The recent victory of IBM’s Watson computer against human competitors in an exhibition round of Jeopardy got computer scientist Stephen Wolfram thinking about how regular search engines might fare in such a match-up. So he took 200,000 known Jeapardy clues and ran them through six search engines (Google, Bing, Ask, Blekko, Wikipedia Search, and Yandex). He excluded known Jeopardy sites from the results, and didn’t test his own Wolfram Alpha because it is not designed for those kinds of queries.

What he found is that the search engines did fairly well, depending on how you measure success. Google did slightly better than the rest, but Bing and Ask were close behind. On average, Google got the correct answer somewhere on its first results page 69 percent of the time, versus 68 percent for Ask and 63 percent for Bing. Google got the right answer somewhere in the title or snippet of text of the very top result 66 percent of the time, versus 65 percent for Bing (and Ask dropped to 51 percent).

In comparison, most humans answer 60 percent of Jeopardy clues correctly, while the top player of all time, Ken Jennings, answered 79 percent correctly. So it is conceivable that a system could be created using regular search engines that could beat most humans. But Wolfram cautions:

Of course, the approach here isn’t really solving the complete Jeopardy problem: it’s only giving pages on which the answer should appear, not giving specific actual answers. One can try various simple strategies for going further. Like getting the answer from the title of the first hit—which with the top search engines actually does succeed about 20% of the time.

Answering Jeopardy clues correctly and consistently is a hard problem for computers to solve because of all the variations and nuances of human language. Yet “just using a plain old search engine gets surprisingly far,” concludes Wolfram.

Barry Diller realizes that his Ask search engine isn’t going to gain market share anytime soon, but search can still power growth for IAC if it just keeps up with the growth in the overall search market. IAC released third quarter earnings this morning. Total revenues were up 25 percent to $422 million. Operating income quadrupled to $36 million, and adjusted earnings per share came in at $0.32 versus Wall Street estimates of $0.27. (However, after stripping out the adjustments due to one-time sales of stock and other assets a year ago, net income was actually down 19 percent).

Search represented nearly half of revenues ($205 million). The search business grew 20 percent, goosed primarily by a 55 percent increase in active toolbars to 97 million. IAC’s toolbar business is its secret distribution weapon, but those searches tend to generate lower revenue per query than those on Ask.com, which itself is still growing and is now ranked as the sixth largest website in the U.S. LAst month, Ask CityGrid Media’s new local advertising network also contributed to overall search revenues.

But search wasn’t the fastest growing part of IAC’s business. Revenues for its media and other properties (such as CollegeHumor, The Daily Beast, Electus, Evite, and Vimeo) grew 44 percent to $63 million. Match’s revenues jumped 31 percent to $106 million, with paying subscribers up 30 percent to 1.8 million (“organic” growth in subscribers, though, was 16 percent). The rest of IAC’s revenues came from ServiceMagic, up 10 percent to $48 million.

In terms of operating profits, the two biggest contributors were Match ($38 million) and search ($29 million). The media businesses showed an operating loss of $4.6 million. The company also bought back $125 million worth of shares during the quarter. The stock is up about 5 percent this morning to $27.85 on the earnings news.

If you ask that question on Yahoo Answers, you’ll probably get something about how a husband’s Xbox was once used for Netflix, but now it doesn’t work. If you ask that question on WolframAlpha, you’ll get get some abstract data about money spent on various things each year. If you ask that question on Bing, you’ll get a bunch of old content. Both Ask.com and Google give you the answer if you dig into their links a bit — or if you click on their first links: to Quora.

What happens when you ask Quora that same question?

You get this: “about $600m”

That’s not just some random person citing some old numbers from an old story. That’s directly from Netflix CEO Reed Hastings.

The data point itself is somewhat interesting, but it has been out there before. What’s really interesting is that once again, Quora is proving to be a go-to place to get such information from fairly incredible sources. No one had yet answered the question someone asked about Netflix, so the CEO took it upon himself to answer it. Boom. Next question.

The fact that Quora is able to foster such an environment continues to fascinate me. If that can scale remains to be seen, but real people answering real questions about their companies is very, very useful. And if the platform can prove itself to be a good source of information relayed anonymously — such as here — it could really explode.

