Introduzione: Verso un nuovo Realismo Digitale

INTRODUZIONE – Verso un Nuovo Realismo Digitale | pdf

“È tempo di muovere le riflessioni tecnologiche e di business contemporanee verso una nuova prospettiva ontologica. Impiegherò, in questo percorso, il termine “nuovo realismo digitale” per indicare questo programma di ricerca. Sono consapevole che storicamente, l’idea di un “realismo digitale” ha avuto un fondamentale e diverso dominio di applicazione. Con realismo digitale si è indicata e si indica, tutt’ora, la capacità dell’arte contemporanea (soprattutto fotografica, videoludica o filmica) di simulare, con perfezione sintetica estrema, il “reale”. In questa tradizione teorica, immagini sintetiche “realistiche” ricercano e ricostruiscono (attraverso le tecnologie digitali) una verosimiglianza, la più prossima possibile, con il loro referente reale (un volto, un corpo, un movimento, un paesaggio, una scena). Chiarisco, qui, che non è questo il realismo cui si fa riferimento. Potremmo, forse, definirlo anche “realismo postdigitale”. Il nuovo realismo digitale di cui parlerò, invece, non descrive un processo di “simulazione”, ma bensì un processo di “attuazione” (Cosimo Accoto, 2014, continua..)

WEBOOK 2013 (cap.4 by Cosimo Accoto)

“A giugno 2013 il termine ‘Big Data’ è stato incluso ufficialmente nell’aggiornamento online trimestrale della Oxford English Dictionary. La definizione descrive il fenomeno come “dati di una dimensione molto grande, in genere nella misura in cui la loro manipolazione e gestione presentano notevoli difficoltà logistiche”. Un anno prima, due delle principali fonti scientifiche e di business avevano dedicato al tema un numero speciale. Nel mese di agosto 2012, Significance (la rivista bi-mensile pubblicata sul conto della Royal Statistical Society e dell’Associazione Americana di Statistica) ha pubblicato un numero dedicato ai Big Data parlando delle dimensioni scientifiche del tema. Due mesi dopo, Harvard Business Review ha pubblicato ‘Big Data: The Management Revolution’ coinvolgendo manager ed esperti per discutere il lato di business e management del fenomeno. L’anno prima ancora (2011), due importanti report avevano acceso il dibattito: un rapporto di 156 pagine di McKinsey che analizzava i Big Data come ‘la prossima frontiera per l’innovazione, la concorrenza e la produttività’ e l’inclusione dei Big Data nell’hypecycle di Gartner per le tecnologie emergenti. Da ultimo, anche il journal accademico Marketing Science ha individuato nei Big Data una top priority emergente nei trend delle ricerche e analisi (vedi figura 1). Eppure, come il filosofo dell’informazione Luciano Floridi ha recentemente detto “non è chiaro che cosa esattamente con il termine big data si intenda”. Ma, cosa sono allora i “Big data”? (continua…Cosimo Accoto, 2013, Big Data tra hype e realtà, Webook 2013, I Quaderni di Comunicazione, from p.34)

Judah Phillips on “dashboarding”

“Dashboarding is a type of reporting; however, it is different enough from traditional reporting to outline it as its own analytics activity. Although there is no rule of thumb, you may have 100 reports but only one dashboard. Or you may have 10 reports and only one dashboard. The relationship is always that you have fewer dashboards than reports. RASTA-reporting principles can also be applied to dashboards. Dashboarding requires special handling and treatment so that the most relevant, useful KPIs and data are presented in such a way that the data can be explored. The high-level KPIs can be drilled into and explored. The concept of LIVES dashboarding presents an easy-to-remember mnemonic for creating useful dashboards:

• Linked: Dashboards may be delivered in hard copy, but it is more common to view dashboards via a browser or application. Hyperlinking then becomes key to dashboarding because linking can be used to link to other relevant artifacts, such as detail reports and written analysis about the business condition expressed in the dashboard. It is becoming more frequent to see mobile dashboards and even apps for reporting data analysis and dashboarding.

• Interactive: Although it is common to see dashboards that do not enable the exploration of the data within them, the best and most useful dashboards enable drilling down and filtering into the data from the charts and graphs. Often these drilldowns are into detailed data or secondary KPIs.

• Visually driven: Whereas reports are mainly composed of columns and rows of data, dashboards communicate data through charts and graphs. A strong visual narrative using data visualization best practices and clear information design and user experience is always helpful for dashboarding.

• Echeloned: Organize information and data presented in dashboards by relevance and priority to the audience. Put KPIs and other visualizations on your dashboard in the best position for the culture. For example, English speakers look up to the top left on a page, but Hebrew speakers look up and to the right.

