In a few days Oracle will release Data Sync, a utility that facilitates the process of loading data from various on-premise sources into BI Cloud Services (BICS). The Data Sync utility comes as an addition to several other already available means of loading data into BICS: data load wizard to upload data from files, SQL Developer cart deployment feature, and RESTful APIs. What is special about Data Sync is that it makes the effort of data loading more manageable from a capability and scalability perspective (full loads vs incremental loads of many tables), and its scheduling, notification and automation aspects as well. This approach makes data loading a lot closer to conventional ETL methods. And if you ever worked with DAC before to run execution plans of Informatica mappings, you will find that Data Sync follows a similar methodology of defining connections, tasks and jobs. However, instead of referencing Informatica mappings, as is the case with DAC, Data Sync itself handles the mapping of source columns to target columns. And it supports SQL Queries so your opportunities for doing data transformations are endless. In this blog I present the key features of Data Sync.
The whole Data Sync software sits in one place. There is no server/client components to it. So you basically extract the zipped file you download from Oracle onto a machine on your network that can connect to the different data sources you wish to source data from. Once you unzip the file, you configure the location of your java home and that’s it! You are ready to launch Data Sync and start moving data.
There are 3 main steps to follow when designing data loads in Data Sync. Below the item menu there are 3 corresponding tabs to configure: Connections, Projects and Jobs. Read the rest of this post »
With an Oracle Business Intelligence Cloud Service (BICS) subscription, you get accesses to two instances: Test and Production. The test instance can be the initial playground where the upfront development work is carried out before pushing the developed components to the Production instance. The development work may entail creating tables in the Oracle Cloud Schema Database, loading data, creating data models, reports and dashboards. You may very well find yourself with a fully functional system on the Test instance and now thinking of how to migrate everything to Production. This blog elaborates on how to achieve such a task.
I recently got published in the Special Big Data Edition of CIO Story (see page 20), where I talked about the “six” essentials for transforming into an Information-driven organization.
Information is a hot commodity. Research suggests that in the next two to three years, businesses will begin to apply monetary value to their information assets by trading or selling them. Throughout history, this notion has been referred to as “Infonomics.”
The principles of Infonomics are based on the premise that information has both potential and realized economic value, which can be quantified and should be managed as an asset. The benefits in doing so include improving the collection and use of company information, determining how much to spend on business or IT initiatives, and improving relationships with customers, employees and partners by sharing better information with them. More and more organizations realize that the trick to experiencing these benefits by better managing assets, however, is to effectively apply the organization’s existing experience in managing other assets toward managing information assets. But, in order to get to that point, executive leadership (business and IT) needs to recognize the barriers to becoming an information-driven enterprise while focusing on certain fundamental strategy essentials.
If organizations are serious about improving the value and speed of information, they must consider the following six imperatives. Doing so will drive their organization’s ability to become an info-centric enterprise:
Read the rest of this post »
All of this focus on the Internet of Things (IoT) is really about the “Internet of Me” (IoM). From social media sites to smartphone apps and GPS systems, loads of data are being generated today about individuals – their interests, their travels, their behavioral patterns, their purchases, and so on. No one in this digital economy can afford to ignore the demands of the “me” generation. It is no longer good enough to tailor marketing based on customer demographics alone. All interactions now need to be customized to your customer’s specific situation and emotions.
With all of this digital interconnectedness, one thing that is very clear is that customer loyalty is at risk. Comparison shopping is as easy as a few mouse clicks, and previously loyal customers can quickly discover new products, new services and new vendors, and learn what other buyers like and dislike — all without leaving their laptops and other mobile devices.
Research shows that more than 50% of consumer interactions are now occurring in this multi-event, multi-channel environment. But, 65% of consumers get frustrated by companies that do not provide a consistent experience through these various media. Those firms that put a priority on the consumer experience and can provide consistency regardless of source have been shown to generate 60% in additional profits versus their less enlightened competitors.
The bottom-line is that a brand is no longer simply what we tell the consumer it is. “It’s increasingly what consumers tell each other it is.” So, how do you ensure brand competitiveness in such a volatile environment?
Well, this is where competitive organizations must recognize and strategically embrace the emerging nexus of social, mobile, cloud and information, where Big Data and advanced analytics serve as revolutionary ways of advancing the digital ecosystem. We must therefore look for big data opportunities across the nexus shifts, and in turn craft a vision that takes into account more of an all-encompassing “personalization” perspective.
It is imperative that we leverage these newer data sources and types for personalizing the digital experience for each customer by considering the environmental factors and circumstances that surround an individual use case (so contextual), or then a customer’s previous interactions to provide an evolving experience that spans across interactions (behavioral), in attempts to align a customer’s preferences with those of a pre-defined target persona through that journey! So, persona and journey- based personalizations.
