Posted on April 20th, 2012
by Jerry Smith
A principle consideration of most companies is the growth of revenue and margin, the top and bottom line. Just look at the modern day enterprise and one can find a wide variety of valuation-generation capabilities, ranging from innovation centers to product management. But a deeper look reveals that many are myopically focused on creating or improving their products and /or services, while woefully neglecting a rich source of value hidden in the simple zeros and ones of their massive data repositories. But why?
The answer to this question will probably come at no surprise – many companies are having a hard time valuing not only the their data assets, but any investment in their enterprise data architecture as well. Since the early 1960s we have spent and spent on numerous data architectures in search of the most efficient means of aggregating and persisting data. At the same time, the volume of data has exploded and its velocity has accelerated as people and businesses have become more interconnected (commonly refer to as the Big Data effect). As an industry, we have become very proficient at understanding only the cost-side economics of the data; but still only treat it as that, a cost.
Companies are now beginning to look at data from a revenue or asset perspective. There is a growing “belief” that data has value and, as such, is worthy of some level of enterprise investment. Conversations stemming from vendors to industry analysts, from engineering to corporate, are pushing the discussion that data is the next financial frontier. But while nobody has qualified the value of their data, all are still asking the same question, what is it worth. However, in order to quantify that value, it is important to take a brief, but necessary side trip that discusses the value of networks.
Dr. Robert Metcalf, recognized for his invention of the Ethernet, noted that the total value of a communications network grows with the square of the number of devices it connects. This scaling law, known as Metcalf’s law, in addition to Moore’s law on transistor density, is widely accredited as the catalyst that has driven the growth of the Internet. His law simply states that if you build a network so that any customer can choose to transact with any other customer, the number of potential connection of the N customers can make is (N-1), giving a total number of potential connections as N(N-1), or N2-N, which is simply proportional to N2. Metcalf’s law implies that value of a connected network grows faster than does the number of access points; that is, merely connecting two networks creates more value that substantially exceeds the original value of the unconnected networks.
In and of itself, Metcalf’s law is a very powerful business concept. However, Dr. David Reed, building on Metcalf’s work, discovered a new type of network, called Group Forming Networks (GFN). GFNs have functionality that enables and supports collaboration and affiliations (e.g., interest groups, meeting, communities) such as eBay, twitter, FaceBook, and LinkedIn. That is, they allow groups of interconnected users to come together around a common goal, interest, or issue. Reed found that these GFNs create a new kind of connective value that exponentially scales with N. Without going into to much detail, the number of non-trivial subsets that can be formed from a set of N members is 2N-N-1, which is simply proportional to 2N. Reed’s law implies that a network that supports group communication has a potential value that grows exponentially with N.
Taken together, one can see the increasing power of self-organizing interconnectivity. At its simplest, transacting with individual members has a baseline valuation that is proportional to N (Sarnoff’s law). Think of cable TV, for example. As one begins to interconnect members, valuation grows proportional to N2 (Metcalf’s law). Email and Yahoo are two prime examples of this kind of valuation. Finally, when members are allow to coalesce around common interests, the overall valuations grow as a function of 2 N. eBay and Google+ are two prime examples of the valuation achievable in GFNs. The question now remains, how does this background relate to the valuation of Big Data?
As it turns out, networks and data have a lot in common; more precisely, networks and ‘data networks’ have everything in common. Enterprise data is organized around three primary building blocks: Data Marts, Data Warehouses, and now, Data Lakes (sources of big data). Data marts, often built on relational data structures of related subjects/entities, are a dominant means of data organization. Subjects are defined, entities are designed, and tables are created, and then rows of data are loaded. This is the core repository where we can seek answers to issues that we know we know. Like with the case of Sarnoff’s law, the value of a data mart grows proportionally to the number of subjects, N.
Data warehouses, often defined as interconnected data marts over time, enables use to address issues in areas where we know we don’t know something. By interconnecting predefined data sets through multi-dimensional, cubic, data stores, we begin to see how things change and how they are interconnected. By allowing data subjects to logically interconnect to other data subjects (both dimensionally and temporally), the valuation is similar to Metcalf’s law; that is, growing proportional to the square of the number of subjects, N2.
The new frontier for valuation, however, comes from data lakes – otherwise known as big data. As data grows in size (often defined as greater than 1 TB), becomes highly distributed (through the cloud, for example), and changes in real time (customer sentiment, transactions, etc.), the ability to monetize its value goes beyond the capabilities explicitly available within relational data stores, marts, and warehouses. For example, analysis of THE Internet cannot be conducted in any one data warehouse, no matter how large or computationally intensive.