I wonder how long until one of the search engines tries to buy these guys? Or rather, I wonder how long until one of the search engines is successful in buying these guys — because I’m sure some have already tried. Again, this Netflix query brought up Quora as the first result in both Google and Ask. Google is all about saving time, so why not just buy Quora and bake this into results?

Of course, whether Quora, which was started by ex-Facebookers, would sell to Google or anyone is another story. But someone is going to offer this company a lot of money. And they’re going to need to.

Update: I decided to put my query on Quora: “Has one of the major search engines already attempted to buy Quora?“

Naturally, Co-founder Adam D’Angelo answered:

No. We’ve been pretty clear with everyone we’ve spoken to externally (including our investors) that our goal is to build Quora into an independent company over the long term and we are not considering acquisitions.

Good for them. Awesome product.

The search engine Ask is about to submit an iPhone app to Apple designed for asking questions on the go. About a year ago, Ask returned to its roots as a place to ask and answer questions. Roughly 35 to 40 percent of all search queries typed into Ask are in the form of a question, versus maybe 6 to 9 percent for other search engines, according to Ask president Doug Leeds. Ask’s main search page centers around questions, and last July it launched a social Q&A feature that is still in private beta.

When the iPhone app launches in the next couple of weeks, anyone who downloads it will automatically be enrolled in Ask’s social Q&A beta. The app lets you ask questions either by typing them in or speaking them. It translates the voice to text and then tries to offer up the most likely answer right away, based on its index of more than 500 million question and answer pairs from sources across the Web, including other Q&A sites, FAQ pages, and more. Right below that best guess is an option to “Ask the Community.” Your question will then be routed to people in the beta who self-selected as being knowledgeable in related categories. (Before the end of the year, you will be able to ask your friends on social networks like Facebook and Twitter as well).

The Ask iPhone app, however, will become more interesting with its second release, which will include a feature called “Nearby Answers.” Based on your location, it will show you a list of places nearby. You can then select one of the places and ask a question about it. For instance, you can ask whether a restaurant is any good or how long the line is at the local movie theater. Your question will be routed to people nearby or people who have answered questions about those places in the past. You will also be able to see questions asked nearby, and adjust the distance from less than 1 mile to 50 miles using a slider. If you are a local know-it-all, that is a great feature. Places data is coming from CityGrid Media, which, like Ask.com, is owned by IAC.

Geo-targeted questions and answers could potentially be very powerful. The social Q&A space is extremely crowded right now with everyone from Yahoo Answers to and Aardvark (now part of Google) to startups like Quora. Adding a geo layer on top of Q&A could be a way for Ask to distinguish itself. We’ll ask Barry Diller (CEO of Ask’s parent company, IAC) about this when he comes to speak at Disrupt next week.

All the fuss about Microsoft finally posing a credible challenge to Google with Bing, its new search engine, misses the real primary target of Microsoft’s search efforts: Yahoo. Microsoft knows it can’t unseat Google anytime soon, but it does have a fighting chance of taking down Yahoo to soften it for an acquisition or simply take over the No. 2 spot in search. Even that day is still a long ways away, with Yahoo commanding about twice as much search market share in the U.S. as Bing.

But data from Bing’s first full month after launch suggests that its initial share gains are coming out of Yahoo’s hide, not Google’s. ComScore released its June qSearch market share figures to Wall Street analysts last night (click on the table below from Barclays to enlarge), and they show Bing making a modest 0.4 percent gain in search query volume to 8.4 percent, compared to May, 2009. (Compete reported a 0.3 percent jump in search market share for Bing from May to June.)

Whlle that gain was modest, and less than what some of Bing’s own self-reported traffic numbers would have suggested, it pretty much all came out of Yahoo’s hide. According to comScore, Yahoo’s search market share declined 0.5 percent to 19.6 percent from May, 2009. Google’s market share stayed steady at 65 percent.

