Saturday, January 28, 2006
Diseconomies of Scale and Edge Competencies
Two articles I saw recently started a chain of thought.
Seeing Fakes, Angry Traders Confront EBay - New York Times
How Click Fraud Could Swallow the Internet
(Hat Tip: Paul Kedrosky)
First thought that came to my mind was that just as network economics has allowed suppliers to scale their business, it has also made them exposed to massive leverage of their own network by their users. More often than not, this leverage is a good thing but if (and when) a vital weakness in the business model is exposed, it can quickly snowball into a major strategic threat. This is another evidence of Umair's key thesis that the boundary between the firm and external world is being obliterated : the world is no longer neatly divided into you the producer and joe the consumer. Now think about what happens when you are learning to live with 100 million Joe Consumers every day. This is when a lightbulb went off : if you can leverage your network, so can your customer - and for whatever purpose they want - because, now they are part of you! Yes, the camel is in the tent and living with you : deal with it. This also illustrates how the firm will need to develop very different competencies in future to work with networked business models - learning to live with your customers every day and working with them is very different from seeing them as passive buyers of your stuff some of whom may be criminal. However, I don't see any realization among web businesses that their businesses have, perforce, internalized their customers. They still think of them in old terms, as in "a few bad apples and fraudsters who need to be policed" (or not, in the case of EBay). Here's the new way to think about it: remember the old adage from David Ogilvy, talked about in every Marketing 101 class at any self-respecting b-school? "The customer is not an idiot : she is your wife". Well, the new adage for the networked business today would be "The user is not the scammer : she is your customer".
The other thought that followed from this is, well, how do we solve such problems? Start thinking of this as a strategic judo, where first you leveraged the network to build your business and now the network is leveraging you to make some money for itself. The natural strategy that follows from this is to once again leverage the network. How? Yes, you guessed it : by developing edge competencies. This is again where Umair's thinking is very useful : once you realize that the boundaries between the firm and the market have evaporated, you realize that all action has shifted to the edge and that's where you should be playing.
Let me elaborate: EBay has a massive strategic weakness due to its exposure to fraud. Where is this happening? At the edge - fraudulent users are also traders and consumers of EBay's service. What mechanism does EBay have to deal with it now? Ratings and user feedbacks, which can be gamed easily. What else is going on? Observe that there's an emerging band of users such as Ms Rogers in the NYT article that are self-organizing to tackle this problem. This is the inflection point : that's a clue right there for EBay to develop a competency at the edge and leverage such users to solve the problem. In other words, EBay must deploy resources to help such users self-organize, open its corporate gates so to speak and give all help to such users. The community will find a way to solve this problem, if only EBay would let them. . All EBay needs to do is let them work with EBay's engineers and managers to come up with tech and non-tech solutions to this problem in partnership with the community. If they could do it, that'd be a killer edge competency.
What are the chances of such enlightened strategic thinking at EBay? Zilch, of course. The lawyers will spread fear, uncertaintly and doubt about legal liability for EBay if it tried helping its users. The preferred strategy, driven by lawyers and beancounters, would be to 1.pretend that there is no problem 2.litigate this all the way 3.cash in their stocks and get out. The old MBA types, if they were really trying to be good strategists, would talke about developing a "core competence" in battling fraud. The CTO would be given a $100M budget every year to develop internal solutions for this problem and his MBA buddy will talk about how EBay can "differentiate" itslef against competition by having the best anti-fraud solutions in the universe. The firm would work with the cops and the CEO would lobby the congress to have scammers jailed. Of course, if you've followed Umair's work up to this point, you can see that none of this will work.
The other thing that I find interesting is that edge competencies seem to be of two kinds : those involving humans vs. those involving algorithms. While the EBay solution outlined above is human-centric, I think Google will repose its faith in algorithms and only invest in tech solutions to tackle click fraud. Will it work? Remains to be seen, but for now, humans seem to be winning. Personally, I'd always bet on the ingenuity of the human to outmanouver any computer or algorithm. Why? Because humans can coordinate and cooperate many orders of magnitude more effectively and efficiently than algorithms. As I learnt in my CS grad school, while we have done pretty well with sequential uniprocess algorithms, the state of the art in distributed algorithms is not really that far out. The profound significance of that understanding is only beginning to dawn upon me now after my groping in the dark attempts to try relating that with network econ and strategy.
I totally agree... I have always thought eBay should acquire a company like escrow.com
and offer this service for free. While this would increase costs it would virtually eliminate fraud...Jason
Great article, a few thoughts:
Your solution is an excellent one BUT you seem to miss the fact that it's the ONLY solution.
The alternate "old school" solution (building internal anti-fraud competencies) that you discuss isn't actually an option for eBay; remember that eBay is a Marketplace, not a retailer. Most importantly, eBay is not (legally or otherwise) the "seller of record". That's why the company can't (internally) do anything about it eradicating P2P fraud other than report on it.
If eBay's model was to receive and consign inventory, then you would have been on the money. It's an interesting point that even within peer-production models, there are many many flavors of strategy.
A good example of a commons-based marketplace that is the seller of record and that has built an extensive internal QA system is iStockphoto, who I've long thought should be following your advice of having the community rubber-stamp the product instead of doing it internally.
Keep it coming ;-)
I might add one bit to your analysis. A key issue here is brand dynamics in network economies. The old top-down brand models, in which the company �owned� the brand and dictated it to �consumers,� are no longer valid. We now have a case where eBay users are fighting for the integrity of the eBay brand, and eBay is saying it has no role in the effort. Effectively, brands are built at the edge, and will be increasingly so going forward. eBay would seem to have a major decision to make in the near future. Either it joins with motivated users to make the eBay brand one of commercial trust, or it cedes brand integrity altogether in the guise of a �neutral� marketplace.
You might also call this a case of brand asymmetry. Ebay favors an ultra-narrow, limited and legalistic notion of �marketplace brand,� whereas eBay customers want an eBay brand based on community values (truth, accountability, fairness, etc.)
Brand asymmetries portend brand changes. One way or the other, something�s gotta give.