Karen McGrane has become one of my favorite UX pros over the last few months, all through UIE’s podcasts. Her closing talk at the IA summit is one of the most inspiring UX-related talks I’ve heard. After listening to that I went back and listened to all the other interviews with her over the last couple years. And bought her book Content Strategy for Mobile.

This morning I listened to her latest UIE interview, with Jared Spool. Here were a couple big takeaways for me:

  • Responsive does not fix deeper problems with your content and workflow
  • Moving to a mobile is a great opportunity to address fundamental UX problems on your site that you should have fixed a long time ago (that’s something I’ve often said about starting an SEO initiative)

But my favorite part was Karen’s response to Jared’s well-articulated question about doing a site redesign all at once vs. breaking it into pieces — something we often wrestle with.

Here’s Jared’s question:

For years, we’ve been telling our clients that you don’t want to just redesign your entire site all at once because that just opens up too many problems and too many holes. It doesn’t really give you a chance to learn anything.

You want to do it in small pieces so you can design the process of learning into your development process. Of course, things like lean UX now are saying basically the same thing, which is you make small changes and you do it very fast. You get to learn all the way through.

One of the push backs that I always get on this, is infrastructure changes like the CMS, like the underlying content, like the underlying servers, it’s really inefficient to do that in small pieces, that you really want to do the whole site at once and just get it done.

Let’s say I’ve bought into this idea that I do need to create a single content management system is not going to work much better than the one I have. It’s going to be a better publishing. It’s going to help me when I’m dealing with smaller screens, and larger screens and slower bandwidth, and higher bandwidth.

I’m going to have all that set. Do I have to do it all at once? Or, can I break it into little pieces and learn as I go along?

And here’s a summary what Karen recommends for this situation:

  • “Let’s build a new responsive site that, for the time being, is only going to serve smart phone and tablet [devices].”
  • “Once we get that responsive site to a place where we’re happy with it, then it will grow up and overtake the desktop site.”
  • This allows you to “go back to the organization with some actual numbers and demonstrated success on this new site” before you touch the “desktop mother-ship” with all of the competing stakeholders
  • Start with a lightweight version of the CMS you’re hoping to move to.
  • Start just with a section of the site that has a relatively small number of stakeholders, and is important but not critical to the business. Something like investor relations or white papers. (This was more Jared’s point than Karen’s.)
  • Once you get the process down there, move to the other sections of the site.
  • Once you show success with the mobile and tablet versions, integrate desktop.

As Jared points out, this approach gives you the “ability to learn as you’re going along without taking too much on at once and without dealing with the political issues.”

Check out the full podcast and transcript here.