Shortly after Bing’s launch, comScore reported strong initial interest in the new search engine, but how much of that will translate into actual search market share remains to be seen. Bing’s 8.4 percent market share just brings Microsoft’s share slightly above where it was in March, and it is still below the 9.2 percent share it commanded a year ago. Bing still has along way to go. But as it gains traction, it isn’t Google that should be worried. It is Yahoo.

U.S. Core Search Share, June 2009 (Source: comScore qSearch)

One of the most active sub-genres of search right now in terms of startup and new product activity is question and answer sites. Some searches are subjective and best answered by another human being. The success of Yahoo Answers proved this and spurred a raft of competitors to try their own hand at making Q&A better. These include Answerbag, Wiki Answers, Mahalo Answers, Aardvark, and Hunch. Now Ask, arguably the original Q&A search engine (in that it encouraged searches to be asked as a question, not that the answers came from other humans), is waving its arms to remind people that you can ask questions and find answers there as well.

In fact, it is doing a little more than that. Today, it launched a Q&A tab on its site which taps into a new database of 300 million pairs of questions and answers, which it has crawled and indexed from around the Web. In other words, it is crawling the other Q&A sites to look for the best answers to a particular question. It is also applying some semantic and clustering filters to group similar questions together and to try to surface the most relevant results. It is more of a search engine for Q&A sites than a Q&A site itself. You can’t answer any of the questions, just search for what other people have answered on other sites.

At first glance, I find it a bit unsatisfying. I asked it, What is the best Q&A site? Yahoo Answers seemed to be the consensus, but no other choices even surfaced. I tried, What is the newest Q&A site? and it turns up only a single result from someone on Yahoo Answers asking how to go about creating a new Q&A site.

Does Ask even search Mahalo Answers? If it did, it would have found this question (“What other question and answer services have you used, tried or have found interesting besides Mahalo Answers?”) that includes a long list of more than 25 Q&A sites, many of which I had never even heard of (including Afraid To Ask, Ask An Owner, Blurtit, and Quenchmark).

It is not just that the answers on the handful of queries I tried weren’t so great, it is that taking a purely algorithmic approach to Q&A is the wrong answer. Obviously there are way too many Q&A sits out there and Ask is trying to find the best existing answers from everything that is out there across different Q&A sites. But offering Q&A search without letting people ask new questions or improve the results by offering their own answers kind of misses the whole point of Q&A. It is people helping out people to find the best answers to their questions. At least the Q&A startups are trying to move the ball forward by building a community and incentives around Q&A (Mahalo Answers), machine-learning and game-play (Hunch), or let you tap into your direct social circle for more trustworthy answers (Aardvark).

These sites get smarter the more people who use them and some of them offer personalized answers as well. The right answer to any question often depends on who is asking. Ask thinks there is one or two right answers for everyone.

(Image above courtesy PhotoXpress).

The company has always been secretive about their partners – they work with both advertising networks and advertisers/agencies and are able to compare traffic across those networks to increase data relevancy, and most of the time these partners don’t want others to know about their specific security precautions. But Technorati, LookSmart, Adbrite, Vivaki (Publicis Groupe) are all announced, and today Ask is also announcing that they’ve started working with Anchor Intelligence.

The product has also evolved since 2007. They aren’t just looking for fraud/no fraud on clicks any more. In addition to tracking fraud, the company is also generally scoring the overall attractiveness of a given click. Traffic with a very high likelihood of conversion can be sent to one type of advertising (CPA-type stuff), and lower quality stuff can be sent to ads that pay per click. The fraudsters can look at display ads all day.

The company says they are also working on a “self serve” product that publishers can use on their own to gauge the advertising quality of their traffic. I’m looking forward to using this for TechCrunch when it launches.

The company says that by next calendar quarter they’ll be scoring over a billion clicks a month, so there are clearly some other large partners working with the company that haven’t been announced. They’ve raised $6 million in venture capital to date.

Sendori developed the technology, dubbed PureLeads and patent-pending, to enable both search advertisers and domain owners to benefit from typed-in domain traffic based on the highest auction bids. With rates for PPC (Pay-per-click) dramatically dropping the past few months, Sendori was quickly becoming a nice alternative for domain name owners who traditionally looked no further than the usual suspects offering PPC advertising deals.

