Data mapping, common taxonomies, and inferences from sparse data are best addressed by using semantic technologies (ontologies). Ontologies not only store data, but store hierarchical relationships between concepts. Thus instead of querying data, the end user queries knowledge about the data. For obvious reasons, ontologies are one of the foundational structures in the Semantic Web (Web3.0).

I feel organisations still rely much on the informal channels of information rather than the formal ones. Above all they have their own justification for taking this path. The efforts of many knowledge workers lie unused or under-used.

To this interesting article one may want to add the organisational culture aspects of MIS and other BI tools, be it as examples of organisational artifacts or as a means to (i) change the “way things are done around here”, or (ii) to create a common organisational language and understanding, with the ultimate goal to align efforts and ensure high and sustainable performance.

The authors of “Data to dollars” respond to reader comments

This piece is focused on treating knowledge-as-an-object. Knowledge can be made explicit and stored in repositories so that it can later be retrieved and reused. The emphasis is on codifying knowledge, interacting with and managing a repository rather than on people and their interactions. In contrast, the knowledge-as-action view emphasizes communicating knowledge. Knowledge is no longer static content with a well-understood context, but is instead situated or acted on; the interpretation of the knowledge is solely dependent on how and who acts on it.

The demands placed on technologies from these two views are at odds. On the one hand, the spontaneity of a knowledge-as-action view demands flexible tools, integrated infrastructure, and fast retrieval of knowledge. On the other hand, a knowledge-as-object view emphasizes broadly applicable knowledge and careful encoding and storage for later retrieval.

Effective and value driven operations is a goal for all organizations; achieving it is a challenge. A variety of professional organizations can provide your readers support. Professional associations are a resource that CIO’s and CEO’s can tap to jump-start their initiatives in a cost effective way. Data management activities are the foundation for business intelligence initiatives. The DAMA Data Management Body of Knowledge represents best practices from more than 100 data management practitioners, along with references to topic-specific resources; it is similar to the PMI PMBOK for project management. DAMA International ( www.dama.org ) is a not-for-profit, vendor independent association of technical and business professionals dedicated to advancing the concepts and practices for data resource management and enterprise information. The primary purpose of DAMA International is to promote the understanding, development, and practice of managing data and information to support business strategies. DAMA International has chapters and members throughout the world. The professional organizations for The Data Warehousing Institute (www.tdwi.org), Data Quality (www.iaidq.org) or even content management (www.aiim.org) may yield great value to your readers.

(1) Business Case - The ability to quantify (within any reasonable margin of error) the benefit of such an investment versus the status quo is a major hurdle in determining its value and importance. The business case usually is made on the non-financial benefits which is difficult—yet essential—for most non-IT executives to value. The CIO should be helping to sell this to the business if it makes sense but the business must buy it wholeheartedly for it to work.

(2) Strategy to Execution - The other aspect to be conscious of is that these types of initiatives are highly dependent on the requirements, design, and implementation. KPI identification, architectural design, and implementation are not all created equal and whoever is running such an initiative will have a great impact on it’s success or failure.

(3) Sponsorship and Appetite - Lastly, the information is useless if it’s not used appropriately which makes it imperative that the business is highly engaged and see it as a business rather than an IT initiative. If you build it (alone) they are less likely to come, so to speak. The reality is that in a large and complex organization such an initiative is likely to impact current processes by introducing more granular and better related data and cut out manual data cleansing and reconciliation. This requires that senior management sponsor the initiative outside of IT as the real benefit comes when business process and thinking is augmented to take advantage of the new information and capabilities. In addition, an initiative such as this will occupy many of your analytical people including senior management for KPI identification, requirements and testing. Many, if not most, organizations that take on a cross-company information transformation initiative will fail the first go around likely due to not addressing one or more of the above challenges.

Cannot a unified ERP address the preliminary part of the whole issue at hand? Standardization and KPIs can be tailored into the system with the use of a single ERP and the next step would be to design appropriate reporting standards specific to the organizational needs to tap into the aspect of organizational intelligence.

The approach presented in the article may be very appropriate for certain types of large corporations with centralized IT organizations. In our conversations with CIOs and business executives of medium-sized and Global 1000 companies we find that the paradigms for analytic applications simpler reporting solutions (and the associated information architectures) are changing dramatically from the ones that had been established as late as 5-7 years ago. We are moving from centralized information architectures exemplified by enterprise data warehouses to more distributed ones. Cloud-based information architectures (particularly the growing adoption of private clouds) such as the ones supported by Vertica and Aster Data, SaaS business intelligence applications such as those offered by Host Analytics and Pivotlink, as well as the move towards the consumerization of enterprise applications (see applications such as Jive as an example) are the catalysts for the changing nature of the next-generation business intelligence solutions.

