Column · feature outcome
Because the buyer evaluates the change, not the mechanism. A list of capabilities asks them to perform the translation themselves, and most will not.
For each feature, write the sentence a buyer would say after using it: "I no longer have to do X." If that sentence is weak, the feature is weak regardless of how well it is built.
The outcome first, the mechanism second, and only if the mechanism removes a doubt. Most pages invert this, and the reader leaves before reaching the part they cared about.
Remove every feature name from your page and see whether the promise still stands. If it collapses, you were selling parts rather than a result.
See the data this is built on
Open the snapshot