Tuesday, October 13, 2009

Ambient Awareness meets Business Intelligence

Clive Thompson in a 2008 NY Times Magazine article reported on how "ambient awareness" is enriching our social lives as we have instant access to more and more real time information about our friends and colleagues I'm so Digitally Close to You). The rapid increase in the number of information sources, the speed new information is being generated, and the quantity of information available is equally impacting business processes from product development to marketing to customer service. In my consulting work this is one of the key technology shifts that I see companies struggling with as they re-examine how data is gathered to support decision making. Redesigning business processes to incorporate business intelligence from ambient sources will be a key part of redesigning companies over the next decade.

(This blog post is an excerpt from a long white paper that I am currently writing which I will post in its entirety when complete)

The term ambient intelligence (or AmI) was, according to Wikipedia, first used by Brian Epstein of Palo Alto Ventures in 1998 in a workshop for Philips on the future of consumer electronics. The world they and others began to describe had a set of key technology trends that together created a fundamental shift in the way we interact with the physical world including miniaturization, wireless communications, software platforms for distributed systems, improved human-computer interfaces, the general robustness of autonomous systems and a continuing reduction in cost for the deployment and maintenance of such systems

These researchers envisioned a world that by 2020 thoroughly connected people with their environments as sensors, transmitters and other devices became increasingly inexpensive to deploy, easier to program, and more connected. Ambient information systems can be generalized as following this common pattern: (1) the translation of the inherent information in our environment, such as the speed of passing traffic, into digital information via a sensor; (2) the transmission of this data through a computer network; (3) use and presentation of this data by a human or machine process (for example, traffic statistics super-imposed over an online map). I'll use the term "lens" to denote any system which is aggregating, analyzing, and presenting this data.

A simple example of this can be seen in the automated toll systems now in use in many western countries. A small transmitter placed in an automobile uniquely identifies the vehicle to a toll sensor, allowing the driver to be automatically charged a use fee as he drives past some fixed point on the roadway. In this case the lens is a machine process designed to associate the location of a specific vehicle with a financial transaction to be processed against a specific driver's account.

Augmented reality systems currently in their first consumer deployments through mobile phones also offer a glimpse at the human side of this coming world of ubiquitous information-rich interactions. The combination of a set of sensors including a digital camera, GPS, and a compass into a portable device with Internet connectivity allows information about an individual's environment to be retrieved as he moves from one place to another. Here the lens is a visual human-computer interface made possible through the video display of the digital device.

In just the same way that the physical world can be instrumented, detected, and thus better understood we can also instrument the virtual environments in which we are now increasingly communicating and conducting business. As Thompson points out, the innovation of Facebook's "news feed" in 2006 was not in the creation of new information but in the way that information was surfaced to Facebook users.

Facebook had already created a system in which inherent information about people's activities ("Tim and Lisa broke up") was being captured through the human sensor network of its users. The news feed suddenly provided a lens through which one could consume all of this information easily, providing users with a tool for comprehending larger quantities of data and presumable making decisions (I guess I should call Tim or Lisa...) and, as Thompson reports, startling and upsetting people who hadn't thought through the implications of putting this information online.

In the same way the inherent information in our business environments is increasingly being collected and stored online. Past Amazon CTO Andreas Weigend enjoys pointing out to clients that "more information will be created and stored this year than in every prior year in human history." Businesses must implement the right sensors for collecting and transmitting and the right lenses for aggregating, analyzing, and presenting this information.

For example, for a B2B client we recently added online social profiles to the set of information that sales people have about prospects as they try to follow up on initial product inquiries. Having more information available about that particular individual measurably increases the likelihood that the sales person can reach a prospect and have a meaningful conversation. But almost as important is how this information is informing the process of deciding who to contact in the first place.

Sophisticated sales organizations have long implemented "scoring" mechanisms for trying to decide who their most interesting potential prospects are within a given list. A weakness in these systems is that much of the information used for such scores is self-reported by the prospects (size of company, title, industry group, etc). Thus the ambient intelligence about these prospects -- the data they are creating all of the time as the use various online services -- can be significantly more useful in assessing the relative value of one prospect versus another.

Another example, marketing organizations are increasingly aware of the vast number of customers talking about their companies and products. Communications teams are establishing "listening posts" (sensor networks) to aggregate this information. Too often this information stops at an evaluation of "influencers" who can then be targeted for media campaigns. We have helped organizations recognize that key product insights, support issues, and other business processes can be informed by the collection of this ambient intelligence from the marketplace.

These are a set of ideas which we are only just beginning to understand about how business will change over the coming decade. When researchers began to define ambient intelligence a decade ago, they envisioned "...a world in 2020 where user-friendly devices support ubiquitous information, communication and entertainment." We can now see that the same technology trends impacting consumer products will also radically transform business processes and decision making. The most advanced companies have already begun using sensors to collect relevant information from their environments and are developing lenses to use this information in their activities. The development of these systems will be critical to competitiveness in the 21st century.