Therefore, it is pertinent that customer-centric organizations be able to create a continuous, seamless virtual cycle of targeting the right customer with the right offer @ the right time by looking at avenues to enhance the existing 360- degree view of the customer. Initiatives focused on such a view have gone a long way toward providing those benefits by synthesizing customer profiles, sales and other structured data from multiple sources across the enterprise. But today, there is more opportunity for growth when you enhance that view with information from more sources, both within and beyond the enterprise. Information in email messages, unstructured documents, web logs, machine data, and social media sentiments – previously beyond reach – is now extending this view.
Organizations that make full use of these data sources can deliver better insights and a sharper competitive edge by making the customer’s experience more personalized, thereby encouraging loyalty and accelerating sales. Therefore, an enhanced 360-degree view of the customer is a holistic approach that takes into account all available and meaningful information about the customer to drive better engagement, more revenue and long- term loyalty. It combines data integration, data exploration, data governance, data access and analytics in a cohesive solution that harnesses the volume, velocity and variety of Big Data. To establish such a view of the customer, you must be able to:
• Eliminate duplicates and rationalize conflicting information through matching, linking and semantic reconciliation of master data to create and maintain a golden record
• Integrate high-quality data across multiple enterprise systems
• Manage new data types and navigate quickly through massive amounts of both structured and unstructured information from within and beyond the enterprise to find the most pertinent information
• Creating a single, up-to-date view of customers and other key entities that can be used throughout the organization, all by Leveraging Hadoop systems so that information of all types, in any volume and at any velocity, can be incorporated into the single view
• Assessing streaming data sources to analyze perishable data quickly and to select valuable data and insights to be stored for further processing
• Federate search, discovery and navigation securely across a wide range of applications, data sources and formats
New analytic opportunities are driven from this centralized, data lake architecture, where Hadoop is increasingly being leveraged as an enabler. What is critical here is to make the analytical process as specific as it can be to each customer’s digital journey by leveraging capabilities such as advanced customer segmentation, predictive and prescriptive analytics to enable cross-sell and up-sell, along with next best offer generation, thereby helping you create and evaluate the consistency of that experience across relevant products and channels.
In essence, digital transformation needs big data analytics technology, at-rest and in-motion, in order to enable a deeper level of analysis across various touch points: Mobile, Social, Web, Multi-channel. And, in order to make that happen, an effective analytical process tied to distinct digital data architectural capabilities needs to be created and implemented.
If history is any indication, companies will encounter false starts in Big Data initiatives, like we did in the early Data Warehouse days. I see similar confusion in terms of the tools, types of solutions. The variety, volume and veracity of new tools and technology companies offering solutions are enormous. For starters Big Data is tech heavy and geeky. One should know when you see the green screens and Unix/Linux prompts and wonder whatever happened to the GUI. Part of the reason why getting business value out of it is difficult.
But not to worry, technological advances and lessons from the history, will keep us straight with options. Before we talk about that let’s look at what it takes to implement a decent Big Data solution. Assuming the dream team with all the right skills at the right price are in place. Just to prove if the Data is worth producing the business value you are looking for will cost at least 715K. This does not include HW and SW. You get the picture where it is going, expensive experiment no matter how bare to the bones you get.
So how to minimize risk and know the business value for real than some hypothetical assumptions and hypothesis, especially when it is going to cost close to a million dollars. The good news is we have options. Depending on the level of or organization’s maturity in handling Big Data Analytics the following options are worth considering
Leveraging cloud option relieves the IT infrastructure delays. It also provides ways to compare different solutions. Partners bring not only technology solutions but also provide industry experts who have done this at different clients. Finally POC will validate the assumptions or even level set the expectations.
Many companies have invested millions in building a successful BI / EDW and are investing in advanced analytics for the future. But the mystery remains about the data quality. Though glaring DQ issues might be contained through constant backend data corrections or through exception handling, many organizations still faces the challenge of poor data quality.
The reason Data Quality does not get addressed in many organizations because of several reason. Typically you find:
So the problem gets buried in various facets of the organization. Everybody knows the problem but no one will step up to own it or sponsor to fix it permanently. The more efficient IT is, harder it is to build a business case for DQ tools or initiating DQ projects.
Having a Data Governance organization becomes very critical in bringing the business and IT together. This is the forum where business and IT can work together to solve the DQ issues and define the ownership and accountability. A day a month of cleaning up each person who uses the data adds up quickly in terms of hours and not to mention the data discrepancies due to manual changes to data.
Matured organizations understand the DQ issues and implements the DQ as part of the overall development / operations. It is an expensive affair if the DQ goes unchecked. One time cleanup of data will slowly decay over time to right where we started. Investing in setting up DQ metrics, data ownership and other quality related policies enabled by appropriate tools is the right way to solve the Data Quality issues. DQ does not mean perfect data but good enough data to do the analysis for right decision-making.
Whenever OBIEE encounters a run-time error, instead of getting a report you typically see an error message like this on screen:
Expand “Error Details” and you are provided with several lines of query information.