• Strategic: Dashboards are not supposed to simply include total counts of this or that metric or data point. Instead, they are supposed to quickly communicate important numbers, KPIs, trends, and data visualizations to the business. These KPIs and data visualizations must be tied to business strategy. The tactics of the business, thus, influence the movement of the data on dashboards. And the movement of data in one direction or another should indicate tactical success or failure to help pinpoint the outcomes of current strategy”

From “Building a Digital Analytics Organization. Create Value by Integrating Analytical Processes, Technology, and People into Business Operations (by Judah Phillips, p.162-163)

Silos Can Threaten Big Data Strategy

“C-level support is also essential to the success of any big data strategy. Why? Because many brands are still working to develop a consolidated view of their customers, and, all too often, that work becomes a struggle. Then, without strong executive leadership, this struggle can pit departments against one another, and give rise to internal turf wars. Maybe that’s not surprising, considering that customer data is typically collected and owned by a variety of different departments. Marketing collects demographic data; product/customer support keeps customer satisfaction data; finance captures transactional purchase data; and so on. Plus, IT is usually involved in data management and control across multiple departments, which means that IT often also has ultimate ownership over data. In fact, Teradata’s recent research shows that less than one-third of marketers own and control customer data, while more than half rely on IT to access their data. In addition to validating the need for partnership between the CMO and the chief information officer (CIO), these results indicate how crippled data-driven marketing would be in a company that can’t squash turf wars and smash through interdepartmental silos.

I’m a realist, and I understand that turf wars will always exist to some extent. As Hayzlett says, “Wherever you have people and systems, there will be turf wars.” For me, the key is to rise above them. Here’s a tip from Hayzlett’s playbook: “Be clear what business objectives you are trying to drive,” he advises. “Emphasize that you’re there to serve the company, to help the company deliver on the promises you all agreed to make. That is what the focus should be on—what the group is trying to accomplish—not what’s yours or what’s theirs. This approach disarms the turf war, since turf wars break out around personal issues” (from, “Big Data Marketing, Lisa Arthur, 2013)

http://eu.wiley.com/WileyCDA/WileyTitle/productCd-1118733894.html

[…] Measurement can make the community. By accurately tracking and understanding your community and its trends you can assess the health of the community, determine where it needs to go, and align its success with that of business goals across your company. The metrics that will be available to you may be platform-dependent, but there are common measures that fall into categories such as population, activity, and value. Your business needs will determine the format and frequency in which these are delivered. Your business needs also drive what tools you use to report. If you are delivering periodic reports, Microsoft Excel may suffice. But if you are reporting across multiple platforms, then you may need more sophisticated tools such as mashup reporting tools or an enterprise data warehouse. Likewise, if you are measuring and reporting on lists and counts for your metrics, you can achieve that with spreadsheets and databases. But if you are looking for deep data analysis of the unstructured data in your community forums, you will need specialized tools for collection and processing. As always, being prepared will help ensure success. Setting a baseline and understanding the context of your metrics will support your efforts to explain the results to your executive sponsorship and quantify the value the community health has to other business units. All levels of your company need continual proofs of the proposition that your communities give you unparalleled access to your company’s most valuable asset—your customers (from Brooks, Lovett, Creek, “Developing B2B Social Communities”, 2013:181)

[…] As a leaky pipe for communication, Enterprise Social Media (ESM) create special opportunities for analyzing social relations and producing insights based on social analytics. The digital traces of communication can be processed with algorithms that can help employees make connections, and help managers understand the organization’s informal information economy. A study by Green, Contractor, and Yao (2006) showed how a social networking application with algorithms to make emergent associations between people and user-generated content spurred cross-boundary interactions and knowledge sharing in environmental engineering and hydrological science research. This increased collaboration occurred because once users learned that others were interested in similar topics to them individuals were more willing to work to overcome disciplinary differences and understand one another, even if they did not share a common store of domain knowledge. The use of digital communication traces that have leaked out of secure channels and are available for mining with machine learning algorithms can also have disadvantages for organizational action.. (from “Enterprise Social Media: Definition, History, and Prospects for the Study of Social Technologies in Organizations, Paul M. Leonardi et alii, 2013)