Seems like a good match with Ask Sponsored Listings, an Ask.com unit which focuses on keyword targeted advertising on a rather large (+100) network of sites including properties like Match.com, TicketMaster, Ask.com, Evite, CitySearch, CNet, etc. Ask.com has also actively been purchasing valuable domain names and websites like Dictionary.com, Reference.com and Thesaurus.com. Sendori’s client roster includes names like Netflix, GEICO, Hewlett-Packard and other familiar brands.

The terms of the acquisition were not disclosed, but we suspect Sendori got a good deal as its model was clearly working: the company said it was providing 130,000 advertisers 33 million page views per month from direct navigation traffic at the time of the acquisition. From the 11 employees (including contractors), 10 will make the move over to IAC.

Update: per commenter Satanish and BusinessWeek, we’ve learned that Ask.com paid a decent $25 million for Sendori.

The startup had raised only $800,000 in Series A financing, from Baseline Ventures, First Round Capital, Maples Investments and Felicis Ventures.

(Source: DNJournal)

The responses, in order of popularity:

Better results (45 percent).
Nothing would make me change (38 percent).
Faster response speeds to search (28 percent).
Less cluttered results (27 percent).
The ability to preview Web content (23 percent).
Less clutter on search sites (27 percent)
Results that include video, web, and music (12 percent)
Other (1 percent)

I’ve recreated the poll below so that you can take it. J.P. Morgan’s poll was only among 766 U.S. residents. We can probably do better than that.

What Would Make You Want To Try A New Search Engine?
( polls)

Here are some more results from J.P. Morgan’s poll, including the demographics of different search engine users. Google attracts both the wealthiest (70 percent of people with incomes greater than $100,000) and the youngest serachers (67 percent of people between 18 and 41).

Update: ComScore has notified me that it has discovered an error in its September figures and will be releasing amended numbers shortly. It appears that the impact of the change will be marginally positive for Google. We’ll update as soon as we get the recalculated figures.

Update 2: Corrected figures are below, as are the originally reported figures for comparison.  The mistake was in Ask’s numbers. Its market share declined half a percentage point to 4.3 percent instead of increasing to 5.4 percent.

Ahead of Thursday’s earnings announcement from Google, comScore just released its search market share figures for September. Google’s overall share of search queries in the U.S. dipped from 63% in August to 62.9% 62.2%. Yahoo and Ask (whose search is powered by Google) saw the biggest gains.

Corrected: U.S. Search Market Share (September, 2008)

Original: U.S. Search Market Share (September, 2008)

On the bright side for Google, both its annual and quarterly search query volume growth rates are accelerating. Year-over-year, Google’s query growth was 38.6 percent, up from around 33 percent each of the past three months. (On a quarter-over-quarter basis, the growth rate was 35 percent). Wall Street will likely focus on this acceleration as a slight positive for the stock.

Google as helped by overall search queries growing 26.9 24.7 percent across all search engines. Only Ask’s search volume grew faster, at 45.5 percent year-over-year. Ask’s growth slowed down to 8.2 percent, from 23.5 percent in August.  But that still helps Google, since Ask is a partner. AOL, another partner, saw 18.9 11.4 percent growth in search queries. Yahoo saw only 7.1 percent growth, and Microsoft saw a measly 3.0 1.3 percent growth (but at least its growth rate did not decline as it had each of the previous three months).

Corrected: U.S. Y/Y Search Query Growth Rates (September, 2008)

Original: U.S. Y/Y Search Query Growth Rates (September, 2008)

Here are the main search market share percentage numbers in the U.S. from comScore:

Company—–July search share—Change from June, 2008

Google:                    61.9                           +0.4

Yahoo:                     20.5                            -0.4

AOL:                          4.5                           +0.2

Microsoft:                  8.9                             -0.3

Ask:                           4.2                           +0.1

According to the latest search query and market share numbers from comScore for July, Google’s U.S. market share inched upward to 61.9 percent (from 61.5 percent in June).  While Google gained 0.4 percentage points in market share inJuly, yahoo lost the same amount.

