ArticleSnatch Blog

You are browsing the archives of San Francisco.

After receiving feedback from the recent updates to Mobile Live Search, the team over at Microsoft has made even more updates. Here’s what to expect:

– Weather
– Traffic coverage for more cities through Live Maps
– Map a Contact
– Speech recognition via Bluetooth (available only on select devices)
– Collections - allows users to search community-generated content. See image below for an example of “dog parks in San Francisco.”

What do you think of these updates? Let us know in the comments.

Related Reading:
Price Check! Products Added to Mobile Live Search

I recently spoke on the multivariate testing panel at eMetrics in San Francisco. You would think that I dropped a hand grenade into the room when I opined that the Taguchi Method was a bad fit for landing page testing. This is a well understood fact to anyone with a solid understanding of basic statistics. Unfortunately this seems to leave out most landing page testers…

In the world of landing page testing there are two common mathematical approaches: A-B Split testing, and parametric Multivariate testing. A subset of Multivariate testing is known as “Design of Experiments” (DoE) and is also called “fractional factorial”. A common fractional factorial approach is called the “Taguchi Method”.

Some online marketers consider A-B Split testing to be kind of wimpy, and endow fractional factorial methods with an almost mythical quality.

I spend way too much of my time explaining to people that at least when applied to landing page optimization fractional factorial methods are a really bad idea. Despite this, the illusion persists that this kind of testing is somehow state-of-the-art, when in fact, nothing could be further from the truth.

For a lot more detail (30 pages worth), download the whitepaper - “The Truth About Taguchi”.

Testing is composed of two important activities:

- Deciding what to test and coming up with good ideas
- Finding the best solution among your tested alternatives

People claim to get really good results with fractional factorial multivariate testing, and they credit this to the method that they use to analyze the data.

In reality, the improved conversion rates are the results of the great ideas for new landing page elements that go into the test. If all of your alternative landing pages designs are better then the original, it does not really matter what method you use to confirm that. Fractional factorial approaches may actually miss the best version of the landing page in your test and often lead you to a sub-optimal answer.

There is a huge mismatch between the original environment in which fractional factorial testing was developed and how it is usually applied to landing page optimization. It was basically transplanted to online marketing because it is relatively easy for a non-mathematical audience to understand, and not because of its appropriateness or fitness for the task.

The principal drawbacks of fractional factorial methods are:

• Very small test sizes
  
• Restrictive & inflexible test designs
  
• Less accurate estimation of individual variable contributions
  
• Drawing the wrong conclusions
  
• Inability to consider context and variable interactions
  

If you plan on using parametric (i.e. “model building” )approaches for landing page testing you should always use full factorial data collection regardless of the subsequent analysis you plan to do. It greatly simplifies your test design, and produces better estimates of the main effects.

All parametric methods (including fractional factorial) are also outclassed by newer non-parametric testing methods such as the SiteTuners TuningEngine, which can be licensed to run your own tests in-house and have the following advantages:

• Very large test sizes (1,000-10,000 times larger with the same data rate)
  
• Much faster data collection (on the same data rate)
  
• More accurate results (consider variable interactions)
  
• Flexible test construction
  
• No knowledge of statistics required
  

Who won Survivor last night? Powerset doesn’t know.

Okay - maybe that’s an unfair question for Powerset since the aptly-named Parvati Shallow only won Survivor Micronesia last night. It might be easier to ask “Who is Parvati? The answer: in Sanskrit, “parvati” means “Mountain’s daughter” one of the names for Shiva’s wife, the Universal Mother.

So what is Powerset? That’s a much more intelligent question to ask.

Powerset is the much-hyped beta natural language search engine metaphor-challenged mainstream media call the Google Killer. That means you can type questions in a search box the way you normally ask them. (Think Ask Jeeves 1.5)

That doesn’t mean natural language search or Powerset can kill Google, or even commit assault and battery on Google.

Powerset launched with a smart concept: better search results than Wikipedia’s own search box. So the play is a “non-Google Custom Search Engine” for Wikipedia. Let’s see about what Powerset can can do..

