So I’ve gotten a lot of mileage by tweaking the destination of bookmarks. For instance:
Make the bookmark for your email client (eg. Gmail) link to your “drafts” folder, rather than the default inbox. Most often, I open my email client to write an email, but I used to catch myself getting distracted by something in my inbox. Also, drafts used to languish - now, I’m much more likely to complete them.
Make the bookmark for twitter to you /following, rather than the default home feed. Most often, I notice I want to open twitter because I’m wondering if a specific person posted anything new. In the past, I could get swept up in the deluge of postings and forget what I was looking for - now I’m much less likely to get distracted.
Make distinct bookmarks for Amazon’s video service and create a bookmark to you “Wishlist” for purchases. You engage with their services for two distinct reasons while they’re trying to make the movement between entertainment and consumption frictionless. Make the decision to either watch content or buy something more deliberate. Similarly, when you go on amazon to consider a purchase, you now confront other things you’ve wanted in your wishlist, and it forces you to weigh the opportunity cost across categories of purchases (eg. that shirt vs. the audiobook) as opposed to only within categories (the wide or narrow-collar light blue poplin shirt). This something research has demonstrated people are pretty poor at doing, decreasing our satisfaction on how we spend our money.
These are just my examples, but the underlying thought-process is the same: “Why do I engage with/get value from this service?” and “When does this service pull me into distraction?” The inbox and aggregated feed, were choices by Gmail and twitter, respectively. Almost by default, it’s worth considering that the home pages of any service that depends on users’ ‘time on site’ is best avoided unless necessary.