BUMPY TRAIL-BLAZING
This story on CNET via Techmeme tells a good tale of a company that is a leader in it's industry, takes a big leap of faith on making old technology work in new areas, and running into some head winds in making it all happen.
The CNET story has to do with Google's efforts to bring it's vaunted automated ad system to the local radio spot advertising market. As the piece notes:
"The departure last week of two top Google radio ad honchos had observers wondering whether the search giant could conquer the old-fashioned $20 billion radio industry with its new-fashioned technology..."
"Applying a national automated approach to a business culture that traditionally relied on personal relationships and local content has proven tricky for Google."
This past week there were two other similar, but unrelated stories about leading companies running into near-term difficulties applying old tech to new opportunities.
The second instance that caught my attention was this story on Walmart's difficulties in getting it's supplier network fully on board with rolling out RFID (Radio Frequency Identification) for it's logisitics management system. Here's a clip from RFID Update:
"As the top story in the WSJ's Marketplace section, Wal-Mart's Radio-Tracked Inventory Hits Static (registration required) focuses on Wal-Mart's initiative and takes a generally gloomy view of RFID adoption.
The article points to a number of negative indicators, including the fact that Wal-Mart's RFID footprint is far smaller now than originally hoped (5 distribution centers have RFID today, while the company was aiming to have 12 up and running by January of last year), persistent high costs and consequent supplier complaints of elusive ROI, and privacy concerns."
Last one up is this LA Times story on difficulties being experienced by Kaiser Permanente, the country's largest not-for-profit managed care provider, in rolling out an ambitious electronic health records system. Again, an excerpt:
"Kaiser Permanente's $4-billion effort to computerize the medical records of its 8.6 million members has encountered repeated technical problems, leading to potentially dangerous incidents such as patients listed in the wrong beds, according to Kaiser documents and current and former employees..."
"Kaiser's effort, one of the largest and most ambitious electronic medical records projects in the country, is seen as a possible national model. With evidence suggesting that digitized recordkeeping can lower health costs and save lives, President Bush is pushing for every American to have an electronic medical record by 2014.
But the glitches illustrate the difficulties a massive healthcare provider might encounter trying to implement a complex computerized system."
The one thing each of the stories had in common, was the no-brainer reaction most of us techies had when the basic proposition of the initial roll-out was first made evident. Of course Google should be able to bring it's paid-search success to the local radio advertising market. Of course, RFID can tremendously improve the inventory logistics of a massive retailer like Walmart. Of course, automating patient records ought to be a win win for all parties involved.
But what all three unrelated stories of tech woes illustrate, is that one cannot under-estimate the difficulties of getting the cultural, political and business model drivers for ALL parties aligned JUST RIGHT, to make things really hum.
It's all so much harder than it looks. And is likely going to take much longer than initially expected.
But early failure doesn't mean that the longer-term goal is not attainable. It's just going to take a couple of steps back to make one or more forward. For a little bit longer.
NLP efforts as purported by Powerset and the revolution of the Semantic Web both fall into this category as well. Fernando Pereira has a good blog post on the NLP issues, having been in research in this area for a long time, expressing the idea that change and development here will be incremental and evolutionary, rather than revolutionary and a magic bullet release so to speak (http://earningmyturns.blogspot.com/2007/02/back-to-future-nlp-search-google-and.html).
Seeing some of the challenges still left to resolved w/the Semantic Web, all the hype about Web 3.0/4.0, and big news articles in the NY Times, all misconstrue these developments as revolutionary rather than evolutionary, as I believe they will occur.
Posted by: P-Air | Saturday, February 17, 2007 at 12:39 PM