A new meta-search shopbot

Today (June 19, 2007), a new meta-search shopbot (up-stream shopbot) emerged. Discountmore.com could simultaneously search all major shopbots. The rational for a meta-search shopbot was: "...the only way to find out who really has the lowest price for individual products." according to Founder-Creator and CEO Bobby Kalili.

As in a paper I have written a couple of years before, the emergence of up-stream meta-search shopbot was inevitable. However, the real issue is more complicated than the pitch in the announcement. For example, is meta-search shopbot really the "last piece to the comparison shopping puzzle" as indicated by Kalili? Certainly not the last one.

For example, choice overload is the first problem a meta-search has to handle. How to balance the needs and present the same number of choices aggregated from different shopbot? This is more complex than individual shopbot handling choices from vendors. Now the choices come not only from vendors but also from auctions, and whatever courses one could get the product.

Another issue is still the coverage. Shopbots covered more or less the same set of online vendors via data-feeding by vendor themselves. A meta-shopbot only expand that set a little bit. There are still a lot of vendors (those do not involve in comparison-shopping at all) not being covered by shopbots.

The last one is the challenge about how to incorporate factors like coupons and promotions that are not covered by shopbots. They make big differences on price. however, unless we find a way to make the information structured, we won't be able to benefit from this in a large scale.

But still, this is a good move and nice search engine. I enjoy it.

Growth, Division of labor, and BPR

Almost finished "Direct from Dell," which illustrated the first 15 years of Dell, the Texas based computer manufacturer. The business acumen of the founder, Michael Dell and the timing (aseembling PC was a high margin emerging industry in early 80s) seems to be the top 2 critical factors leading to the growth and success of the company.

Another issue impressed me is how Dell handle the growth of the business. When a business growth from millions to billions and then to tens of billions within a short period of time, the company has to constantly re-structure to meet the growth need. The major theme seems to be the division of functions. For Dell, it first realized the importance of P&L (Profit and Loss Statement), and then how to apply it to different product categories. Then they found a manager who used to manage one product category would soon be outgrown by it. Thus, what they did is whenver the product category became too large, they will separate it into two or more product categories and have additional managers for each new category (it somehow reminds me of how eBay handle product categories). For each product category, P&L would be applied and subsequently, strategic decisions were made based on data collected.

This is very similar to most enteprirse did in 50s and 60s in the US. Presciely because both Dell and those companies in 50s and 60s faced the same seller's market. PC is a new and hot product, the demand will exceed the supply so this method works.

Anyway, though BPR became popular in 90s, Dell seemed not being affected by it, which makes sense for a company that focus on growth.

Business Process Reengineering revisited

I am reading Hammer & Champy's "Reengineering the Corporation" these days and found more historical contexts about BPR. My first contact with BPR was during my doctoral study from those academic papers. The missing part is why BPR was needed or why corporations were not optimizing their business process from the very beginning. It turns out they did but later they failed to adapt the process to changing environment.

Hammer and Champy's seminar HBR article ("Reengineering Work: Don't automate, Obliterate," which led the latter book) revealed this a little bit via a fictional example:

"... Many of our procedures were not designed at all, they just happened. The company founder one day recognized that he didn't have time to handle a chore, so he delegated it to Smith, smith improvised... business grew, Smith hired his entire clan to help him cope with the work volume. They all improvised...The hodge-podge of special cases and quick fixes was passed from one generation of workers to the next...institutionalized teh ad hoc and enshrined the temporary...Why do we send foreign accounts to the corner desk? Because 20 years ago, Mary spoke French and Mary had the corner desk. Today Mary is long gone, and we no longer do business in France, but we still send foreign accounts to the corner desk..."

Later in their book, the historical context was explained in more details. According to their view, the existing business processes in American corporations in 80s were originated from 50s and 60s, a time of mass production. At that time, the market was so good that customers will buy whatever produced thus the main goal for a corporation was producing as many products as possible to grab the market share. Thus a division of labor, hierarchical style corporation management structure was scalable and preferred (need 1000 products? Add 100 more workers, 5 more managers, and 2 more human resources staff).

However, when market power shifts from producer to consumer and the choice of products are increasingly abundant, which is the situation in 80s, such style and structure was no longer effective. Even from departmental perspective, the highest efficiency could be achieved, the overall efficiency of the corporation can not be guaranteed because there was no alignment between the departmental goal and corporation goal. A good example cited in the book is the purchasing department. A company may spend $100 internal cost to buy a $3 battery! Also, in terms of customer experience, it may took six days for IBM credit to finance a customer purchase though the actual process took only 90 minutes! Most of the time was wasted in the transition of documents among departments.

The reengineering thus called the efforts to focus on business process intead of steps in the process because if the process is wrong, even each step of the process is efficient, the whole would still be inefficient. By reengineering the process, critical elements like customers and suppliers could be brought into the focus and being organically linked with business processes within the enterprise. This is in contrast with the previous view that the internal process of an enterprise is separated from customers and suppliers for mere efficiency goal.

The negative influence of division of labor on current corporation behavior was especially illustrated in the book and argued by the authors to be the main cause of current problems of American corporations.