Press "Enter" to skip to content

Feature Spectrum Analysis

I’m sure what I’m going to show exists under a different name but I call it Feature Spectrum Analysis. You take a feature and evaluate it across as many spectrums as you wish. This process helps me identify what features are the most important to my users and achievable by me.

How this works

  1. List all the features you’re thinking of implementing.
  2. List all the measure categories you’d like to evaluate against.
  3. Fill in the grid. Honesty is critical. (I use a 1-10 scale, but that’s arbitrary)
  4. Reassess your numbers. Do they make sense when you compare them? Is feature X that much easier to implement than feature Y?
  5. Make a stacked bar chart, or any other chat to help you look at the data.

Example

I recently did this for my Lazy Gardener app as I was trying to decided on the direction the app should go. Should it go in a utility direction, or keep it moving towards being a journaling app. I decided on the following metrics.

Utility: Is this feature beneficial to people who want the app for utility reasons? One example would be using local weather to remind the user to water their plants since it will be hot a few days in a row.

Journal: Is this feature beneficial to people who want the app for journaling reasons? Sharing your garden’s timeline to the web for friends and family is a great journaling feature, but serves no utility.

Implementation difficulty: How hard will it be for me to implement the feature at a user testable state?

Usefulness: How useful will the be for the user in their everyday usage of the app?

Visibility: Is this something that the user will see an notice, or is it an non-discoverable feature/improvement?

Personal enjoyment: How much do I want to work on this feature?

It’s ok to make multiple passes at the values, add and remove columns, even add or remove features. The purpose is to evaluate what features are most important for your users and the product.

Evaluation

Surprising to me was that most features tended towards being utility based. The web site generation feature is one that I want the most, but serves no utility purpose. Worse than that it will be one of most difficult to implement. Looking at the graph I see that scheduling reminders, online backups, and Siri support should be my next features. A UI pass and new themes shouldn’t be neglected as they bring value to the app with very little effort.

Conclusion

I like having this tool as part of my decision making process. The steps descriped are not ridgid, but they work for me. At the moment, each metric is weighted evenly, but should Usefulness be the same as Visibility? With weights you could then create a formula and more objectively assess the data. I’m keeping some subjectivity this time around because I’m primarily making this app for me. The data gives you another perspective, it should influence, but not dictate.

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *