The answer to this question relies on users. More and more the question is emerging in smart data marketers about the quality of data. After a long drought of winging it, marketing departments have feedback that is timely and specific.
Campaigns would tell the story over time ten or fifteen years ago, mostly at the level of what rocked and what flopped. These last few years the data is flowing so fast marketers were splashing around in lakes of information.
Now we are learning to swim and curious about data quality. My favorite questions are where do the respondents sit? I know where they stand from feedback but what is the demographic, interactive and response profile of the commenter.
We all know those who have an opinion on everything or share hearsay as fact. Data for new features or market strategies needs to flow from real user data though. Yes if the core value proposition was based on ROI then the CFO is a user. We solve problems with any solution and often one solution solves many problems cross functionally if we hit the mark.
Flipped on its’ side if we have sought to help multiple functions and only one responds, data point from the others are as valuable as the first. Personally I like the label of data scientist but see little science in their activity. True science test to failure and creates models that can explain the past events as well as predict with some accuracy the future.
Truly if you can’t understand the past with your concepts then I question your value going forward. This requires looking at your features and products impacts in hypothetical, I agree. You could not test the new mobile app six years back when smart phones as a platform did not exist. You can however test the hypothesis. Would being able to understand the fare and availability of a black car service helped you choose a service over a taxi?
Frankly this may be a truer test of the value prop and business value rather than the tech that delivered it!