Those of us working in and around the internet are generally in agreement that it's a good idea to reduce the number of clicks/taps it takes a user to complete a task. By reducing frustration in an experience it's likely that someone will return and use your product again.
But sometimes it is good to introduce some friction. If you just give users exactly what they expected to find, what is there to set you apart from other sources?

Picture credit
Google earned our trust early on because they were able to give us relevant results quickly and with zero fuss. Their competitors at the time were bloated portal sites like lycos, aol, yahoo, and Google's simplicity won the day. Compared to MySpace, Facebook 'just worked' (I can think of another example there). But those sites haven't stopped there.
Having my mind switched on to this in my current role has made me realise that this is going to be a continuing trend as the world truly becomes digital native and the internet of things surrounds us in our 'offline' lives - it will not be enough that your product works the best, or is the most popular right now, it will need to surprise and delight people, connect emotionally, and seamlessly enhance the user's own self-projection in a meaningful way.
I guess branding people have understood this for a while, but it totally applies to user experience design as well. That's why we see Google adding circles, +1, places, and updating the look and feel of their entire suite of products. Their core business is still search, but they compete on a much broader plane because of the extended value they can offer to that search.
I just the other day learned from John Gibbard that there's a well-establised term for this concept; it's called 'poka yoke'. It was traditionally used in manufacturing, as a fool-proofing method. By adding steps or barriers to factory processes, companies are able to reduce production errors. Essentially, if you put an unavoidable thing in the way, you will make sure that it's always done. On the internet, it's sometimes been little things like suggesting purchase add-ons as part of a payment funnel.

And we've always been doing our own little poka-yokes in our own lives as well. There are some great examples on Dan Lockton's blog.
So I guess what I'm realising is this: you can deliver the best user journey with the most rewarding end points, but what is there to stop a competitor popping up and 'building' on your 'best practice'?
As well as being best-in-class we need to start thinking about the opportunities to delight people. What steps will users be ok with us adding to processes so that we can make sure that they're getting the most from their time with us, and that they're taking away a memorable experience they can share, and a brand they can return to?