The 11th Principle of Good Design

Software is tricky, the needs of the users can change over time, and the developers understanding of the project will most likely change over time as well. If not for any other reason than keeping up with competing technologies, or the introduction of new technologies.

So I would agree, iterating your application's design is very important. But ideally, if it was designed in strict adherence to the 10 principles of design, the essence of the application would be long-lasting. This would mean any needed changes would be intuitive.

First I take a bit of umbrage at the term the "agency" mindset because it seems to indicate its our fault. Its more of the "agency relationship* mindset.

Unfortunately despite that its a bit true. We generally are in the business of getting things done and out the door so we can go onto the next project. Its very much the nature of the business. But a lot of this is also driven by our clients. The best clients realize this is a phased a approach and build in separate phases to accomodate user testing A/B testing and realize that the work isn't done after the work is dropped. Its great when we get clients such as these, because we get to really focus on the product. Unfortunately they are few and far between. Most are much more budget focused and when we bring things up like A/B testing and phased approaches, they get excited, when we show them the costs for this kind of work, most of the time these kinds of "extras" are what gets dropped.

Now part of this problem also comes from larger agencies. There's some folks out there who won't do a project thats not $500k or larger. Trying to tack on an additional $100k of improvements post delivery isn't something they're interested in doing. This isn't just hubris, if you're running a shop of 100 folks a "small" 100k project presents problems, it utilizes resources better used elsewhere, and you'd be running your sales and account folks ragged managing 4 or 5 of these small projects (One account manager on one 500k project is a lot better then one on five 100k projects).

This tends to get worse on the dev side, for dev we do the best we can to get a nice clean codebase out the door, but frankly once it "works" its done. Unlike design which actually does have some iterations and polish built into the design process dev is more "does it work to spec" and done.

A solid agency will help to identify any caveats and get the design to a point where it doesn't have to be iterated on.

I'm not sure this is true, the better the creatives working on a project the closer you'll get to it, but its always a tradeoff against budget and time.

It simply makes it the final iteration in that development branch.

Edit: and it has in fact been improved in several regards, mostly materials: http://www.nytimes.com/2012/03/08/garden/dieter-ramss-606-sh...

I'll add another quote: "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." - Antoine de Saint Exupéry.

Eventually you get there. Or close to it.

Too: Rams worked in physical media. Eventually you've got to ship, and in his case, he couldn't push updates twice a day. This happens even in software and web design. The "bones" of Linux were layed down in AT&T's labs over 40 years ago. For mainframe computing, history starts over 50 years back.

Even today's major websites have their own mass and inertia in the form of their code base -- you can push changes every 12 hours, if you like but things have to work together, and either modularity or fragility will impose limits on what you can reasonably change and expect to have functioning, stable code.

For us humans we're not as intuitive with nature. We have many distractions and conflicting goals. It takes us more, and longer, iterations to figure out these things. We're just not lazy-enough.