Understanding Google Search Engine

The search results come from Indexed pages [i] and Knowledge graph pages[ii].

Then, Google analyses the words typed into the search bar, both  literally and semanticly.

The literal search consists in the engine looking for the words or phrases exactly as they are entered.

The semantic search system considers context of search, location, intent, variation of words, synonyms, generalized and specialized queries, concept matching and natural language queries to provide relevant search results[iii]. In order to understand what your search means, the term is also broken down looking at your Google+ account (using your location and account history), language use (both syntatic and semantic) and synonyms.

Finally, five other factors determine the results: site structure relations, page structure relations, external link relevance, internal link relevance, and a common set of schemas for structured data markup on web pages[iv] (schema.org).

The right hand side of the browser shows the result coming from the knowledge graph.

The left hand side is a combination of literal search results in accordance to their PageRank and relevance and traces of semantics through the use of synonyms in the rich snippets of the results. In February 2012, 5% of the searches came from semantic.

 


[i] A collection of web pages stored to respond to search queries

[ii] A separate database with the ability to differentiate between words and phrases with different meanings and finding out their relation ship to each other

[iii] Wikipedia Semantic search, http://en.wikipedia.org/wiki/Semantic_search, last retrieved October, 19th

[iv] Wikipedia on Schema.org http://en.wikipedia.org/wiki/Schema.org. last retrieved October, 19th

Estrategia de Business Analytics en retail

Los principales objetivos que aportan valor a una estrategia en retail pueden ser:

  1. Análisis de los patrones de comportamiento del consumidor como la experiencia de compra
  2. Desde el punto de vista y en coordinación con los proveedores, la gestión de tiendas, clientes, categorías y disponibilidad, fundamentalmente para toma de decisiones futuras.

Entre los elementos de Business Analytics que pueden ser incorporados y aportar valor destacamos:

  • Visualizaciones de datos avanzadas en el cuadro de mandos. Las novedades en cuanto a una  a las visualización más avanzada incluiría histogramas, mapas de calor, gráficos de burbuja. Así como una mayor interactividad y exploración con los gráficos en sí.
  • Data mining y análisis predictivos. En concreto, la posibilidad de que los resultados de los mismos se incorporen en la herramienta de BI (y no tanto que el software lleve a cabo dichos análisis predictivos)
  • Web y social media analytics
  • Análisis on line de la reputación
  • Inclusión de sistemas de recomendación
  • Incorporación de sistemas de previsión empresarial y simulación de escenarios

Todos ellos implican la necesidad de un análisis continuo en tiempo real y la elaboración de informes predictivos a partir de los mismos que permitan llevar a cabo una toma de decisiones que aumente la rentabilidad de la empresa

RETOS

La implantación de una estrategia de este tipo no debe olvidar los retos a los que se enfrenta (y que deben ser requeridos en la solución que se demande al mercado). El principal es analizar correctamente el punto de partida o estado de madurez de las herramientas analíticas de las que dispone la empresa antes de avanzar.

Dicho lo cual, business analalytics es ambicioso porque:

–      Requiere de grandes volúmenes de datos que además sean de la calidad necesaria. Es necesario garantizar en el proceso de integración dicha calidad así como la conciliación con los datos existentes. La cantidad de datos exigirá también un estudio de su tipo de almacenamiento

–      Es esencial asegurar la escalabilidad en la integración, almacenamiento y análisis de los datos

“In god we trust, all others bring me data”

Deming, an American statistician, has made the sentence rather famous. From Erich Schimdt (former Google´s CEO) to Barry Beracha (former Sara Lee Bakery Group´s CEO), they all have used it.

At the heart of the matter lies the importance of analytics today. Its role has shifted to the heart of the decision making process. Decisions are based on data.

Indeed, in some sectors (consumer products, finance, retail, travel or entertainment) analytics was the key to success many years ago (in firms such as Marriot, Amazon, Harrah´s, the Boston Red Sox, Capital One, UPS, Barclays Bank, Procter and Gamble) so that nowdays reluctant firms have been obliged by the market to be more analytical ( ). However, not all the companies have been succesful in implementing and adopting analytics.

CHARACTERISTICS

These companies have three differential attitudes:

  • Widespread use of modeling and optimization. Companies look for a comprehensive understanding of their customers, prices, and products through predictive modelling (instead of descriptive statistics).
  • An enterprise approach. Data needs to be consolidated in just one source (instead of spread in different departments). It is also important to consider having a centralised human resources group dealing with the data and the visibility of  analytics within the company.
  • Senior executive advocates.  However important everyone agrees analytics to be, it needs a “big push” from top level executives.

PRACTICES

  • The right focus: different leading points
  • The right culture: achieve a companywide respect for analytics: measuring, testing and evaluating. For instance, base decisiones on hard facts
  • The right people: analytical talented, business oriented and relationship skilled
  • The right technology: competing analytics require a technology which tales into account or includes a data strategy, business intelligence software and high capacity computing hardware

WHICH ARE THESE COMPETITIVE ADVANTAGES?

COSTIn terms of efficiency, and being activity based.

PRECISION

REPEATABILITY AND COMPLIANCE

AGIL

 

HOW TO IMPLEMENT THEM?

Investment in technology

Storing and builidng a data management strategy

Data culture

Timing/Long term perspective: even if a company decides to implement an analytics strategy,  people and technology will need to be adjusted, that is to say: the company needs to give training to its existing employees, hire new “fresh” ones and preserve managers. “On the technical side”, as has been previously said, data management (from integration to quality control and tests)  requires large periods of time.

 

WHERE DO YOU FIND ANALYTICS IN A FIRM?

SUPPLY CHAIN

PRICING

HUMAN CAPITAL

PRODUCT & SERVICE QUALITY

FINANCIAL PERFORMANCE

R & D

CUSTOMER (Selection, Royalty and Service)

Source: