar resource advantages to market their product? Jeeves has to go about it in a multibrand approach in my opinion.

I am still scratching my head Andrew on how you as an industry observer who writes about Jeeves doesn't have any sense of what the company is about. I remember you writing on Traffick about how the key to the ISH deal is Excite. JEEVES DOES NOT OWN EXCITE SEARCH. You guys write about Jeeves and don't ever consider the fact that MyWay and MyWebSearch have MORE USERS THAN ASK. You put down iWon like its a piece of garbage saying its not search related. iWon users search on that site DOUBLE OR TRIPLE the amount of times that Ask users search on Ask.com every month. ISH properties have MORE QUERIES each month than Ask. Andrew I really have no issue with Barry or Danny or Mikkll or any of the other SEM guys here. My opinion is you are posting here about a company that you really don't understand at all. Thats why I post here. There are ton of blog writers that claim to be experts on search companies. Some are great like Search Engine Lowdown and Battelle's Search Blog. A lot of them don't do research and think that Excite is the main part of the Jeeves ISH deal, and anybody who enjoys following search outside of SEM is nothing but a stock pumper. Do some research.

[rustybrick]

We understand search. Ask Jeeves, I believe, is a search company. If not, then one aspect of their business is search. I do not research companies from a perspective if their stock will do well or not.

All I am looking for is a search company that will provide the best user experience.

Are we so bad? I know, you did not mean any of it personally. But, I can speak for myself. I am not looking out for Ask Jeeves financial growth or stability. All I care about is the Ask Jeeves search experience. I know very little about how well a stock will perform or if Ask Jeeves will be around in 10 years from now.

Barry - You know I think you are a great guy. You guys understand search. Do you understand that the majority of queries for Jeeves come from MyWay, MyWebSearch and iWon and not from Ask? If so why is every post about Ask.com on this thread?

[rustybrick]

No, I don't understand that. Sorry. But I never claimed to.

We will see, give it a year.

Also Barry, like I stated before there are a ton of guys that know the search business. I think Beal, Battelle, Danny, you and others are all ok. My problem is with guys posting about companies they don't understand. I mean when I read about stuff like Excite is gonna be the best part of this ISH deal for Jeeves when INFOSPACE OWNS EXCITE SEARCH it makes me wonder what people are thinking?? Its nonsense.

[rustybrick]

Originally Posted by MUSCLE13

Also Barry, like I stated before there are a ton of guys that know the search business. I think Beal, Battelle, Danny, you and others are all ok. My problem is with guys posting about companies they don't understand. I mean when I read about stuff like Excite is gonna be the best part of this ISH deal for Jeeves when INFOSPACE OWNS EXCITE SEARCH it makes me wonder what people are thinking?? Its nonsense.

OK, but this thread is meant to be about the Ask Jeeves appeal to the SEM community.

Originally Posted by rustybrick

No, I don't understand that. Sorry.

We will see, give it a year.

Barry - Jeeves revealed those query numbers when they made the deal. Its a fact. And in the quarter that followed the deal ISH grew faster than Ask in queries so the spread got even wider. Look at the numbers.

Originally Posted by rustybrick

OK, but this thread is meant to be about the Ask Jeeves appeal to the SEM community.

Barry - This the question you posed in the first post of "The Little Engine that Could"

"Can Ask Jeeves get beyond their current market share? Do they want to or are they happy with their current position?"

If a majority of searches for Ask Jeeves are coming from MyWay, MyWebSearch and iWon, these properties should not be addressed in answering your question? Just Ask.com?

Barry I don't want to get into an argument with you. I think you are probably the most open minded guy here. My problem is with guys like Goodman who laugh off posts when they themselves have done no research at all.

[Mel]

As a FireFox use I see only the box at the top of the page and exactly zero sponsored results.

Way to Go Jeeves!

[dannysullivan]

Do you understand that the majority of queries for Jeeves come from MyWay, MyWebSearch and iWon and not from Ask? If so why is every post about Ask.com on this thread?

Because Ask is the leading brand. Ask isn't going to be on TV putting out iWon or the other brands it owns in front of audiences. As I've posted above, the other brands it bought in my view have already declined. It grabbed some valuable traffic -- and that's helping to fuel revenues -- but the long-term support of those other brands doesn't seem there.

We simply have not seen any major player do a good job in maintaining support for both a flagship brand and other brands -- Excite is the classic example, killing off WebCrawler and Magellan. Yahoo killing off AltaVista and AllTheWeb. Lycos largely killing off HotBot and only pitching it to revitalize it when the Lycos flagship itself was diminishing.

Maybe Ask will come around and keep all these other brands alive. History doesn't have to repeat -- but neither do you ignore it.

