Demand Media and the Decline of the Content Farm
December 19, 2013
An article in Variety, “Epic Fail: the Rise and Fall of Demand Media” can be read as a cautionary tale for the search engine optimization and content marketing crowd. Just a few short years ago, the newly-public, five-year-old company out of Santa Monica saw its market capitalization top $2 billion. Now, however, writer Andrew Wallenstein contrasts that success with the company’s status today. Demand Media is now worth about a quarter of its peak value, and was forced into several rounds of layoffs this year.
The company is prudently turning its focus to its successful side—domain registrations—and Wallenstein predicts the media side will soon be sold off or taken private. What went wrong? Well, we’ve always maintained that a building a business, nay an industry, around gaming Google‘s search engine was just asking for trouble.
The article recalls:
“Early on, Demand used [its domain-name registration service’s] 1 million generic domain names (such as ‘3dblurayplayers.com’) to serve up relevant ads to people searching for specific topics. These ‘domain parking’ pages were immensely profitable, generating north of $100,000 per day, according to a former Demand exec who requested anonymity. ‘That’s $35 million-$40 million per year without doing any work,’ the exec said.
But the tactic was fundamentally a bait-and-switch. Users landed on the pages expecting to find information on a subject and instead found an ad. To try to drive up traffic, Demand shifted its strategy, populating the sites with thematically related content….
Demand then continued to build out the content-farm strategy, treating the domain-name registration business as largely separate from the content-production arm. Paying contributors comparatively little — usually less than $20 for a single article or video — it built up a stockpile of content against which it sold targeted advertising.”
This tactic (responsible for much of the useless content populating the Web) exploded and was adopted by copy-cats. Since then, however, both users and Google’s algorithm have wised up, making content farms much less profitable. On the bright side, this means less incentive to fill websites with baloney. Hooray!
Cynthia Murrell, December 19, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
More Trouble for Barnes and Noble
December 19, 2013
Oh, dear. Have desperate times led to desperate measures for Barnes & Noble? A brief post at theneeds informs us, “SEC Probing Barnes & Noble Accounting.” Writer Susan Lulgjuraj reports:
“According to the [AP] report, the SEC told the book company it was looking into two issues: the company’s restatement of earnings announced in July and an allegation from an employee who said that B&N improperly allocated some information-technology expenses between its Nook and retail segments. B&N said in a July filing that it had restated results from the fiscal years ending in April 2011 and 2012 due to material errors in the financial reports for those periods, according to the AP report. B&N said it was notified on Oct. 16 that it would be investigated.”
In an effort to adapt to the changing publishing landscape, the company began reporting its Nook revenue separately from the rest of its sales in late 2012. Naturally, news of the investigation was reflected in the company’s share prices, which dropped almost 12 percent right after the announcement. Will Barnes & Noble recover, or is this confirmation that the primarily brick-and-mortar bookseller is on the way out?
Cynthia Murrell, December 19, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
Whither the Bing Thing in 2014
December 18, 2013
I found the data in the “2013 Bing Infographic” surprising. I continue to think of Bing as a search and retrieval system. I don’t use the system directly. I prefer to run queries on metasearch systems that use Bing as one source of content. The reason for my indirect access is that I don’t want distractions, social media content, and videos. In case you, gentle reader, have forgotten, I prefer to read. I read more rapidly than I can watch a video unfold in real time. I understand that some people find videos just the best possible way to locate information. I don’t.
The infographic has a number of data points. Let’s look at three in the context of locating a white paper, information about a person of interest, and a fact.
First, Bing reports that if people looking at a Bing home page each month were to hold hands, the length of that “chain” would be the circumference of the earth. Got it. What’s that go to do with precision, recall, and access to information? Nothing. Okay. That’s fact one.
Next, Bing has more video. That is super. I don’t want video. Period. Well, Bing had twice as much video search in 2013 than in 2012. Got it. I don’t care.
And Bing is the search engine for Facebook (really?), Yahoo (ah, that’s the problem with Yahoo search), and the Kindle Fire (I don’t use a Kindle Fire).
What does the infographic reveal about search at Microsoft?
- Search is not the point of Bing. I thought Powerset and Fast Search were going to improve Bing search? Guess not.
- Why is it getting * more * difficult to locate information instead of easier? Maybe the vastness of the Web and economic pressures are forcing Microsoft to shift from search to some other type of service? That’s okay, just knock off the use of the word search.