And its search query volume in the U.S. held steady at a healthy 33.2 percent year-over-year rate (and accelreated slightly to 11.7 percent on a quarter-over-quarter basis).  That quarter-over-quarter rate is what caused investor concern earlier this year, when it troughed at -0.3 percent in February.  Since then it has re-accelerated every month to 4.4 percent in March, 6.1 percent in April, 9.9 percent in May, 10.0 percent in June, and now 11.7 percent in July.

Woldwide, Google’s search share declined from 67.9 percent in June to 64.1 percent in July. This drop was largely due changes in the way comScore measures search and Web traffic in China, Brazil, and Russia.  As a result, Baidu’s global search market share went from 7.7 percent in June to 12.9 percent in July (based almost entirely on its strength in China alone).  That makes Baidu the third largest search engine the world after Yahoo.

What the recalculation highlights is that Google faces more challenges to its dominance abroad than in the U.S.  Baidu’s global quarter-over-quarter search query volume grew 98 percent in July, compared to 3.2 percent global growth for Google.

Here are the global numbers with Baidu (China) and NHN (South Korea Russia):

Company—–July search share—–Change from June, 2008

Google:                    64.1                                  -3.8

Yahoo:                     14.6                                  -1.0

Baidu                       12.9                                  +5.2

Microsoft:                  3.6                                   -0.2

NHN                          2.1                                  -0.2

Ask:                           1.6                                  -0.1

AOL:                            1.0                                  -0.1

Here are some select scores, with the comparable 2007 numbers in parentheses:

American Customer Satisfaction Index Scores (2nd Quarter, 2008)

Google:                86 (78)
Yahoo:                 77 (79)
MSNBC.com:        76 (74)
ABCNews.com:     75 (74)
MSN:                   75 (75)
NYTimes.com:      75 (73)
Ask:                    74 (75
CNN.com:            73 (73)
USAToday.com:    73 (72)
AOL:                   69 (67)

The only surprise here is Google’s massive jump. Is it really doing that much of a better job than last year, or is it just that its halo effect keeps growing?

(Photo by Bing Ramos).

No word on what will happen to Ask’s Teoma search technology (the rumor was that Google would be replacing it, since it already handles Ask’s search advertising). Safka is obviously taking more of a marketing than a technology approach. But without improving actual search results (with technology), Ask is going to have a tough time maintaining its 4.5 percent market share. Ask’s search sites collectively brought in 41 million unique U.S. visitors in January, which was up from December and November, but still below October’s 44 million, according to comScore.

According to Silicon Alley Insider, Ask may abandon or selling its Teoma search engine in favor of using Google for its search results. Teoma has powered Ask since it was acquired in September 2001. The decision will result in “bad news for Ask Engineers.”

Paid Content puts the downsizing figure at 100 in April, although they note that the final decision on the switch to Google hasn’t been signed off on yet.

The decision to abandon Ask’s in-house search engine comes following a $100 million advertising campaign in 2007 that succeeded in growing Ask’s market share, but not to a significant level in the overall market. Google already provides Ask with its search ads through a recently renegotiated, five-year, $3.5 billion deal.

Silicon Alley Insider reports that Ask has just launched a news site called Big News in partnership with Digg that, from the looks of things, only partially incorporates social news functionality.

There were rumors just earlier this week that Digg was white labeling its technology for Ask. However, Big News is more akin to Google News or TechMeme than to Digg. The bulk of the news items collected and displayed from around the web are identified algorithmically, not socially.

Digg’s only clear influence on Big News shows up in the footer of the site, where you can view the current top five Diggs and five stories collected by Big News algorithmically that haven’t been Dugg yet. This real estate will help drive traffic to Digg and encourage the identification of interesting news stories. What does Ask get in return? That’s not altogether clear, although SAI hears that “Digg ratings factor into the site’s algorithm.”

