In my last post, I told you about how you could use a decision tree to automate decision-making processes, thereby making it faster for you to arrive at decisions and make the outcomes of those decisions more consistent. In this post, I’m going to show you how to actually build a decision tree using an example from the last lesson.
We need decision trees and decision-making processes for “regular decisions” – decisions which occur several times throughout a given period of time. A good example of this kind of regular decision is the decision to add a new feature to a product or service, and that’s the example we’re going to use today.
Example Decision: “Should we add a new feature to our product?”
Step 1 - Begin with the First Key Question
The ultimate question in this example is “should we add a new feature to our product?” We have to produce a “yes” or “no” answer that is consistent with our market positioning, needs of the customers, and our product design aesthetics.
So the first question that we need to factor into our decision making process is our customers – if an insignificant number of our current customers or easily accessible prospective customers would be willing to use this feature, it might not be cost-effective for us to implement this feature. Thus we begin this decision tree with the simple question, “Did our customers request this feature themselves?”
We also include all of the likely answers to this question, which in this case are simply “yes” and “no.”
Step 2 – Ask the Next Level of Key Questions
We now have two paths in our decision tree:
- The feature was requested by customers
- Or the feature was not requested by customers.
Features that have been requested by customers are going to be considered with a different set of criteria than features that were not requested by customers. So I’m going to represent that on my decision tree by asking two different sets of follow-up questions:
Next we provide some answers to these questions.
The concept of paths here is important—remember those “choose your own adventure” books that you read when you were a kid? Your adventure ends in a different way depending upon the choices you make as you read the book. A decision tree is a graphical way of depicting those different choices and outcomes—each decision you make moves you along a new path.
Step 3 – Keep Asking Questions until You Have Enough Information to Make Some Decisions
At this point in the decision tree, we’ve considered whether or not customers recommended the feature themselves, why customers would want to use the feature if they didn’t recommend it themselves, and how many customers actually made the recommendation for this feature.
We really still do not have enough information to arrive at any decisions for any of these paths at this point, because we still don’t know how expensive it would be to implement the feature in the case of customer-recommended features. And we don’t know how many people would use the features which were not suggested by customers. Let’s ask those questions next.
I decided that it would be easier to figure out the “feature requested by customer” path since the decision tree was starting to get a little big; I simply hid the right-hand side of the decision tree.
Now we’ve reached a point on the decision tree where we are able to arrive at some solid decisions:
- If thousands of customers have requested this feature and it’s easy (cheap) to implement this feature, we should do it.
- If only a few customers have requested this feature and it’s not easy (expensive) to implement this feature, we shouldn’t do it.
These seem like reasonable outcomes thus far. For the remaining undecided part of this “feature requested by customer path” we will need to ask and answer a few more questions.
The next question we ask for both of the undecided cases is “could this new feature attract a lot of potential new customers?” Obviously this is a question that might take a while to answer; in fact, it would probably take some market research to answer the question. But, nonetheless, it is the next key question that must be addressed in the decision-making process for adding a new feature to a product.
By now you get the idea—you need to simply keep asking questions until you reach a point in all branches of your decision tree where you feel that you can make a consistently good decision.
And that’s all there is to it. If you have to make regular decisions all the time, developing a formalized decision process (akin to standard business processes or sales processes) is well-worth the time and trouble.
If you want to try making your own decision tree, download a free trial of SmartDraw.
If you liked this post, make sure you subscribe to our RSS feed or follow us on Twitter.
