[Sigia-l] Re: Large Orgs

Stewart Dean stewart at webslave.dircon.co.uk
Thu Aug 7 18:19:52 EDT 2003


At 10:36 07/08/2003 -0500, you wrote:
>Stewart,
>
>I appreciate you sharing the anecdotes, but I disagree with your
>analysis.
>
>You said:
>"Unfortunately there is a link between the size of the company and the
>quality of their IA. Often what happens is rather than trying to fix the
>IA
>it's easier to develop microsites. Microsites I feel are a symptom of
>bad
>overall IA."
>
>I work for a very large company.  We are very diverse and operate in a
>number of industries.  We have businesses in over 65 countries.  We have
>lots of web sites.  The complexities involved in undertaking an
>"Enterprise IA" are absolutely mind-numbing.

The Dolphin example I gave was 26 countries. The solution to large and 
scary problems like these is to switch to a bottom up approach as the BBC 
are currently using.


>Consider just these few points:
>- Legal and regulatory differences between business units and
>geographies (e.g. The EU has different requirements for privacy and
>financial services have different regulations from food processing)
>- Language and cultural differences.  Not just from a end-user
>(customer) perspective, but from an internal business unit perspective
>and from the perspective of the CMS user (author, editor, etc.)
>- Difficulties of collaboration and consensus as the number of
>"stakeholders" increases.
>- Etc., etc., etc.

I feel your pain but I have dealt with all above and the answer is often 
about allowing the business unit people to maintain their own 
content.  Provide a forum for requires for functionality and maintain a 
common interaction style across all languages with strong guidelines for 
nomenclature. You cannot control content centrally - that's a nightmare. 
What you create is an environment in which people create their own content 
according to local needs but within the bounds of your IA rules. Constrain 
the users too much and use a centralized method of content control and 
publishing and you will have splinter projects.



>The fact is it's not Large company = bad IA, but rather Big company =
>big, nasty puzzle to solve from an IA standpoint.

Very much so. Without polictical support any attempt at IA might as well be 
thrown away. Being able to deploy guidelines is vital in a situation like 
this.

I have done some work on 'whales' and, to be honest, the problem is not one 
of being able to simplify the IA - that is usually not the issue. Spend 
some times doing as is site maps and overall patterns soon become clear - 
common solutions for like business units becomes apparent, as does shared 
functionality.

The problem is, as you point out, one of the whole acceptance, it's the 
political issues. Not strong political support - not go. This is the 
problem with large organisations. I read Louis work and was not totally 
convinced by his approach. It is very traditional top down IA and when he 
starts on low hanging fruit I have to part company with his point of view. 
It's not about tackling everything but setting up an environment where good 
IA can grow, where there are common guidelines if people want to start 
using them.

In other words its about the ability to set up standards and guidelines 
that grow in detail over time and offering support for the mess that people 
with too many good ideas create.

Regards

Stewart Dean






More information about the Sigia-l mailing list