I would like to address two related issues: 1) The metaphysics of engagement: Does customer engagement exist? and 2) The possibility of a uniform definition of customer engagement.
First off, metrics do not exist in the human-independent way that a tree does. If there was no human race, a mountain would still exist, but the metric called ‘customer engagement’ would not.
The only justification for creating a metric is to measure the property of something that really exists – e.g., the height of a particular tree (presuming, of course, that that property exists).
Second, engagement is a bipartite relation, i.e., X is engaged with Y. Indeed, it denotes the very fact of the relation. Any relation is an engagement and vice versa. The use of the word engagement in marketing is best expressed as the degree or intensity of a person’s psychological investment and involvement with an object (I don’t mean physical object – it can be anything).
Engagement is, hence, not a psychological state, but the degree of a person’s psychological investment in some object. This investment involves psychological states, such as — in the case of positive engagement — sympathy, love, pleasure, pride, happiness, gratitude, empathy, affection etc. Each of these states induce, in turn, a physiological state that is characterised by cognitive, somatic, emotional and behavioural components. The somatic effects of fear, for example, involve the tightening and priming with oxygen of muscles used for physical movement, increased heartbeat, etc.
For marketing purposes, however, measuring the somatic effects of certain states is rarely undertaken. Instead the more easily measurable behavioural effects of psychological states are the target of customer research. Although the behavioural effects of a psychological state are subject to culturally-specific variation and hence are not as universal as the somatic effects, they are closer to the final object of customer research, i.e., understanding and predicting customer behaviour for the purpose of improving ROI. Try to get likelihood to perform target action from data such as an increase in adrenaline release along with 10 other somatic symptoms.
Although the behavioural manifestations of psychological states are culturally constructed, the strength of social conventions commands a high enough degree of uniformity that any skepticism regarding their predictive value based on some form of radical cultural relativism (linguistic or other) is largely unfounded.
So although the behavioural manifestations of an engaged human being will vary — depending on whether it is a positive or negative engagement, as well as across different objects of engagement and media through which these behaviours are/can be expressed — engagement understood as the degree/intensity of the subject’s psychological investment, does not vary.
In so far as the psychological states that constitute positive or negative engagement are objective, their degree or intensity, manifest via public behavioural symptoms, will also be objective and measurable. Finally, if measuring that intensity is useful for predicting behaviour then the metric is also useful. Although I believe the answer is a resounding yes, it’s another topic altogether.
[Note: Theo and his colleague, Richard Sedley, have done a lot of great work regarding customer engagement. Check it out.]