Reminds me of a course I took in my MBA school in the 70’s, “Seven Myths of MIS”. One of the myths was: “You need information to make a decision”!! The idea here is not that information is useless, rather, a management decision is required only when there is lack of information. Decision making is a person-centric process, not data-centric. We often find the same piece of data being acted upon differently by different people.

Unless IT clearly understands the business management process this issue of IT versus business will continue forever. That the stereotypical CIO is not a business person is a case in point. No organization that has a non-business-savvy CIO can ever hope to build a useful IT organization. IT needs to know firsthand how business is actually using the systems provided by IT to manage day-to-day business operations. If this is not the case, business will neither trust IT nor rely upon it. This can explain the sense of inferiority IT feels in many companies. The positive examples we can quote in this context are Walmart, P&G, and GE while the negative ones include companies like GM. I also recall, in the 80’s when decision support systems were beginning to be deployed, a leading consumer goods company always paired-up the brand manager with an IT business systems analyst, so that they worked together in analyzing the huge volumes of sales/inventory data created on a daily basis, to do what-if simulations and make decisions on pricing, promotions, etc. The productivity gain for the business user was immense and the knowledge of the business processes gained by IT was equally immense. The synergy between business and IT is created by shared knowledge and common objectives rather than by organizational structure or job descriptions.

A point apparently overlooked in this article, or excluded for the sake of brevity, is the potential for the SaaS phenomenon to engender more formalised system interfaces thus dramatically reducing integrated reporting costs—this is an important point for an executive information system positioned as “next generation”.

Possibly a minority view. My experience suggests that IT should focus on the high cost highways (enterprise computing data/infrastructure) while executives should recognize business intelligence (BI) experts understand the off-ramps to the fields and streets where the real work is done. BI experts should be embedded in the business units and key functional organizations to constantly understand evolving client needs and ever changing BI demands.

The CIO role should be to orchestrate the relationship between IT and the BI experts embedded in the business units. The paradox will always be “logical” (the way IT experts attack BI delivery challenges) does not always equal “relevant or usable”. The best BI implementations I have seen are when BI Experts are embedded in and accountable for business or functional organizations, and fully empowered to design for cross-organizational reuse, and expected to align to the highways the CIO helps ensure are available. This, hopefully combined with adequate off ramps that matter.

The differentiator between a database and EIS/MIS is the effective use of the data, and the biggest problem in this is that it’s lead by the CIO rather than the CFO or CEO!

The system needs to be driven by business rather than technology and until the time business has a buy-in in to the system, they will always work around the system.

Sometimes I think that a bit too much time and attention is given to creating the “intelligence” and a bit too little given to leveraging the intelligence. For an information management program to drive breakthrough performance it needs to focus on infusing fresh, high-quality, and relevant information into well-designed and efficient business processes.

The divergent views of business and IT operations are a standard problem plaguing the effectiveness and efficiency of data delivery. Realizing this upfront if both business and IT colloborate from requirements to delivery using agile software practices, they can bridge the gap in ensuring right data at right time.

Technology is critical to make business intelligence (BI) work well. Blending technology and business needs is what makes BI successfully taken-up and used.

I’m somewhat skeptical of such an initiative being lead by the CIO. For example, the CIO developed a taxonomy seems like a recipe for a disaster to me. The approach described is still technology lead and it is only towards the end that the business people are being brought in to any great degree. They should be more inside the process from the beginning. However, the process described is valid and it seems to describe a good overall BI project plan for any organization.

Management involvement is key; what do they want to accomplish (vision/mission/objective) and how do they want to manage getting there (strategy measured via KPIs). In this case the CIO took the lead in setting KPIs, which is good but also shows the problem in many companies. Management has a vision but lacks involvement and interest in setting by what key information they want to measure the path towards the vision. This work can not be “delegated” to IT unless there is a clear concept of how the organisation is supposed to operate and measure its results. Many IT projects & systems that fail or do not reach their full potential lack this critical first step. Very important in this first step is to discover what really drives the long term/strategic direction rather than tactical/short term issues. If you want a robust system, understand what is key and constant in your industry, and what is tactical and changing. This also shows where to spend what portion of your time and efforts.

I think the article lacks the understanding of new trends in information management and decision-making processes. Although BI is important for the construction of a comprehensive executive information system, methodologies and the representation of business processes into the design are often lacking in such initiatives.

