Margaret Stewart, YouTube's head of user experience, talks about how the ubiquitous video site works with copyright holders and creators to foster (at the best of times) a creative ecosystem where everybody wins.

Earlier this year, we touched upon the MASS MINGLING trend in our ‘10 trends for 2010’ briefing, but this phenomenon now warrants its own, full briefing:
As predicted by digital gurus more than a decade ago, hundreds of millions of people are now living large parts of their lives online (and lovin’ it!). However, this has not turned entire generations into homebound, anti-social zombies (another popular forecast). Au contraire: social media and mobile communications are fueling a MASS MINGLING that defies every cliché about diminished human interaction in our ‘online era’.
So (for now), forget a future in which the majority of consumers lose themselves in virtual worlds, with cities dying and kids never seeing the light of day; expect people to mingle and meet up like there’s no tomorrow. A definition:
MASS MINGLING | Thanks to the online revolution, hundreds of millions are now actively searching for, finding, connecting/signaling, and staying in touch with likeminded souls in the virtual world. Constant updates, GPS and mobile online access is now bringing this explosion of dating, networking, socializing and mingling to the real world domain.
Here’s what driving this trend, in more detail:
MASS MINGLING follows the pattern of any consumer trend, whereby an existing human need is unlocked in a new way. In this case, interacting with other people - a fundamental need (which goes beyond simply enjoying one another's company, or being emotionally dependent on other people) - has become easier thanks to new technology.
So, no surprise here that hundreds of millions of people are now adding and tending to personal profiles (listing likes and dislikes, interests, preferences, physical assets, and opinions), making it easier than ever before to ‘discover', or stay in touch with, likeminded others*.
Think friends and family, colleagues, romantic interests, and those sharing similar hobbies, interests, political preferences, grievances or causes. And all this 'befriending' takes place in unprecedented quantities: never before were people able to build and maintain such extensive and relevant personal networks.
Some numbers:
* Hundreds of millions of personal pages, feeds, status updates, tweets, profiles, blogs—courtesy of the Facebooks, the Twitters, the LinkedIns—are building an eternally up-to-date encyclopedia of individuals. Some thoughts on how this will lead to 'forever connected' amongst younger generations, from our fave media guru Jeff Jarvis:
“Thanks to our connection machine, they [young people] will stay linked, likely for the rest of their lives. With their blogs, MySpace pages, Flickr photos, YouTube videos, Seesmic conversations, Twitter feeds, and all the means for sharing their lives yet to be invented, they will leave lifelong Google tracks that will make it easier to find them."
Image courtesy of flickr/rodrigofavera
While the rise of the online world remains one of the premier (consumer) stories of our time, URBANY and the EXPERIENCE ECONOMY /LIVING THE LIVE, with their decidedly 'offline' overtones, are enormously impactful too.
This incredibly powerful tandem of mass urbanization and experiences has resulted in an orgy of real world activities and happenings that are all about mingling; from countless cultural and not so cultural events, concerts, festivals, and seminars, to a burgeoning and truly global bar/dining/party scene, to a Warholian retail renaissance, to tourism & travel now being one of the world's largest industries, employing approximately 220 million people and generating over 9.4 percent of world GDP. In short, people have always, and will for a long time continue to enjoy interacting with other warm bodies.
The mobile web has eradicated any wired person's dilemma whether to be offline in the real world, or being online but stuck in one location (in a room, or worse, in a basement). Offline is now online, and online is offline. More on that below ('An info-layer on top of daily life').

A quick side-step: Traveling (and thus meeting up) has of course become easier, cheaper and yet more sophisticated on a global scale. Forget recessions, strikes and Icelandic volcanoes: the NO-FRILLS CHIC travel-eco-system now includes low cost airlines from the Middle East to Asia, and funky Yotel/CitizenM style hotels from Taipei to Toronto. Making it possible to go anywhere and meet anyone, at low costs, without having to sacrifice too much comfort.
Photo courtesy of Mistdog on Flickr
Back to the beginning of this Trend Briefing: MASS MINGLING is happening because people can. There's now an all-encompassing information layer* on top of real-world daily life, that (especially when mobile and location-based), turns 'connecting' into 24/7 and 'on the go', further blurring the boundaries between online and offline.
