Implementing I/A and web strategy when you don't have control of the content

24 Jul 2008 - 5:55pm
5 years ago
2 replies
557 reads
Todd Moy
2007

Hi - I'm working on two similar projects, which are site redesigns of
entire web presences. The common factor between both is that we have
varying ability to alter or change most of the content--or the I/A it
supports. For example, consider redesigning NY Times, but only being
able to affect the main page and some personalization features...but
not the layout, I/A, or presentation of any section like Health,
Finance, etc. Much of this is political, some of it is technical. Some
of the content _is_ unique enough to look and operate differently. But
my question is this... Does anyone have any thoughts or case studies of
a redesign under similar constraints? In particular, I'm interested in:
* Incentives & disincentives to get others to adopt common standards *
Strategies to achieve buy in * Ways to manage expectations about what
is covered in the redesign * Heuristics for determining where to draw
the scope line A lot of this will be, "it depends." I'd really like to
hear "here's what we did, here's what worked, and here's what didn't."
Thanks in advance, Todd

Comments

24 Jul 2008 - 6:23pm
bminihan
2007

I had a very similar political/IA problem a few years ago redesigning
an intranet portal for a very large company (120K employees, 42
countries, 18 languages, and extremely distributed content
development, including 9000 custom "widgets" we had no control
over).

To cut to the chase (it was an 18 month project, more or less), we
got reps from each BU to agree on the depth and scope of the top of
the taxonomy, and commitment to adopt that throughout their
navigation. Everything below the first 3 levels was their own
domain, and it worked very well considering once you got past the 1st
3 layers of navigation, you were in "your own BU space" anyway and
most of the content down there had to be very specific to what you
were looking for.

Some groups had very loose, shallow taxonomies (e.g. HR), but our R&D
dept, for instance, had at least 10 more levels below the top 3, and
about 9 different vectors across which they sliced.

The UI was a little easier, because we already had a common global
stylesheet and color theme everyone had agreed to adopt. On the
other hand, we tended not to strictly enforce design standards in
highly-specific apps, and opted instead to call attention to any
content that appeared to "general employees". That meant about 80%
of the publicly navigable UI looked and behaved the same, while the
BUs had a large number of apps to feel comfortable in and design for
specific expert groups (fewer complaints, and cheaper for them to
outsource, that is =]).

As for your questions:
Incentives? For the UI, we used a ton of industry accessibility
reports and user feedback illustrating how the standard styles and
fonts we chose were more usable and accessible. For the taxonomy,
the big incentive was that if we could optimize everyone's ability
to get down to the deepest levels of the company as fast as possible.
Our central group maintained the motto: Our job is to get people to
your content without having to think about it. That helped everyone
see that we weren't trying to reorganize their BUs, but were just
trying to help everyone work better.

Managing expectations? After convening our BU reps, they did much of
the work for us, but we had a pretty good roadmap following our
initial design and a dozen or so "road shows" to get everyone
comfortable with what we were, and weren't, planning to do.

Determining boundaries? The rule of thumb we used was: If a taxonomy
change generally makes the "global news", it probably belongs in the
global taxonomy. Or better yet, if a changed required very
domain-specific experts to work through a global bureaucracy to
change one of their apps or add a menu item, it was probably too
specific to be in the global taxonomy.

What didn't work?
- Trying to map out what the entire taxonomy might look like,
including every BU and global item in a single layer. No one
understood it and it was obsolute 2 hours before we finished - waste
of time. We wound up showing the top 3 layers and a few sample sites
to illustrate how it might appear.

- Using dummy 'lorem ipsum' text in our mockups. If I had a dime
for every time someone asked us for "the english mockups", I'd be
rich.

- Trying to delegate maintenance of the global taxonomy to specific
groups, or trying to create a group to manage it. Wound up
leveraging existing governance teams who had very little time to
review/approve suggested changes. This also helped clarify how much
throttling the global taxonomy could take, and winnowed out a few
misplaced items early on after the redesign.

If you have any specific questions feel free to contact me offline.
I couldn't resist posting here since your situation sounded exactly
like where I was a few years back. Best of luck, I hope you have
some good contacts throughout the business =]

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=31556

24 Jul 2008 - 7:39pm
Todd Moy
2007

Brian - thanks for the thorough and thoughtful response.