“What is Powerset?” we asked. Powerset separates results by combining the primary keyword (Powerset) with related verbs and nouns from Web pages and Wikipedia. Here’s the answer we weren’t looking for from Powerset itself.

Factz from Wikipedia: we found the following about Powerset
Powerset opened : community and Powerlabs.

Results for Powerset opened community

Powerset (company) In a form of beta testing, Powerset opened an online community called Powerlabs on September 17, 2007.

Results for Powerset opened Powerlabs

Powerset (company) In a form of beta testing, Powerset opened an online community called Powerlabs on September 17, 2007.

Results for Powerset displayed advertise

Powerset (company) (Powerset is not currently selling or displaying any advertising.)

Results for Powerset searched language

Powerlabs Let’s roll, is a prerelease of Powerset’s general natural language search.

Wikipedia Articles: results 1 - 10 of 78

1. Powerset (company)
Powerset is a company based in San Francisco, California that is developing a natural language search engine for the Internet.
2. Power set
In mathematics, given a set S, the power set (or powerset) of S, written , P(S), or 2S, is the set of all subsets of S. In axiomatic set theory (as developed e.g. in the ZFC axioms), the existence of the power set of any set is postulated by the axiom of power set.

6. Powerlabs
Powerset Applications Currently, Powerset has released two applications: Powermouse and a beta of its search engine.

Only when you click on the expand snippet (powermouse) miniviewer button, do you find the answer you’re looking for (sort of):

Powerset (company)

Powerset is a company based in San Francisco, California that is developing a natural language search engine for the Internet.

The company’s stated desire is to build a search engine to compete with Google and Yahoo which would act on a user’s questions, as opposed to keywords. For instance, a user who wanted to find out which U.S. state has the highest income tax would enter “Which state has the highest income tax?” at Powerset, as opposed to “state income tax” at Google. The advantage to the user, aside from using questions similar to what one person would ask another verbally, is that a natural language search engine would, depending on its underlying programming, return a result that is more relevant to what the user seeks

Powerset is limited by the poorly-written Wikipedia entry. The comparison to Google doesn’t work. If someone’s searching for the state with the highest state income tax, they’ll type in “highest state income tax.” No one would type in “state income tax” in Google if they’d type “Which state has the highest income tax?” in Powerset.

At this stage in beta, Powerset can be considered a search engine survivor.

But unless the technology yields better results — and fast — people will think of Parvati and Powerset as “shallow” contestants.

I just got back from speaking on a multivariate testing panel at eMetrics in San Francisco.

A conversation I had with Marshall Sponder triggered this post. Marshall was bemoaning the fact that web analysts can’t even get “a seat at the table” (i.e. serious consideration) within many companies. To me this was a statement of the obvious. There are three main reasons that web analysts are not taken seriously.

Trying to do too many complicated & custom things

Data mining and analysis is pretty open-ended. A smart person can think up many ways to slice and dice the data. The very word “analyst” conjures up images of complexity. This is “rocket science” and no amount of simplified Key Performance Indicators (KPIs) on a customized dashboard for your boss will change that. If you dumb down the data too much then others may jump to the wrong conclusions. If you let them peek under the hood, then the complexity comes roaring back.

Looking in the rearview mirror

Analytics pores over data that was collected in the past. No matter how detailed or insightful it is, it can not necessarily be translated directly into actions because the conditions now may have changed significantly since the data was collected (e.g. traffic mix, seasonal factors, actions of competitors).

Not speaking the language of business

Analysts speak the language of numbers - but unfortunately not the right ones. Management only understands numbers in the context of finance. Does it make more money for us? Does it cut down costs?

So what can a web analyst do to get some respect?

1) Focus on mission-critical projects only - unless it can make a big impact on the business you should not waste your time on it. Intellectual curiosity and “what if” open-ended idea exploration should be limited to a small proportion of your time.

2) Get proactive - You should be working on actionable forward-looking activities such as landing page optimization which can actually lead to significant changes, and will not be filed away and never looked at.

3) Learn to speak in the language of finance - The only numbers that matter to your managers are the financial ones. Once you start presenting your projects and priorities in terms of their contribution to business performance you will get a lot more respect and attention.