This layer has created a space in which following, finding, tracking, connecting to, and ultimately (spontaneously) meeting up in the real world with interesting known and unknown people will be easy, automatic, instinctive, convenient, and even natural. And thus, for many, connecting to 'strangers' is rapidly becoming second nature.
For a glimpse of things to come, dive into location-based 'meet-up' services like Foursquare (now doing 700,000 check-ins per day), Gowalla, Google Latitude and Loopt.
By the way, the first official Foursquare Day took place this April. News of the event spread by word of mouth, leading many users to organize parties and gatherings celebrating the social network.* Let's not forget that this emerging layer also taps into a vast reservoir of local knowledge and content that has been growing online for years. This incredible info-infrastructure further helps (if not encourages) MASS-MINGLING prone consumers to select venues and activities (based on profiles, preferences, reviews etc.) before venturing out.
Via | Trendwatching
Mini lanzó una campaña en Facebook a través de la que desafía a Porsche, con autos conocidos como rápidos, a una carrera que parece no ser justa (porque el Mini tiene 172 hp contra los 385 hp 911) pero será una muy buena publicidad:


Pero, este foco sobre las cuestiones de modelización hacen que el miner, a veces, se “olvide” de un preliminar fundamental: la preparación de los datos que serán el insumo principal de las tareas de modelización. No solo se olvida del aforismo de que “entra basura, sale basura”, sino que también tiende a subestimar la complejidad de la tarea (complicada y bastante específica) de llegar a una tabla de análisis apropiada. Lamentablemente, esto atenta peligrosamente contra el feliz desarrollo de un proyecto de data mining.
ETL es un actividad de IT que apunta a generar eficiente y eficazmente tablas de bases de datos a partir de diversas fuentes (muchas veces otras tablas). Las preocupaciones pasan principalmente por la estructura y tamaño de las tablas fuente y destino, las tareas específicas de extracción, transformación (incluyendo joins) y carga, la generación de jobs automatizables, la calidad de los datos, etc.
Si bien cualquier proceso de generación de tablas puede considerarse de incumbencia genérica del ETL, las características específicas del mismo para tareas de data mining hacen que la persona que lo realice deba utilizar una estrategia vinculada a conceptos y técnicas de data mining. Pero, cuando esta persona proviene del área de IT muy probablemente no tiene los conocimientos necesarios. Por otro lado, el miner no posee, por lo general, las destrezas necesarias (programación, conocimiento de las fuentes de datos subyacentes, etc.) para realizar por sí mismo la tarea. De esta manera, el proceso (fundamental) de preparación de datos cae en un “agujero negro” del proyecto, con las consecuencias obvias.
¿Qué hacer entonces? La respuesta no es demasiado complicada: capacitar al miner o al personal de IT (¿y por qué no a los dos?) en las necesidades peculiares de la preparación de datos para data mining. En esta capacitación, las personas involucradas deben aprender una estrategia y las técnicas apropiadas para realizar eficazmente esta actividad.
Parte de esta estrategia consiste en implementar un conjunto de etapas que permitirían pasar de tablas inespecíficas a la tabla de análisis que el miner registraría en el SAS Enterprise Miner para comenzar su tarea de exploración o modelización. No conviene generar tablas faraónicas que luego es bastante probable que resulten inútiles. Más bien, conviene pasar por algunas fases de refinamiento progresivo, comenzando por una tabla muy simple que permita empezar a trabajar lo antes posible en las tareas de modelización. Es decir, es más valioso una tabla mala pero que se puede usar mañana mismo para modelizar, que sentarse a esperar seis meses para obtener una tabla enorme que muy probablemente haya que achicar considerablemente para poder generar un modelo.
Para producir una tabla con características apropiadas hay que preocuparse por criterios netamente estadísticos: no generar muestras heterogéneas (poblaciones mezcladas), contaminadas temporalmente, sin control de la debida estacionareidad o estacionalidad de los fenómenos. Hay que preocuparse por las “patologías” de los datos, es decir, los problemas (fundamentalmente estadísticos) que los datos podrían contener. Hay que evaluar si es necesario enriquecer los datos y de qué manera.