Also to consider is that non-structured data—this is information that is not available within the structured databases or systems in the organization—is sometimes more important than strict aggregated or statistically relevant data. Case in point is that many decisions today are taken when relevant information appears in external sources like internet sites, blogs, or newsfeeds. These data have to be integrated to a comprehensive Business Intelligence platform that not only supports KPIs, OLAP analysis and Relational data reporting but also incorporates external non-structured data to any given report. No such platform is still out there yet, although some inroads are under development from several vendors and consulting companies.

An excellent article, however in these financially trying times the focus is always going to be on the benefit or return on implementing such an information management system (IMS), particularly when your existing topology is ill-equipped for easy integration. Increasingly the users of this IMS have mutually exclusive requirements from the IT function (i.e.: accurate information delivery at short notice competing with the need to directly drive down the ICT overhead cost).

Another factor for consideration is that of the type of business the IT function supports and the value of that information. In a non-profit organisation, for example, an IMS of this nature would certainly provide the necessary performance improvement but would not contribute much to strategic or business objectives. Finally, due consideration needs to be given to the scope of integration. Many organisations have a significant blend of in-house and shrink-wrapped applications and systems, written in different languages and running on different platforms. My experience with this has generally revealed that data migration and integrity retention of legacy data during this process is the most costly and risk prone element.

The article very clearly brings out the key items CIOs should look for while building the next generation Executive Information System. As discussed in the “Designing a business intelligence architecture” section of the article, the creation of a centralised business intelligence data store will help in a great way to analyse historic data and perform business reviews. The centralised datastore will provide one version of the truth across the entire organisation. This architecture can be further enhanced by integrating it with “real-time” feeds from operational or downstream systems.

Great discussion and we could not agree more. The thing I was hoping you would discuss here is the actual costs involved in the operation of the infrastructure and the ability, and from these costs suggest reliable best practices to expose and reduce. The other aspect I had hoped you would discuss is from the view of the CFO not the CIO. I say this because we consistantly see a lack of solid needs and (or) requirements for the IT department to work from.

This is an interesting case study and a great business intelligence (BI) success story. It is worth calling out a couple of the key success factors that led to the actual successful implementation. First, the initiative had executive sponsorship. Regardless of the buiness needs, technical issues, or internal drivers, executive sponsorship and engagement throughout the process is vital to acheiving the overall goals in BI projects which tend to span multiple internal organizations and take a very long time to implement. Second, the initiative was business driven rather then technically driven. The KPIs were first identfied, then the data architecture defined before the technical solution was developed. This helped ensure better adoption from the business user community since they had input into the solution and the solution was designed to delivery what the business needed. I’m sure there were many, many technical issues to overcome during the development. But the fact that the actual BI system delivered a solution that was usable by the business users was the best indicator of its success.

In the coming years, we are going to see corporate performance management reporting become a major component of information systems. With increasing support for international data standards such as XBRL, CIOs will be able to leverage a core set of taxonomies to provide exceptional value by mashing internal and external data streams.

One more step is needed and involves extending the data-to-information-to-intelligence process to include “wisdom”—by which I mean the product of executives reviewing and reflecting on the actual use of intelligence in decision-making to ascertain the contingencies that diminish or enhance the value and utility of data. “Knowledge management” may be a more modern and more sophisticated term than “wisdom,” but the latter (1) captures and depicts the reality of people seeking to use intelligence effectively and (2) the distinctive characteristics each organization brings to bear in using data, information and intelligence. A good CIO will want to capture that wisdom and make it part of the information systems she/he creates, if only to ensure accountability but also to make the organization one that learns as decisions are made and actions are taken.

It is rightly said that data plays a major role in driving an organization. However, the article lacks a very important point and that is the contribution of people with statistical backgrounds who can play a major role in handling and analysing the data over and above the IT people. As per my opinion, an organization must have a statistician along with IT people in its core team of data management while designing the architecture of its data base and report generating modules.

I really liked the problem statement. But, the answer to align existing KPI’s seems like polishing up the rear view mirror rather than helping the organization to look ahead. What’s missing is any talk of new mindsets. Of responsive leadership in the form of: agitation about the transition required, alignment around mileposts, anticipation of community reaction (inside and outside the IT office and the organization as a whole), agility in dealing with the unpredictable and ongoing disruption, more discipline in marching forward using the same indicators seems like a pretty typical IT response rather than the innovation one expects from McKinsey.

The problem does not lie in poor IT systems, it really lies on assuming that techies are supposed to know how a business runs. In my company we have an army of high-tech specialists, but before them there is a team with a business management experience which understands the design behind the IT. Blame the IT managers and most of the strategists as they still think that the organization of a company is something only for IT. I am a Finance expert with a good understanding of SAP, and I will never let my mechanic drive the Paris-Dakar Rally even if he/she is extraordinarily good at fixing the vehicle.