Big data is also the realm of where the most important issues can be found – it is where the stuff we don’t know we don’t know is contained. Unlike data marts and warehouses, big data does not presuppose organization or structure, nor does it selectively filter out information that is not of interest. The unstructured and unfiltered nature of data lakes allows analysts and algorithms to form around common interests in the search of knowledge. This is a kind of Data Group Forming Network (dGFN) and, as noted by Reed’s law, has a valuation that grows exponentially as a function of the number of implied subjects, 2^N.
Taken together, we see that there are three value categories that data can provide: the linear value of addressing questions that we know through relational data marts, the square value from facilitating the transactional interaction between data stores in through data warehouses order to address issues we know we don’t know, and finally, the exponential value from allowing the facilitation of very large, unstructured, time varying data in the pursuit of understanding about the areas we don’t know we don’t know. By moving up this data architecture chain, one pursues every increasing value as corporate data goes in size.
As companies look for way to grow their top and bottom lines, that value of big data (data lakes) should be of prime consideration. If only looked at from a simple return on investment – where the sunk cost of identification, collection, storage, and maintenance have already been incurred – the exponential value that could come through big data capabilities provides credible return on investment model that exceeds most targeted internal rates of return (IRR) of other investment opportunities.
Posted in Application Development Trends | No Comments »
Posted on March 9th, 2012
by Thomas Hannigan
For much of 2011, Pharmaceutical Manufacturers were working to enhance systems to properly capture information in support of the Sunshine Physicians Payment Act. Most companies already did a pretty good job of capturing that data so projects have been more about coordination and planning with vendors. The process of integrating those streams of data together to meet the reporting requirements is a task many organizations left for 2012.
Now that the integration work is underway, forward- thinking companies are discovering the added strategic benefits of spend compliance reporting.
Posted in Application Development Trends | No Comments »
Posted on February 21st, 2012
by Raymond Bordogna
In 1987, an IBM business systems planning consultant published a seminal paper: A Framework for Information Systems Architecture. Although the article’s focus was on how best to model information systems, the resultant framework was later extended to model an entire business. This extension led to a discipline coined Enterprise Architecture (EA).
EA prescribes a set of viewpoints in which to define, communicate and align the strategic intent and tactical operations of a business. These viewpoints are commonly categorized into subject areas. The Open Group, a global consortium of 400 member organizations and a leading proponent of EA defines the following 4 principal domains:
1. The Business Architecture defines the business strategy, governance, organization, and key business processes.
2. The Data Architecture describes the structure of an organization’s logical and physical data assets and data management resources.
3. The Application Architecture provides a blueprint for the individual applications to be deployed, their interactions, and their relationships to the core business processes of the organization.
4. The Technology Architecture describes the logical software and hardware capabilities that are required to support the deployment of business, data, and application services.
Despite its inherent logic, I think it’s fair to say that EA has more than struggled to gain business acceptance. And, I believe the primary constraint to be a combination of 2 factors:
1. A dearth and immaturity of business architecture viewpoints in EA programs
2. A lack of integration between business architecture and other domain architects – i.e., cross-domain architectural viewpoints.
Business-Oriented Architecture: weaving the “business” into architecture:
In our experience, research and observation, most Enterprise Architecture programs’ business architecture work streams simply create business process models. While certainly necessary, an organization needs to leverage additional business architect viewpoints* into their practice including:
1. Business Model Canvas
2. Business Strategy Canvas
3. Business Strategy Map
4. Business Process Taxonomy [Level 0, 1]
5. The Balanced Scorecard
6. Organization (Chart) Model [People]
7. Skills Inventory [People]
8. Business Social Media (Collaboration) Strategy [People]
9. Business Financial Statements
10. Business Performance Metrics
LiquidHub employs the term, Business-Oriented Architecture to better signal the importance and necessity of incorporating these business focused viewpoints into Enterprise Architecture programs. The advantage of specifying these artifacts is that most “business” professionals are familiar with them. The indispensable skill that remains then is to how best cross reference these viewpoints with the other 3 principal domains which “IT” professional are most familiar with – i.e., data, architecture and technology architectures.
* Based upon viewpoints gathered from LiquidHub experience, HBR article by Kim and Mauborgne, The Balanced Scorecard, Business Model Generation and client input.
Posted in Next Generation Architecture | No Comments »