You must reinvent yourself - instead of being viewed as a passive “analyst” you should become a business process architect who is relentlessly driven to improve the economics of your company.

If you do these things then you will be much more likely to get a seat at the table.

The Web Analytics Association has announced a new board of directors, and here they are:

• Vicky Brock, Highland Business Research, Inverness Highlands, Scotland
• June Dershewitz, Semphonic, San Francisco, CA
• Andrea Hadley, NetSetGo Marketing, Vancouver, BC Canada
• Avinash Kaushik, Author, Analytics Evangelist — Google, Mountain View, CA
• Alex Langshur, PublicInsite, Medford, MA

Neil Mason, Applied Insights, Oxford, England

• Lauria Paxia, I.C.C. S.r.l., Catania, Italy

Seth Romanow, Microsoft, Redmond, WA

• Marshall Sponder, Monster Worldwide, New York, NY
• Robbin Steif, LunaMetrics, LLC, Pittsburgh, PA
• Jim Sterne, Target Marketing, Santa Barbara, CA
• April Wilson, The Dallas Morning News, Dallas, TX

WAA Chairman Jim Sterne said, “The WAA is growing and maturing, and this new Board is a wonderful mix of talent and geography. The cumulative experience of these people is exceeded only by their cumulative IQ. We have a great team for tackling our objectives for the coming year.”

Related Reading:
Bryan Eisenberg Talks about Persuasion Architecture
Interview of Analytics Expert, Eric T. Peterson

Google CEO Eric Schmidt has named legendary Bubble 1.0 i-banker Frank Quattrone as an adviser to help reverse engineer the Yahoo-AOL and MSN- MySpace entanglements, according to the NYT blog DealBook .

For eagle-eyed readers of press releases, the news should come as no surprise. As Dealbook noted, Eric Schmidt gave a testimonial for the recently launched financial services venture last month:

“The launch of Qatalyst is an important development for the technology industry,” said Eric Schmidt, Chairman and CEO of Google. “Frank and his team bring unparalleled industry knowledge, a unique 25-year market perspective and candid, insightful judgment that CEOs greatly value on important strategic initiatives. I look forward to working with him again and am very enthusiastic about Qatalyst’s prospects for success.”

Quattrone has advised tech companies globally since 1981 while building technology banking departments for Morgan Stanley, Deutsche Bank and Credit Suisse. He was also the superstar banker who faced 18 months in jail but won his appeal in a U.S. court in Manhattan.

Qatalyst Group is a technology-focused merchant banking boutique based in San Francisco. Qatalyst Partners, its investment banking business, will provide high-end M&A and corporate finance advice to technology companies globally.

Qatalyst Capital Partners, its investing business, will make selective principal
investments, typically alongside leading venture capital and private equity firms.

“Video search is a false messiah,” claimed Steven Chao at this week’s Video Search Summit in San Francisco. Video search is not about words, optimization or getting into the traditional web search engines. Instead, he says we must look to video consumers and more collaborative approaches.

It’s telling that Chao, who’s all about our cultural zeitgeist, is currently running an online portal for How-To videos. Chao is best-known as the originator of America’s Most Wanted and Cops, and later served as USA Network’s president.

According to Chao, we’re just at the beginning of video consumption on the web. Today, average Americans are still attached to their televisions and he believes it’s only a matter of time until Americans shift and become “internet viewing zealots.” Improving findability is critical for this transformation.

Click to read the rest of this post…

Google launched a new branded product today: the Google I/O Web Forward developer conference, a two-day, event for developers to share knowledge about Google’s developer products in specific and Web application development in general. For marketers, I/O means insertion order; for developers, input/output.

I/O (input/output) means communication between an information processing system (i.e. computer), and the world – possibly a human, or more likely in Google’s scenario: an information processing system (computer).

Sounds like Search Engine WarGames to us.

Google I/O will be held May 28–29 at the Moscone Center in San Francisco. Five tracks will run simultaneously with several related to local mobile search and social search.

Social Search: The web is an inherently social place. A new wave of APIs and frameworks, including OpenSocial, are opening up interesting ways to interact with people through applications.