- How do professionals at Microsoft locate information? I don’t have any hard data, but I think that Google (an outfit doing a rather questionable job in search) may be good enough. That is indeed chilling to think that Microsoft professionals trust Google to point them to hard to find Microsoft research papers and obscure FAQs about Microsoft products.
Bing had a shot and spent some money shooting blanks in my view. So for 2014 I don’t expect much improvement. I hope libraries in my area have enough money to remain open and provide access to commercial online information resources. The free Web stuff does not strike me as getting better. Oh, if you want video and social media, you may be in business.
How often do I run a query on a Windows 8.1 laptop and want Web hits and not a list of documents on my local hard drive matching my key word query? Never. There you go.
Stephen E Arnold, December 18, 2013
Box Fills Hole for Mac Users
December 18, 2013
A number of good enterprise solutions are on the market. An organization usually decides between them based on their individual needs. For some, usage of iOS and Mac OS platforms means that SharePoint is not a viable option. CITE World expands on this idea in their article, “How Box is replacing SharePoint and custom software at Scripps Networks.”
The article explains:
“Now, people throughout the company ‘use it for exchanging everything,’ he said. ‘I use it the way I used to use SharePoint.’ Box ended up solving an additional problem for Scripps: Filling a hole that SharePoint left for Mac users. SharePoint isn’t very Mac friendly, Hurst said, which became an issue for the 35 percent of Scripps workers who are on Macs.”
Box is touted as user-friendly and enterprise ready. For those reasons, it may just give SharePoint a run for its money. Add to this the fact that Mac OS and iOS users are on the rise and Box has something that SharePoint can’t offer – agility between operating systems. Stephen E. Arnold of ArnoldIT covers the latest in SharePoint and enterprise search. He has recently and often said that SharePoint may be in trouble, and this type of stiff competition adds to that argument.
Emily Rae Aldridge, December 18, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
BayesDB for Easier Analysis
December 18, 2013
Interesting—it seems the venerated Thomas Bayes is now with us in database land. BayesDB is being developed, in conjunction with an analysis method called CrossCat, by a team of folks from MIT‘s Probabilistic Computing Project and the Shafto Lab at the University of Louisville.
The project’s page explains:
“BayesDB, a Bayesian database table, lets users query the probable implications of their data as easily as a SQL database lets them query the data itself. Using the built-in Bayesian Query Language (BQL), users with no statistics training can solve basic data science problems, such as detecting predictive relationships between variables, inferring missing values, simulating probable observations, and identifying statistically similar database entries.
BayesDB is suitable for analyzing complex, heterogeneous data tables with up to tens of thousands of rows and hundreds of variables. No preprocessing or parameter adjustment is required, though experts can override BayesDB’s default assumptions when appropriate.
BayesDB’s inferences are based in part on CrossCat, a new, nonparametric Bayesian machine learning method, that automatically estimates the full joint distribution behind arbitrary data tables.”
The database is designed for two types of folks: those with no statistics chops who nonetheless have tabular data to analyze, and those proficient with statistics who have a non-standard problem or who have no time or patience for custom modeling. The team credits CrossCat in part with making BayesDB possible, but also say the BQL language was key to its development.
The description includes examples, a discussion of which types of data and problems the database addresses best, reasons to trust the results, why they named it BayesDB, and more. Check out the page for all the details.
Cynthia Murrell, December 18, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
HP Leaps into E-Commerce Software Sales with HP Pronq
December 18, 2013
The article titled HP Launches Portal to Sell Its Software Online on eweek introduces the HP Pronq effort. No, that’s not a typo, but a new business that will sell HP software. Still confused?
The article states:
“The Pronq portal currently offers HP’s Fortify on Demand security service, Agile Manager, Vertica, Performance Anywhere and Service Anywhere solutions. “Pronk” is an actual word that is defined as jumping up into the air or moving forward by leaps and bounds, Caroline Tsay, vice president of Web and eCommerce at HP’s Software division, told eWEEK. “It’s a metaphor for what we’re trying to do with Pronq, with the attributes of agility and ease of use,” Tsay said.”
The launching of this new business is meant to draw more attention to HP as a software provider. Often HP gains more attention as a PC and printer brand, and with Pronq the company is working to put their software in the limelight. Pronq is using Drupal, the content managing system, and a private cloud different from the HP public cloud OpenStack-based infrastructure. Careful to avoid the insinuation that by selling software directly through Pronq HP is trying to outmaneuver or work around their partners, Tsay claims the e-commerce business will actually “complement” the sales teams and “support” HP’s channel partners.