[…] A specific area of cultural readiness is the ability to collaborate. This activity measures the amount of collaboration or other cooperative behaviors in existence and, in some organizations, may include Facebook-type constructs or even Twitter. This assessment is important when content management, document management, and workflow are within the realm of the DG team. Additionally, the assessment is handy if the business has picked up on social networking as a possible enabler of business goals. This assessment is usually done via examination of the technology available, its extent of deployment, and its usage. Additionally, a brief survey, similar to the “Change Capacity,” can be used to see if the organization even wants to collaborate. This is not a trivial subject. As organizations become more sophisticated in their ability to reach across organizational boundaries, the need to leverage and manage the collaboration increases. There is also an opportunity to improve how an enterprise makes decisions by instituting and managing collaborative and social technologies. If anything like this is on the enterprise radar, then this assessment should be considered. Lastly, often companies have a situation where SharePoint or Lotus files are out of control. This assessment offers a chance to zero in on this issue” (Data Governance, Ladley, 2013: 78)

From “Principles of Big Data” (J.J. Berman, 2013: p.xxiii, Morgan Kauffman)

“Generally, Big Data come into existence through any of several different mechanisms.

1.  An entity has collected a lot of data, in the course of its normal activities, and seeks to organize the data so that materials can be retrieved, as needed. The Big Data effort is intended to streamline the regular activities of the entity. In this case, the data is just waiting to be used. The entity is not looking to discover anything or to do anything new. It simply wants to use the data to do what it has always been doing—only better. The typical medical center is a good example of an “accidental” Big Data resource. The day-to-day activities of caring for patients and recording data into hospital information systems results in terabytes of collected data in forms such as laboratory reports, pharmacy orders, clinical encounters, and billing data. Most of this information is generated for a one-time specific use (e.g., supporting a clinical decision, collecting payment for a procedure). It occurs to the administrative staff that the collected data can be used, in its totality, to achieve mandated goals: improving quality of service, increasing staff efficiency, and reducing operational costs.

2.  An entity has collected a lot of data in the course of its normal activities and decides that there are many new activities that could be supported by their data. Consider modern corporations—these entities do not restrict themselves to one manufacturing process or one target audience. They are constantly looking for new opportunities. Their collected data may enable them to develop new products based on the preferences of their loyal customers, to reach new markets, or to market and distribute items via the Web. These entities will become hybrid Big Data/manufacturing enterprises.

3.  An entity plans a business model based on a Big Data resource. Unlike the previous entities, this entity starts with Big Data and adds a physical component secondarily. Amazon and FedEx may fall into this category, as they began with a plan for providing a data-intense service (e.g., the Amazon Web catalog and the FedEx package-tracking system). The traditional tasks of warehousing, inventory, pickup, and delivery had been available all along, but lacked the novelty and efficiency afforded by Big Data.

4.  An entity is part of a group of entities that have large data resources, all of whom understand that it would be to their mutual advantage to federate their data resources.  An example of a federated Big Data resource would be hospital databases that share electronic medical health records.

5.  An entity with skills and vision develops a project wherein large amounts of data are collected and organized to the benefit of themselves and their user-clients. Google, and its many services, is an example (see Glossary items, Page rank, Object rank).

6.  An entity has no data and has no particular expertise in Big Data technologies, but it has money and vision. The entity seeks to fund and coordinate a group of data creators and data holders who will build a Big Data resource that can be used by others. Government agencies have been the major benefactors. These Big Data projects are justified if they lead to important discoveries that could not be attained at a lesser cost, with smaller data resources” (J.J. Berman)

DATA GOVERNANCE AND GOVERNANCE

[…] The concept of managing information assets in a formal manner has been established. Now we need a process to ensure that management actually takes place and is being done correctly. Unplug your technology thinking and turn on your accountant thinking. Accountants manage financial assets. Accountants are governed by a set of principles and policies and are checked by auditors. Auditing ensures the correct management practice of financial assets. This is what data governance (DG) accomplishes for data, information, and content assets. DG is defined in the DMBOK as, “The exercise of authority, control, and shared decision making (planning, monitoring and enforcement) over the management of data assets.” In turn, governance is defined as, “The exercise of authority and control over a process, organization or geopolitical area. The process of setting, controlling, and administering and monitoring conformance with policy.”1 This definition is, of course, roughly synonymous with government. Slightly different definitions are often stated with an emphasis on the policy and programmatic aspects of DG. The one we use in our consulting work is, “Data governance is the organization and implementation of policies, procedures, structure, roles, and responsibilities which outline and enforce rules of engagement, decision rights, and accountabilities for the effective management of information assets.” Regardless of style of definition, the bottom line is that DG is the use of authority combined with policy to ensure the proper management of information assets. Make sure you do not confuse the management of data with ensuring data is managed […] from “Data Governance. How to Design, Deploy, and Sustain an Effective Data Governance Program” (John Ladley, 2013, p.11).