Quantifying qualitative data a bacon perspective

In my Business Relationship Manager role I often want to measure how the various relationships are going, I could ask the business partner but depending on how I ask and how often would give me different results. Here's an analogy to make my point clear:

Here is a quick graph showing my love of bacon over time.

As you can see my love of bacon goes up and down over time, these key 10 marks are probably when I’m hung-over.

So depending on when you ask me the answer will be different due to a number of factors such as how much I've had recently, have I just smelt someone cooking some and of course if I am hung-over. So a fair assessment of how much I love bacon would be my average quantified love of bacon which over time is 8.2 out of 10.

But how do you collect this data, well you could ask the question over and over but would that affect the results? I think it would….

Here is a quick graph showing how I get sick of being asked if I love bacon all the time.

Then my overall love of bacon has dropped to 6.7, you’ve made me fall out of love with bacon!

So what is the moral of this story?