Imagine you want to give your living room a new look, and you come up with two ideas: rearranging the furniture, and moving a switch from one wall to another.
Which one would have a greater impact on how the room feels? The furniture, probably. And which one would be harder to implement, or to revert if you change your mind? Moving the switch, assuming it involves rewiring.
Those are two different dimensions — impact, and how hard a decision is to change or revert — and sorting decisions along both gives us a very useful framework.
| Easy to change | Hard to change | |
|---|---|---|
| High impact | D1 | D2 |
| Low impact | D3 | D4 |
In the living room example, rearranging the furniture is a D1 decision, and moving the switch is a D4.
Building products means facing a lot of decisions, and knowing which type you are dealing with can radically improve your team’s velocity and impact.
D1 decisions are perfect for rapid testing and experimentation in production. This is where your experimentation resources should go. The key is to move fast and find the best alternative as quickly as possible, without extended research or long discussions. Example: the position and color of the “Buy now” button on an e-commerce site.
D2 decisions are where the research heavy lifting belongs, and where the discussion hours are worth spending. Take your time with these, because once implemented they are hard to change or revert. Example: moving from one database technology to another.
D3 decisions are so abundant that we usually don’t think of them as decisions at all. They are the small stuff. Make them fast and don’t spend time or resources on them, and above all, don’t run experiments on them. Example: the name of the “work with us” section in a website footer.
Finally, D4 decisions. These are rare and dangerous, because something that is low impact today can become high impact later. Be careful not to sink too many resources into them, but analyze them well enough to catch their long-term strategic implications. Example: the font combination for a brand.
This framework helps you point your team’s time at the right problems and spend your experimentation and research resources where they actually matter, which translates into speed and impact.