My column at Daily News and Analysis (Dec. 27, 2007)

Here is the latest column at DNA.
AxG

Lessons from HP’s “Always-On” Intelligence system
Alex Gofman

Charles Darwin inherited his love for experimentation from his genes. His grandfather, Erasmus Darwin, believed that “A fool ... is a man who never tried an experiment in his life.” It did not matter if the experiment was expected to have a negative outcome or if everybody believed it was crazy and called them ‘fool’s experiments’. In fact, the creator of the evolution theory remarkably acknowledged: “I love fools' experiments; I am always making them”.
Although his views on the history of living species cause some controversy, the implicit applications of genetic approach to New Product Development (NDP) is quite interesting.

The legendary success of Tom Kelley, General Manager of IDEO, is in part based on his belief in the “cheaper, faster, simpler approach.” If you have the luxury to spend months and months on research and in-depth observation of your customers and at the end come up with a “perfect” product, you are lucky, but you occupy a rather unusual position in today’s market. Your competition might not be willing to wait that long and might grab the market share before you. Remember the run­away success of Microsoft Windows 3.0 (followed by 3.1) released shortly before the competing IBM’s product? The latter system, OS/2 2.0, was superior to Gate’s cre­ation in many aspects, but it sadly failed. One can argue that there were many reasons for this failure. But most agree that in that case as well as in many others one formidable aspect of the competition - the timing – played an indubitable role in the success and failure. Particularly, in the high-tech industry, where new products frequently become outdated before they are released.

Just a few years ago our average modern day mobile phone with a camera, an MP3 player, a personal organizer, etc. would sound like an impossible proposition. Forget about a built-in TV with live and stored programs long enough to drain the batteries long before they are over.

What some ingenious designers manage to coalesce into mundane gadgets is astounding. It does not have to be new features – just a recombination of what is known. Sometimes the product hits the ‘button’ and creates a runaway success. But in a majority of cases, it finishes collecting dust at a discount store or goes into a recycling bin. How can designers and marketers find the right combination of the features? In nature, the never-ending recombination of genes, through cross-breeding and evolution, helps species to survive. A quite popular random experimentation in NPD is much faster than nature’s process but still is inefficient and slow by our modern measures.

Critics might say that this notion of innovation by combination by itself is simply too mechanical, too utilitarian, and, therefore, is certainly void of the charisma of creativity. Others disagree. Michael Vance, a well-known American creativity expert, lecturer and Dean of Disney University, once said, “Innovation is the cre­ation of the new or the rearranging of the old in a new way.”

If you had a chance to read the book Selling Blue Elephants, you might know about the “Always-on intelligence system” established at Hewlett Packard a few years ago. Applicable to virtually any part of the process, the system based on Rule Developing Experimentation (or RDE - see my previous columns) brought the consumer to the table in every design initiative or marketing decision in a way and scale that was unprecedented for HP. RDE changed the way the company thought about answering the problem of “What shall we put into this prod­uct to make consumers want to buy it?”

Unlike the data from most ad hoc research projects, which varies in structure and topic, HP used RDE’s dis­cipline to uncover the broader “meta patterns”—patterns that reveal the bigger pictures, across products, across categories, across countries, and over time.
The accumulating library of RDE studies opened a new, virtually effort-free opportunity for the consumer insight team to integrate data across diverse knowledge-development tasks. It became clear that across its many product lines, HP attracted two radically different seg­ments of consumers, with drastically varied mind-sets:
· Segment 1—Technologically savvy individuals who mix and match separate components, and who enjoy and occasionally even revel in the challenge of getting them to work together.
· Segment 2—Individuals who prefer a complete package with all the accessories that work straight out of the box.

This knowledge helped HP to focus and target its ongoing design and marketing efforts, making them more efficient and, as time would prove, far more profitable. RDE provided the specific numbers—what ideas compelled and just how compelling the ideas could become when properly framed. The latest RDE tools also estimate the synergies between the elements – what works well together and what does not.
In turn, the process of data-basing the ideas’ performances could reach the critical mass with a profound impact on effectiveness in the generation of new ideas. Here is another example from a major high-tech company where five years ago, more than four out of five marketing ideas were ineffective (only 18% of about 1800 tested ideas had a positive consumer influence). In 2006, the “institutional learning” through disciplined deployment of RDE has more than doubled the number of ideas with positive consumer influence to about 38%!

Evolution is based on nature’s ‘fool’s experiments’. Good combinations of genes thrive, bad ones do not. NDP does not have to be that painfully slow and cruel. Ideas that come out of the genomics recombination of elements from different products and that perform well in these types of studies tend to do well in subsequent tests and in the market itself. The reason is pretty simple. Unlike “beauty contests” whose goal is to pick one winner from a limited set of contestants subjectively pre-selected according to a ‘heavy-weight’ HiPPO (Highest Paid Person's Opinion), RDE is more like a ‘torture test’ - with the mixing and matching and the rapid-fire presentation of test concepts to the consumers. Any element that does well in this type of survey stands out against many thousands of combinations in which it appears. Betting on that element is like betting on a horse with a great track record in many races, climates, on many different tracks, with many jock­eys. The odds are that winning elements, like winning horses, have some­thing good going on that’s worth incorporating into a product. Good genes do increase the chances of survival, don’t they?

_______________________________________
Alex Gofman is VP of Moskowitz Jacobs Inc., a NY based company, and a co-author of the book Selling Blue Elephants: How to Make Great Products That People Want Before They Even Know They Want Them (www.SellingBlueElephants.com) written with Dr. Moskowitz and recently republished in India (it is also currently translated in twelve countries). He may be contacted at alexgofman@sellingblueelephants.com.