Mobile: Android and Mobile Gears offer the opportunity to build applications on the mobile Web. Search is moving to mobile devices and opening up applications that span multiple platforms.

Maps & Geo: Local search is not only going mobile, it’s going geo. These sessions focus on geo applications, including KML (Google Earth apps)and the Google Maps API.

There will also be tracks for APIs (Google AdWords) & Google Web Toolkit.

Tickets are $400 for developers (or any interested people, we presume), with discounts available for students.

Google Rock Star devcelopers: attendees will\\ learn first-hand from Google engineers like Mark Lucovsky, Guido van Rossum, David Glazer, Alex Martelli, Steve Souders, Dion Almaer, Jeff Dean, Chris DiBona, and Josh Bloch.

Want a snapshot of the day’s search marketing news? Here we’ve collected today’s top news stories posted to the Search Engine Watch Blog, along with search-related headlines from around the Web:

From the SEW Blog:

• Google Defends Security Policy
  Douglas Merrill, Google’s VP Engineering, published an impassioned defense of Google’s security policy.
• Cable Firms Collaborate to Compete with Google
  Executives from the six major cable companies have been gathering at monthly meetings in an attempt to offer national advertisers innovative methods of reaching their customers.
• BoomTown Calls Out TechCrunch: No Google-Digg Buyout
  It seems that Digg’s canoodling with bankers Allen & Co. is more along the lines of delegating the task of handling incoming interest rather than putting a “For Sale” sign up at its San Francisco offices.
• The New Multitaskers: Kids Split Attention Between TV, Internet
  A new study on social networking by Grunwald Associates reveals that kids are no longer glued to the television. Now, while the tube is on, many kids are also splitting their attention between the TV and the Internet.
• Introduction to Search Engine Marketing at SES New York 2008
  It’s hard to provide “a clear and concise overview of the key concepts involved in Search Engine Marketing” when half of them have changed significantly since last year.
• Panama or AdCenter? Microsoft Has Already Decided
  If its proposed acquisition of Yahoo goes through, don’t expect Microsoft to rush to integrate Yahoo’s technology into its platform.
• CAPTCHA Systems are Under Attack
  Should spammers ever succeed in breaking through CAPTCHA, a lot of social media web sites and blogs are going to have a big, big problem. Now, Virtual Blight reports that this is on the verge of happening.
• Best Web Site Traffic: PPC vs. SEO (Search Engine-Optimized) Sites
  How does your PPC traffic compare to natural search traffic and direct navigation? A study by Engine Ready finds that that type-in traffic is the highest converting and “most valuable” traffic.
• SEW Experts: Every Picture Tells a Story: Non-text Contextual Ads
  Like contextual text ads, image ads must distract the user and sell a product quickly.
• Jason Calacanis TechCrunched in Techmeme: Deathmatch at SES New York
  Duncan Riley of TechCrunch fame bodyslammed Jason after his Calacanis.com post on how to run a startup.

Click to read the rest of this post…

Last week’s TechCrunch article about Microsoft and Google in a bidding war for Digg is the stuff that soap operas are made of. Or tabloids. But Kara Swisher at Boomtown has the real scoop, albeit a less sexy one.

It seems that Digg’s canoodling with bankers Allen & Co. is more along the lines of delegating the task of handling incoming interest rather than putting a “For Sale” sign up at its San Francisco offices.

And if an offer does come in from Microsoft or Google, don’t expect to see the purported $200 million. Swisher says that a $60 to 80 million price tag is a more realistic offer, according to the acquiring types she spoke to.

Even Digg CEO Jay Alderson broke Digg’s “we don’t comment about things like this” policy and blogged that the rumors were out of control, denying the very existence of a Microsoft/Google bidding war. But TechCrunch stands by its original sources.

And why shouldn’t they? As Kevin Heisler opined last week, a Digg acquisition would make strategic sense for Google. And with or without the Yahoo deal, Microsoft has last year’s ad deal with the social search engine as an incentive to raise their auction paddle in the air.

In the meantime, like a policeman trying to keep traffic moving around a fatal accident, everyone but TechCrunch seems to be saying, “There’s nothing to see here. Keep moving along.”