Saturday, November 10, 2007

Death to the Silo

I feel the need to explain myself before I share my next suggestion. Not a day ago, I posted a blog about creating duplication in government to spur competition. Now I'm about to suggest the opposite in the corporate world. Am I becoming one of those pseudo-intellectuals who get along simply simply by taking contrarian viewpoints on everything? Hopefully not. Bear with me ... read on and I promise to explain myself before this blog comes to an end ... so here goes!

The silo business model is endangered. Many of their activities are duplicative and ineffective by virtue of being underresourced. Single corporate-level activities can take some of these things over, reducing costs while increasing effectiveness, standardization, and homogeneity. There is ample room for increased standardization within business lines as well. Ops and functions like Compliance can be more cost-effectively centralized, allowing for greater oversight, staff mobility, better MIS. This "reverse hourglass" leaves only the marketers, RMs, and product specialists divided.

Obvious issue #1 is how various businesses share nicely. Do you split based on backlog of work? On cost per widget? On profitability of work? On growth potential? On executive fiat? On risk?
Answer is yes. Via some heirarchical and adjustable algorithm. What yo do NOT do is let the businesses fight politically for attention. You must assert the independence of the management of these ops and functional groups. You must put the right people in charge of those groups, not discount managers. Then you must empower (and indeed incentivize) them to determine the best allocation rules. To do so is simple: internal pricing. Clearly outline enterprise-level policy to the business lines (for example, "you must have accounting which conforms with corporate norms, GAAP, and is detailed and timely enough to enable seamless integration with the enterprise-level systems and reports."). Then, allow them to bid for the attention of the ops groups .... OR go outside the firm if there are no good options inside. True, this puts a new burden on the organization to determine, communicate, and police the enterprise-level policy. That's money well spent. The idea, however, of insourcing these tasks into their marketing silo (which is largely what has happened to date) puts them under the thumb of managers who are lacking in operational expertice, and moreover, perversely incentivized toward cutting so many corners they end up with round desks.

Obvious issue #2 is how a single group can be both standard enough to find efficiencies ... and yet expert enough at each product/service to handle it without killing the efficiencies.
Answer is that this is exactly what the middle managers are well-paid to figure out. That's not a punt on my part. It's simply a reminder of the "ideal state" of corporate organization. Managers need to assess staff capabilities and match that against work to be done in the most efficient, highest quality, and least risky manner possible. The silo approach forces these groups to work extremely lean. They are unable to take advantage of economies of scale, especially in terms of managerial mindshare. On the other hand, a single manager (or team of) who is able to able to lift his eyes up a level from the daily fire-fights and focus on that balance of efficiency vs. risk vs. quality would have no trouble finding the right balance. This would then dictate the balance between standardization and specificity. These managers can structure their teams however they want. I'd recommend they look into the idea of competing teams, all complying with department-level policy requirements. Someday I'll post a blog on how decentalization and standards play nicely together. For now, let it suffice for me to say that the establishment of high-level standards (or said differently "requirements") is an enabler of decentralization.

At the beginning of this blog, I promised to redeem myself from my apparent inconsistency of view. My juxtaposition is, in fact, intentional ... and that intention is to highlight several necessary underpinnings of both my arguments. Competition, Decentralization, and Specialization. Basic tenets of pure capitalism.

No comments: