Bryan Jalowitz

Consider that adaptation is not always the best evolutionary strategy. Usually, its better to keep some sort of a backbone. Embrace change, but then freeze for a while and let your users settle in with a consistent IA. Just as you shouldn't move your furniture around every week just to please every guest, you shouldn't be changing your IA so much that it becomes a distraction for your users. Sure things change and you need to be agile, but uncertainty in architecture is not good. If you have an uncertain IA, its because its not based on deep enough human traits that really don't change that often. Instead you are most likely chasing trends, and trends come and go, and so shall your users.

Consider that all websites need to be both visually appealing and user friendly, so this is just a baseline not a high-bar request. Additionally, take a good look at the messaging, strategy, and content. If the content is low or even mid-grade, and there is no strategy to engage and make it visible to the masses, it really doesn't matter how good the UX or visual appeal is, your not helping the brand that's launching it one bit. A lot is said nowadays about UX, and its about time. But lets not loose site that, mostly still, websites are fundamentally images and text. Interactivity and visual jazz can slow the dissemination of this simple information. A website is a container, not an entertainer, in an of itself.

In additional to all of the safeguards mentioned in staging and availability, aeparate your business logic from CMS apps, and make sure whatever system you are using has reliable caching. Make sure you can shut the db off and your static pages are still served quickly. Always develop a comprehensive caching strategy that protects production from CMS integration and db issues. Keep your CMS apps focused on content management and delivery and caching.

Stop researching behaviors and start talking to people. Then get them to sign off on a design before and after you release it. Unless you have fresh data from real people its all speculation to a point.

User flows are but one exercise in information architecture. I apologize for the brevity of my response, but we are talking about architecture, so hierarchy of ideas and tools is important.

Why is any leadership important? Because experts get entrenched in details, can loose objectivity, and miss the big picture afforded to leadership. Without leadership, things still happen, so it can be confusing to some if it is indeed important, especially with uncaring leadership experiences abounding with huge and lasting impact. With good objective and empathic leadership, experts can focus on the details and work in a state of free flow, whereas without it, there is always something that comes up to prevent it. Without clear leadership in at least one of its many forms (expert, borrowed, or direct etc.), its tough to make clear decisions that stick and personal objectives tend to come into play more often than we care to admit.