Chelsea Kerwin, December 18, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
The Future of Semantic Search
December 18, 2013
The article titled The Stealthy Rise of Semantic Search on Search Engine Journal relates the outmoding of SEOs with natural language search. The article explains that Boolean expressions leaped search forward with the abilities of query modifiers. Many people use semantic search without realizing it, but it works nonetheless to determine intent instead of just matching search terms.
The article explains:
“Vertical search engines such as Hakia, Lexxe, and VSW were at the forefront of semantic technology long before Bing, Facebook, and Google. These engine providers built their business around that potential and are using it to create new distribution and business models to deliver options for content makers well beyond keyword search and SEO.
Advertisers will reap huge benefits from semantic search because it increases the relevance of all forms of advertising.”
Semantic search will also decrease the likelihood of ads being paired with unsuitable content. Soon, users will be presented with information they “didn’t even know they wanted.” This may even lead to interconnected devices within the home. The article offers the possibility of the refrigerator notifying the homeowner that they are running low on milk. (Smart House, anyone?) Whether scary or exciting, the article would have us believe that these developments are inevitable.
Chelsea Kerwin, December 18, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
Computer Scientists Announce Possibility of “Smart Drugs” Through Chemical Computational Devices
December 17, 2013
The article titled Programming Smart Molecules: Machine-Learning Algorithms Could Make Chemical Reactions Intelligent on Science Daily reports on the work of computer scientists from Harvard and the Wyss Institute. The possibilities are exciting, “smart drugs” that can adjust to an individuals chemistry? Yes, please. The article explains the development of chemical computational devices aimed at certain algorithms, which have already made robots capable of more complex decision-making.
The article states:
“This insight opens up interesting new questions for computer scientists working on statistical machine learning, such as how to develop novel algorithms and models that are specifically tailored to tackling the uncertainty molecular engineers typically face. In addition to the long-term possibilities for smart therapeutics, it could also open the door for analyzing natural biological reaction pathways and regulatory networks as mechanisms that are performing statistical inference.”
The article goes on to compare bioengineers facing incomplete molecular pathways with the movement into big data attempting to make sense of huge amounts of data. The ability to leap these obstacles with modeling between random variables is in the near future, and already found to be an efficient method of gathering data through random events. Will Google’s synthetic biology group adopt these methods? It seems likely.
Chelsea Kerwin, December 17, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
Big Data Experimentation: a Business Model for the 21st Century
December 17, 2013
The article Big Data Demands Nonstop Experimentation on Infoworld targets old school executives living in what the article sums up as a conservative dream world where experimenting with Big data is a radical time waster. The article suggests a shift in the fundamental approach to Big data from the “causal density” model to a continual and ideally randomized experimental model.
The article explains, quoting blogger Michael Walker:
“Under this approach, business model fine-tuning becomes a never-ending series of practical experiments. Data scientists evolve into an operational function, running their experiments 24-7 with the full support and encouragement of senior business executives… Any shift toward real-world experimentation requires the active support of the senior stakeholders — such as the chief marketing officer — whose business operations will be impacted. As Walker states: ”…Business and public policy leaders need to support and adequately fund experimentation by the data science and business analytics teams.”
This may just be the information cash-strapped, revenue-hungry firms need. No one said Big Data was easy or quick, but what is clear is that a commitment to experimentation is needed, lots of experimentation. The article cites Google and Facebook as 21st Century success stories already embracing this model, in fact founded on the idea of experimentation.
Chelsea Kerwin, December 17, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
Languages Supported by Google Translate Increase
December 17, 2013
The article on eweek titled Google Translate Adds Support for More World Languages announces Google’s addition of nine languages to its service, making the total number 80 languages. These included several African languages spoken in Nigeria, Somalia and South Africa. There are motions in progress to add Mongolian, Nepali, Punjabi and Maori. The last was only made possible by New Zealanders, as the article explains:
“The Translate effort for Maori was made possible due to the “volunteer effort of passionate native speakers in New Zealand,” he [Arne Mauser, Google Translate software engineer] wrote. Users who want Google Translate to add other languages to the service can participate by “volunteering to help us gather and translate texts in your language,” he wrote. “We’re also constantly fine-tuning our translations. You can help with these efforts by clicking the translated text and editing it to be correct.”
The app itself is simple enough to use. You can now speak into an Android device and receive a translation or use the handwriting feature. You can also take a picture of text in another language and highlight the words you want to learn. The app is appealing to travelers especially now that it is possible to use language-translation on their phones even without an internet connection, by downloading offline language apps.
Chelsea Kerwin, December 17, 2013
Sponsored by ArnoldIT.com, developer of Augmentext