Posts Tagged features

Prioritizing Features

Last Friday I blogged about user features and how we should not give in to every request… Over the weekend, I realized I didn’t give any advice about how to prioritize these features. So, here I am to rectify that! A colleague of mine,  Matt Gregg , shared a technique with my old UX team. It’s called the Kano Method.

The Kano Method allows you to figure out which features really need to be included into your product. According to the method, each feature falls into 1 of 6 categories:

  1. Must have features. Users MUST have these features to purchase/use your product.
  2. Desireable features. These are really useful features that can add value to your product.
  3. Exciting features. Great bonus features that will excite and entice your user population.
  4. Indifferent features. These features are condered ok, but not really worth much notice from the user.
  5. Questionable features. Results for these featuers were not clear. There was conflicting data gathered. Quite likely, your questions were too confusing and should be rewritten, or the user truly does not care one way or another about the feature.
  6. Reversal features. These features may actual cause your users to stop buying/using your product.

To categorize the features, you use a specific survey style. Users rate each feature in a “functional” way and a “dysfunctional” way. So, for example, if you want to rate a password reminder feature. You would have 2 questions (functional and dysfunctional, respectively):

  • How would you feel if you had the ability to request a password reminder on the website?
  • How would you feel if the ability to request your password were not available on the website?

Another example (functional and dysfunctional, respectively):

  • How would you feel if a cooking website contained more recipes than most websites online?
  • How would you feel if a cooking website contained less than the average number of recipes?

Your users would then categorize each question using a set scale. Note: It is not a good idea to list these features and rate them using a Likert-style scale. The user needs to understand that this is NOT a rating scale, instead they are categorizing each feature.

  1. I like it that way
  2. I expect this feature to be in the product
  3. I am neutral
  4. I can live with it that way
  5. I dislike it that way

To analyze your results, you simply calculate how each feature was categorized functionally and dysfuctionally and use the following matrix to analyze the results:

 Kano Matrix

Think through an individual feature, e.g. being able to request your login password. Let’s assume this is a mandatory, must have feature that the user expects. For the first question listed above (How would you feel if you had the ability to request a password reminder on the website?), most users would select category 2 – I expect this feature to be in the product. For the dysfunctional version of the question (How would you feel if the ability to request your password were not available on the website?) most users would select 5 – I dislike it that way. Therefore, an “M” square would be selected in the matrix and the feature would becategorized as a must have.

Now, lets look at a more confusing item on the matrix. Let’s say our feature is addition of a jacuzzi to a hotel room. The functional question could be: “How would you feel about having a jacuzzi in your room?” Let’s assume most users categorize this as a 1 – I like it that way. Next, they read the dysfunctional question: “How would you feed about having a standard tub in your room?” We’ll assume users will select a 2 – I expect this feature to be in the product. So, we have a feature that users like, but it appears like they do not expect it, since they do expect the dysfunctional feature. On our matrix, this feature would be categorized as an “E” – an Exciting feature.

Features can then be ordered based on the priorities of your team and the users. One of the wonderful benefits with using this method is that you can prioritize user interview findings with a large audience very easily via  a web survey. Additionally, it can be useful tool when bartering with your team!

Leave a Comment