Centralization vs. De-Centralization in Global Web Ops

I have never operated in a multilingual web environment, managing the so-called “localization” of content into multiple languages. At IDG, global publishing was handled on a very de-centralized model, with the flagship brands writing in English and then licensing that content to country-based operations who in turn would pay to have the content translated into their local language, adding in local reporting in that same tongue to build a country-specific superset of the original brand.

Decentralization to gain operational agility is a noble cause, and one I support, but in IT enabled business models it can quickly grow a lot of hair, particularly when corporate messaging and brand management come into the picture. Look at CIO.com and compare it to the Polish version, CXO.pl, and you’ll see what I mean. The Polish operation completely rebranded the domain, creating a variant against the CXO brand, using their site as a portal into other c-level titles.

Having just read IBM Redux, an account of the Gerstner turnaround of IBM in the 1990s, one of the biggest issues that Gerstner and his CFO Jerome York had to confront was the extent to which the company’s “Geos” or geographic businesses, had completely gone off on their own, competing internally and raising havoc with the financial and managerial controls across the company.

Pat McGovern, the founder of IDG, says he adopted a very loose, de-centralized structure after returning from a business trip to find a packed inbox, realizing that he was the bottleneck and that he had to loosen his controls so the business could thrive.

Decentralization was, I think, a necessity in the days before ERP and content delivery networks. The one thing that technology cannot remove is the reality of time zones and the complexity of cross-country meeting and calendar coordination. But time-shifted communications — I’m talking fancy talk for email — and voicemail, has all but obviated the need for a decentralized management model.

If the corporate model for a global enterprise is viewed at three levels — worldwide operations at the headquarters level, geographic which encompasses regions: (EMEA, Asia Pacific, etc.) and then country-level — then the importance of a rational command-and-control structure becomes clear. The trick, for a CEO, is, to borrow the phrase from McKinsey’s Dick Foster in Creative Destruction, to “loosen control without losing control.”

I raise this issue of global governance as I enter the early stages of organizing a network of over 70 sites. While there are obvious economic and operational benefits to a centralized hub model, one predicated on a master corporate database, there is less clarity on how to organize centrally while extending local control and translation down to the country level where the expertise resides. Last week I met with Eli Singer, CEO of Web Collage, and he said the notion that translation must be decentralized is misinformed and that cost savings and managerial control can be achieved in a central hub.

I could always follow my brother’s advice, one echoed by an Englishman I met at Ogilvy & Mather last week, and that is what I call “Texan Translation”: wear a ten-gallon hat and yell English very loudly until people understand you. (The Englishman smiled and said in a loud voice: “I SAID, MAY I HAVE A CUP OF TEA?”) All kidding aside, and abject apologies to the world at large for being yet another American mono-linguist, there is no Web esperanto or precedence for English taking over the world of ecommerce any time, ever, soon. Airplanes, ships .. some industries and professions have standardized on English. Not commerce.

Author: David Churbuck

Cape Codder with an itch to write

0 thoughts on “Centralization vs. De-Centralization in Global Web Ops”

  1. Ugh…regionalized websites. One of the demons that lurk in the closet for every corporate webmaster. At one time, I counted at least 10 Atex regional domains. None in the proper format save the main site, many with incorrect logos, and virtually all completely out of date. Some used older website brand guidelines, while others used designs which I’d never, ever seen (and I, of course, was responsible for brand adherence…). In one memorable case, we struggled for days to find the server so we could kill it.

    It’s not a bad thing to let your regionals have the flexibility to provide what their market needs. The problem comes in when they fire and forget, leaving their efforts to wither on the vine. You end up with a market that you’ve done worse than failed to serve, you’ve abandoned them.

    I’m in the process of moving to a centralized control of the regional sites. Not because it’s better than the other way, but because it’s what the winds of management dictate. Luckily, css properly used makes it a little easier to spin up a regionalized version. But in the long run, I think we’ll find that it’s not serving the needs of the guys on the ground…

    I’d love to hear how you solve it.

This site uses Akismet to reduce spam. Learn how your comment data is processed.