The "we'll go to three" levels of navigation is one that I'm wrestling
with right now. We're using a similar criteria as our first cut, and
then diving deeper or coming up where appropriate. Like anything I
expect, determining where existing content lives in the new IA is
based upon a set of criteria. For me, these include:

* place in current IA
* place in physical domain directory structure
* whether the content seems globally relevant vs. very specific
* who owns it currently
* the nature of its current IA or design

'Course, none of these are absolute qualifiers -- they're odors, if
nothing else. And, there's always the concern that not enough
attention is given to really reorganizing the deep content. But I
guess at some point, life moves on. :)

Cheers,
Todd

On Thu, Jul 24, 2008 at 7:23 PM, Bryan Minihan <bjminihan at nc.rr.com> wrote:
> I had a very similar political/IA problem a few years ago redesigning
> an intranet portal for a very large company (120K employees, 42
> countries, 18 languages, and extremely distributed content
> development, including 9000 custom "widgets" we had no control
> over).
>
> To cut to the chase (it was an 18 month project, more or less), we
> got reps from each BU to agree on the depth and scope of the top of
> the taxonomy, and commitment to adopt that throughout their
> navigation. Everything below the first 3 levels was their own
> domain, and it worked very well considering once you got past the 1st
> 3 layers of navigation, you were in "your own BU space" anyway and
> most of the content down there had to be very specific to what you
> were looking for.
>
> Some groups had very loose, shallow taxonomies (e.g. HR), but our R&D
> dept, for instance, had at least 10 more levels below the top 3, and
> about 9 different vectors across which they sliced.
>
> The UI was a little easier, because we already had a common global
> stylesheet and color theme everyone had agreed to adopt. On the
> other hand, we tended not to strictly enforce design standards in
> highly-specific apps, and opted instead to call attention to any
> content that appeared to "general employees". That meant about 80%
> of the publicly navigable UI looked and behaved the same, while the
> BUs had a large number of apps to feel comfortable in and design for
> specific expert groups (fewer complaints, and cheaper for them to
> outsource, that is =]).
>
> As for your questions:
> Incentives? For the UI, we used a ton of industry accessibility
> reports and user feedback illustrating how the standard styles and
> fonts we chose were more usable and accessible. For the taxonomy,
> the big incentive was that if we could optimize everyone's ability
> to get down to the deepest levels of the company as fast as possible.
> Our central group maintained the motto: Our job is to get people to
> your content without having to think about it. That helped everyone
> see that we weren't trying to reorganize their BUs, but were just
> trying to help everyone work better.
>
> Managing expectations? After convening our BU reps, they did much of
> the work for us, but we had a pretty good roadmap following our
> initial design and a dozen or so "road shows" to get everyone
> comfortable with what we were, and weren't, planning to do.
>
> Determining boundaries? The rule of thumb we used was: If a taxonomy
> change generally makes the "global news", it probably belongs in the
> global taxonomy. Or better yet, if a changed required very
> domain-specific experts to work through a global bureaucracy to
> change one of their apps or add a menu item, it was probably too
> specific to be in the global taxonomy.
>
> What didn't work?
> - Trying to map out what the entire taxonomy might look like,
> including every BU and global item in a single layer. No one
> understood it and it was obsolute 2 hours before we finished - waste
> of time. We wound up showing the top 3 layers and a few sample sites
> to illustrate how it might appear.
>
> - Using dummy 'lorem ipsum' text in our mockups. If I had a dime
> for every time someone asked us for "the english mockups", I'd be
> rich.
>
> - Trying to delegate maintenance of the global taxonomy to specific
> groups, or trying to create a group to manage it. Wound up
> leveraging existing governance teams who had very little time to
> review/approve suggested changes. This also helped clarify how much
> throttling the global taxonomy could take, and winnowed out a few
> misplaced items early on after the redesign.
>
> If you have any specific questions feel free to contact me offline.
> I couldn't resist posting here since your situation sounded exactly
> like where I was a few years back. Best of luck, I hope you have
> some good contacts throughout the business =]
>
>
> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> Posted from the new ixda.org
> http://www.ixda.org/discuss?post=31556
>
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

--
____________________________
http://www.oombrella.com
oombrella /a/ gmail.com

Syndicate content Get the feed