Ask is also likely the largest traffic draw of all the holdings. Latest comScore stats are here. Excite Network has a 4 percent share; Ask 2 percent. But Excite is made up of Excite, iWon, MyWay and MyWebSearch. Do any of those have at least a 2 percent share alone? I don't know, as we don't have those stats -- but it seems unlikely. If they did, you'd think Ask would get behind one of the other stronger brands.

As for Excite, Infospace does NOT own it, to my knowledge. It's part of the Excite Network which Ask Jeeves owns. I believe Infospace continues to operate the "search and directory" components of Excite, which is what I think you also understand when you said "Excite Search."

So Infospace may own -- or may be entitled under the agreement -- to keep powering the search at Excite. As a result, that 4 percent Excite network chunk I mentioned may not be entirely all in Ask's control. Then again, Ask Jeeves almost certainly gets a share of some of the search revenue at Excite. But I'll check. The deal at Excite was always incredibly complicated because of this strange deal of the Excite search component.

You put down iWon like its a piece of garbage saying its not search related. iWon users search on that site DOUBLE OR TRIPLE the amount of times that Ask users search on Ask.com every month. ISH properties have MORE QUERIES each month than Ask

According to what. I've not seen any doubling or tripling of Excite Network search traffic over the past few months. It's been pretty static. And if Ask doesn't control the searches at Excite itself, the fact that this traffic is mixed with someone like iWon makes it impossible to know what exactly is happening at iWon. In other words, we know Ask has a 2 percent share. That other 4 percent share for the Excite network includes some Excite search traffic that we don't even know should be counted to Ask. The remaining unknown gets split among the other properties that may -- or may not -- exceed Ask's 2 percent share.

Why is the number of sponsored links on Ask the whole issue on this thread?

To conclude, because that's the leading brand. And when Ask puts that brand out via television and other ads to compete with the likes of Google and gang, it gains more attention. And if the number of ads are heavy compared to the others, it's going to get criticism in some quarters -- say places like Consumers Union, that looks and puts out reports. They aren't going to review iWon. Ask will take the heat. You needn't agree with this, but others are not necessarily foolish or ignorant because they are focused on the Ask brand. They may have their own reasons for seeing it as important.

[rustybrick]

As promised, Mike Grehan delivered:

Mike Grehan in conversation with... Apostolos Gerasoulis and Jim Lanzone

It's been quite some time since we last had an "in conversation with" feature (but don't worry, there are more in the pipeline).

I'm particularly pleased (and honoured) with this one. Not only did I get to speak with Jim Lanzone, Senior Vice President, Search Properties at Jeeves, I also got to speak to Apostolos Gerasoulis, former Professor of Computer Science at Rutgers University and founder of the Teoma search engine (which powers Ask Jeeves). read it here >>

[Mike Grehan]

It's come to my attention that there was a typo in the (probably) most important quote of my transcript of the interview with Jim Lanzone and Apostolos Gerasoulis at Ask Jeeves.

At the point where I mentioned to Apostolos that, I've made Paul Gardi shudder by mentioning Kleinberg's HITS algorithm and Teoma in the same sentence, I then led into a question regarding the possibility of Kleinberg's algorithm being an influence on PageRank.

Let me just get everything into context here. As visiting professor, Jon Kleinberg worked on a further development of his algorithm at IBM's Almaden Research Center in San Jose. Here, the project was known as CLEVER (Client-side EigenVector Enhanced Retrieval) and it ironed out some of the wrinkles in the original algorithm. But the one final piece that couldn't be solved, was how to get the query response time into sub-second speeds.

It was Apostolos who solved the problem and it's his solution which enables Teoma to do just that.

So, back to the interview. When I asked Apostolos: "Is it fair to say that Kleinberg's work was also an influence on PageRank?"

The answer was transcribed as being: "There is some debate about this. I think both of them were independently discovered. Has PageRank influenced Google? The answer is: later, yes. Have they implemented PageRank? The answer is no."

But his actual answer was: "There is some debate about this. I think both of them were independently discovered. Has CLEVER influenced Google? The answer is: later, yes. Have they implemented PageRank? The answer is no."

To have someone as esteemed in the industry, apparently misquoted in this way, on such a highly topical matter, deserves an apology. And I'm happy to apologise to both Jim and Apostolos on behalf of e-marketing-news. It was a pure typing error. But the fact still remains: it's not likely that Google has used PageRank

As for the idiot who typed the transcript, he should be taken to one side and thrashed severely with a fearsome stick of Rhubarb. However, it may not be a wise idea to ask room service at the hotel I'm in, if they could pop up and oblige!

[Both the HTML version and the pdf version have been amended for those who'd like to download the correct version]