For an administrator, the more information the better so one can figure out what went wrong without having to search through sessions log files to identify the problem. However, this much information may be a lot more than what a typical report user would like to see when a failure happens. In addition, the default error messages reported on the screen are more tailored for IT, someone who reads “ODBC” and deduces that the problem is related to the data source. Is it possible to make these error messages more user-friendly for business users? The answer is Yes.
I have split the default error message into 3 sections as denoted in the image above. Here is how to customize each section.
Users denied this privilege see a much shorter message if they expand “Error Details”. Users who are granted this privilege, such as report developers or administrators, can still see the full message with the query. Here is what it looks like when the privilege is denied.
For example: E:\OBIEE\Oracle_BI1\bifoundation\web\msgdb
Generally when customizing OBIEE messages, you will need to find the file under the above out-of-the-box directory that includes the messages you are customizing. Once you have located the file, copy it into a custom folder and customize the message in the copied file. The reason why the custom file needs to go under a custom folder as opposed to the out-of-the-box msgdb folder is to avoid having any customization overridden when OBIEE patches/updates are applied.
Similarly for the error messages I referenced earlier, create a custom directory called customMessages under analyticsRes to store the custom message files. The folder structure under customMessages should resemble the folder structure of the corresponding files under the out of the box messages folder.
Let’s go back to my earlier error screenshot and in particular Section A which contains the message: “View Display Error”. This default message is defined in:
Copy this file over to the custom directory:
Modify the copied file by searching for the following line and replace the text between the HTML tags with your custom message. You may also leave it empty to not display this message if you chose to do so.
<WebMessage name=”kmsgEVCViewDisplayErrorTitle”><HTML>View Display Error</HTML></WebMessage>
Copy this file over to the custom directory:
Modify the copied file by searching for the following line and replace the text between the TEXT tags with your custom message.
<WebMessage name=”kmsgOdbcAccessOdbcException”><TEXT>Odbc driver returned an error (<sawm:param insert=”1″/>).</TEXT>
Once done with all the above changes, restart BI Presentation Services for the custom files to be picked up by the server. Your custom error messages should now show up instead of the default messages.
As new companies embark on the Digital Transformation leveraging Big Data, key concerns and challenges get amplified especially for the near term before the technology and talent pool supply adjusts to the demand. Looking at the earlier post Big Data Challenges, the top 3 concerns were:
Big Data Skills can be broadly classified into 4 categories:
The value creation or the monetizing of the Big Data (see Architecture needed to monetize API’s) depends on the Business and the Analytical talent. See talent gap on the right specifically in the analytical area. Educating and augmenting the talent shortage through partner companies is critical for the niche and must have technology. As tools evolve coping up with the Architecture becomes very important as past tool / platform short comings addressed with new complexities.
While business continues to search for the Big Data gold, System Integrators and Product vendors are perfecting the methods to shrink the time to market, best practices and through Modern Architecture. How much of the gap we can shrink depends on multiple factors of Companies and their partners.
See also our webinar on: Creating a Next-Generation Big Data Architecture
As companies start adapting to handle Big Data, the challenges still remains. Barring the obvious applications, the challenges of getting the value out of the new-found data continues to be on the top of the list. ROI’s and potential revenues are yet to be realized. As the technology and the usage becomes more sophisticated we will start to see the results.
From IT perspective top two challenges are Governance and Skills. Securing the Big Data for greater use within the organization is complex and the technology is evolving. Securing the right people with in-depth knowledge in managing Big Data bigger challenge. And these two aspects will feed into the bigger challenge of ‘How to get value out of Big Data’.
Organizations find that the key resources who needs to be driving the Big Data are also the same resources who are so vital to managing existing core enterprise applications. Balancing the precious time of key resources and leveraging external thought leadership / expertise is key to successful Big Data initiatives.
Identifying the strengths of the organization and prioritizing the critical areas of investment is key to successful Big Data initiatives.
IT spending is primarily focused on technologies to run the business primarily operations. With new ways of doing business, technology platforms decide the winners and losers. Typical brick and mortar versus online stores. If you look at the CIO’s budget, more than 70% goes to operational systems, infrastructure and keeping-the-lights-on-type-of applications, and the rest is spent on customer-facing applications/systems.
With Digital Transformations happening at many enterprises, the shift in IT budget is also tracking the trend. Customer experience is one of the key strategies for successful companies. With smartphones and tools for accessing information, customers are one step ahead of the traditional organizations. Investing in new technologies like Big Data, Fast Analytics and pro-active customer experience strategies through converging technologies are not just futuristic but has to be fully functional now.
CIOs are looking for ways to invest in new technologies for enhancing customer experience and leveraging the data (internal and external) to accurately deliver customer experience not just operational systems. As more and more CIOs get invited to the business leadership table, business technology investment becomes a strategic asset to manage, leverage and deliver greater customer experience. (see spending shift in CIOs face the “Age of the Customer” ).
Connect with us on LinkedIn here.