Como en muchas otras áreas, SAS posee su propia metodología para la tarea de preparar los datos de data mining. Se parte de una focalización en un conjunto de casos representativos de una población de referencia o interés. Solo después se recolectan variables predictivas vinculadas a los casos seleccionados. Por lo general, las variables más interesantes son variables que sumarizan la interacción de nuestros casos (transacciones). Esto se hace mediante un proceso relativamente complejo que exige algunas destrezas de programación. Luego de esta etapa se puede pensar en corregir patologías, construir nuevas variables, hacer enriquecimiento de tablas, etc. Algunas de estas últimas tareas pueden realizarse mediante procesos de ETL o dentro de un flujo de nodos en SAS Enterprise Miner. Cuándo es conveniente una u otra alternativa depende de consideraciones específicas a las diferentes tareas.
Recuerde que lo peor que le puede pasar es tener que dedicar un 90% o más del tiempo de un proyecto a las tareas de preparación de datos y que, por lo tanto, no tenga suficiente tiempo para la tarea medular de modelización.Vía | SAS EducaciónImage via Wikipedia
Via | NinjaMarketing
Image via Wikipedia
Forget Foursquare or Gowalla: Soon every website and service will be able to tell where you are, opening up the floodgates for location-based marketing and blurring the budget lines for advertisers.
"What used to be called point-of-purchase is now called mobile advertising," said Kip Cassino, VP-research at Borrell Associates. "Mobile can be an extension of a retailer's storefront."
The potential of knowing when and where a consumer is -- within privacy constraints, whenever those get hammered out -- means thinking outside the interactive-advertising budget and dipping into other marketing disciplines' coffers. "We're talking about the buckets that were so nicely separated between advertising and promotions starting to fade," Mr. Cassino said.
Some of the money will be sourced from localized print media budgets (such as the Yellow Pages), newspaper classifieds and direct mail, as location helps people find businesses and coupons; some of it will come from interactive budgets as location will tie into the ways we access the internet in the future.
New formats coming
Paul Feng, Google's mobile-ads group product manager, said about one-third of mobile searches have local intent. To re-imagine search for phones, Google is evolving ad formats from just text and display; it recently let advertisers add phone numbers to mobile ads so consumers can click to call. Mr. Feng said to expect more new ad formats in the coming months that could potentially incorporate new types of interaction, such as navigation.
"We think of location as a hugely important signal," said Mr. Feng. "Foursquare is really innovating for small businesses and that's fantastic. But ultimately that's just one piece of the opportunity." (Microsoft's Bing is also building out its mapping tools and expanding its local listing databases.)
Borrell, in fact, projects in its upcoming mobile report that location-based mobile spending will hit $4 billion in 2015, up from $34 million in 2009. Including ad spending with promotions, events and research, Borrell estimates mobile as a whole will dominate U.S. interactive marketing spending as soon as 2014 with 70% share, or $56 billion. While it's difficult to imagine mobile advertising beating existing commitments to web development, banners and search, those commitments won't be dedicated to PC browsers but increasingly to smartphones, Kindles, iPads, portable gaming devices and even cars.
"The demand for stationary computers is going to dry up on a consumer level," Mr. Cassino said.
Outside of apps, there's the mobile web, e-mail, text messaging and Bluetooth. Companies such as Placecast "geo-fence" retail locations and, once you opt in, blast FYIs or offers texts when you're within a determined radius, blurring the lines between signage, couponing and mobile. In a four-month study with the American Eagle retail and Sonic restaurant chains, Placecast reported 79% of participating consumers said geo-fencing programs increased their likelihood to visit stores, and 65% made purchases during the program.
Right now, of course, given the hype machine at South by Southwest, the buzz is all about Foursquare and similar services, such as Gowalla. Its social appeal has not yet been emulated, though social giants such as Facebook and Twitter represent a threat. Today, one-quarter of Facebook's 400 million users access the site through mobile devices; this set is twice as active than non-mobile users. Facebook is also expected to reveal a location-based feature in April, and Twitter released geo-tagging to developers late last year, which ties location to tweets. Facebook is an especially big contender in the mobile-location rush, with its existing scale and user base.
"From a consumer perspective, the decision is going to be made on where their friends are," said Jeremy Lockhorn, director-emerging media, Razorfish. "The fun of being on Foursquare or Gowalla is that our friends are there. If that's Facebook for other consumers, that's where they're going to go. You're going to need some level of scale to attract dollars."
Via | Adage.com