The trigger for this major Brussels Airlines redesign project was pulled after we did a competitor analysis (#1: Delta) and proposed a first alternative concept.

There were 2 main issues to resolve:

1) Stress BAir's USPs more clearly; this mainly meant bringing forward more prominently BAir's wide/unique range of destinations, particularly in Africa — historically.

2) Rearrange the content according to the different needs of the different audience types (example: those looking to buy a ticket vs. those who already have a ticket).

The trick was also to communicate the USPs (1) nearly implicitly via the way the content was structured (2) and presented through the main paths of navigation.

To help distinguish the different audience types and their associated content, a color coding was used in the initial working documents: red for the commercial angle, dark gray for all practical – post-sales – info, and light gray for any other tertiary corporate content.

Once approved, we used the conceptual sitemap also to indicate which pages would be most important to start wireframing in more detail. Not surprisingly, these matched the more 'weighty' elements on the conceptual sitemap.

First however, we also created high-level page zonings for all those page templates, using the color coding yet again. This way, we illustrated not only which content would fit where in the overall structure, but also where on a page level. Later, in the visual design too, the practical info section was intended to have its own distinct color temperament.

Along the way, a fully detailed sitemap was, of course, also assembled.

For now, some technical issues seem to be holding up the launch of this new website, testing my patience too, because I'm dying to see the first real-life stats to learn what works, and what doesn't :-)

The OTC (or "Over-The-Counter") eTrader tool allows financial traders (mainly in London and Paris) to buy/sell certain financial products.

The 2 main product classes involved were interest "Rates" and "FX Options", which are not traded on a typical exchange, like Nasdaq, but between the parties directly; hence OTC.

Clearly, the end-user of this type of tool is highly knowledgeable about its subject and she will likely use it dozens of times per day. The UI contains plenty of abbreviations and jargon, like "Notional denomination", "Coupon floor", "Fixed" vs. "Floating", "Walk away", "30/360 Basis", "EURIBOR 3 Month", "Live spot" vs. "Spot reference", "Strike", etc.

Of course, jargon in itself is not necessarily a usability problem; that depends on who the audience is. But it can make things trickier for a designer. You must understand what you're designing for.

My formal education (I hold a degree in Economics) was helpful, a bit – I did not specialize in Finance though, but Business IT :-) And it's true that, starting from the existing UI, you can make some recommendations to iron out generic usability issues. But still, you're quickly confronted with your limitations.

It's also often about flows. For instance, when requesting a quote, you not only needed to know what the constraints are on the input fields (error handling) but also that a successful quote would expire after x amount of time.

That's why, for this project, I spent a few weeks in London to work on site.

I was seated, literally, next to the traders/end-users. This allowed me to experience the actual context-of-use, and the easy access allowed for redesign alternatives to be proposed and then rejected/refined in rapid iterations.

Some of the main changes we made together:

• Adjusting the layout width to avoid horizontal scrolling
• Rearranging the global nav: grouping & separating
• Introducing the color coding ( + , already used offline) to – at quick glance – recognize the asset class involved
• Rearranging the input field alignment
• Reducing the prominence of input field labels
• Introducing a ★ Favorite quick access feature
• Improving error handling
• Adding a ✔ Check indicator as positive feedback to see – at quick glance – which default values were manipulated (only rarely will all be changed)
• Various other little clean-ups

Some of these changes may seem unorthodox, like making the labels less prominent, but are a result of the high frequency-of-use, where the expert-users will (almost like muscle memory) quickly know by heart what is where. The default values guide. The labels become merely a means to verify if need be.

Like I said, this was an exceptional project because of its focus and the close collaboration with the end-users. And it was great fun working in London!