In a comScore ranking of the top-10 global search engines as measured by number of searches during the month of December, 2007, Yahoo comes in at a distant No. 2 with only 13 percent of global share. (Although, in the U.S., Yahoo actually gained a half-point of share in December, whereas Google dipped 0.2 percent). The big surprise, though, is the strength of local search engines in countries that don’t use the Roman alphabet. No. 3 on the list is not Microsoft, but Chinese search engine Baidu (with 5 percent share, versus Microsoft’s 3 percent). No. 5 is Korea’s NHN Corporation, which operates the Naver portal and search engine. Creeping up on Ask’s No. 8 spot, is Russian search engine Yandex. And Alibaba (which may include Yahoo China) brings up the rear at No. 10.

Shouldn’t the best search technology win no matter what the language? These market share figures suggest that culture and marketing play a big role as well—unless, of course, you are Google.

Direct traffic on Ask.com grew from 29.8 million unique visitors in November 2006 to 46 million in November 2007 after dropping to 24.4 million in February for an impressive 54% growth rate.

Ask subsites saw some amazing growth rates, but mostly off very low bases. The Algorithm has a growing market in Europe, with Ask Spain experiencing 2062% growth rate, Ask Germany at 3006% and Ask France with 606%.

Some humble pie from me: back in May I slammed Ask.com for its advertising campaign suggesting that it was too clever by half; I haven’t changed my dislike of a campaign that suggested that “The Algorithm constantly finds Jesus” but the numbers don’t lie: it worked and worked well. Congrats and Christmas well wishes to the team at Ask; you’ve still got a long way to go to catch up to Google, Yahoo and Microsoft but at least you’re heading in the right direction, and competition is always a good thing.

Ask’s full numbers below:

The problem is most people don’t. A September Comscore report showed Ask was responsible for about 4.7% of all search traffic in July, which declined to 4.5% in August.

The move to privacy is simply not going to make a difference to their business. Google, Microsoft, and Yahoo have existing privacy plans in place since March, deleting personal information within at most 18 months (13 for Yahoo). Ask announced an 18 month policy in July. For years these companies succeeded with lackluster privacy promises.

The press loves to run stories about the hidden privacy concerns caused by data collected online, but consumers have taken an “out of sight out of mind” approach. DoubleClick has logged user data based on IPs and cookies for years, with only an obscure opt-out option that makes Beacon look pro-privacy (BTW, you can opt out here). It’s only going to be worse when Google’s search and analytics data is married with DoubleClick’s on site advertising information. Only when Facebook was upfront about what they were doing with user data, did people revolt. However, none of these invasions are affecting market share, nor have caused anyone I know to leave Google or Facebook.

We’re finding that people are willing to pay for the best free products, with their privacy.

The positive headline results did not flow through to the struggling 4th ranked search engine Ask.com, which saw a decline in revenues. The second quarter decline comes despite a $100 million Crispin, Porter + Bogusky advertising campaign that should be resulting in increased traffic and revenue to the site.

The exact amount of the decline was not disclosed.

According to the Wall Street Journal, Microsoft will officially announce Monday “new policies and technologies to protect the privacy of users of its Live Search services” and Yahoo will announce plans for “a policy to make all of a user’s search data anonymous within 13 months of receiving it.”

The same report goes on to detail plans by Microsoft and Ask to start an “industrywide initiative” to establish standard practices for retaining users’ search histories.

The 4 major search engines with major privacy initiatives in the space of a week and attempts to establish industry wide practices. Privacy would appear to be the new black. But why, and why now?

The Wall Street Journal correctly notes that in part, growing concerns among consumers and privacy groups is driving the move towards improved user privacy. It then goes on to cynically suggest that with Microsoft and Ask it may be a case of the search minnows trying to find a marketing edge over the much larger Google and Yahoo.

A stronger reason lies with Government pressure. In Europe, Google is currently being probed by an Independent EU panel that is investigating possible breaches of EU Privacy Laws. Although the probe is currently focusing on Google, it’s not an unreasonable assumption to make that it could easily be extended to other search companies. The FTC probe into Google’s acquisition of DoubleClick may also consider privacy issues relating to the acquisition along with anti-trust considerations.

No matter the reasons, the steps towards improving user privacy are welcomed. Expect to hear a whole lot more about privacy in the coming weeks and months.

Update: Microsoft has released details of its privacy changes here and its corporate initiative with Ask here.

(image credit: NewSchool)