Guest Post: The Value Of Customer Engagement

cScape recently released the results of its annual Online Customer Engagement Survey. One of its contributors is Theo Papadakis who wrote the following:

Writing in ClickZ, Neil Mason denigrates the value of the concept of customer engagement on the ground that it is impossible to define and therefore useless. Neil proposes an alternative concept with which he seeks to capture what he thinks the proponents of engagement find useful in that concept. I wish to argue that Neil is wrong in thinking that customer engagement is an elusive concept and that his proposed alternative fails to do the task a properly defined concept of customer engagement can carry out.

A bit on the definition of customer engagement

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 concept of customer engagement however only deals with a particular kind of relationship:

  • Object: A customer’s relationship with a brand, company, product, or consumption topic.
  • Character: A customer’s emotional relationship with the object.
  • Kind: Positive or negative.
  • Content: Engagement is not itself a psychological state but involves psychological states, such as — in the case of positive engagement — sympathy, love, pleasure, pride, happiness, gratitude, empathy, affection etc.
  • Degree: The degree of engagement lies on a continuum that ranges from low engagement, namely, the psychological state of apathy, to high engagement. An engaged customer is a customer with an above average psychological investment/involvement with his or her object of relatedness.

However, given that psychological states are very difficult to measure and, above all, to relate to marketing objectives, publicly observable behaviours that reveal a customer’s object, kind, content and degree of engagement are measured instead.

As there is no simple measurement of emotion, a love or gratitude metric for example, customer engagement needs to be a synthetic concept and metric in order to provide a holistic understanding of a customer’s emotional involvement with an object.

Customer engagement can involve any number of the following token behaviours: word-of-mouth advocacy, provision of feedback, reduced focus on price, increased tolerance of mistakes, willingness to participate in new product development or redevelopment, create fan or anti-fan clubs etc. With the advent of the Internet, the engagement of customers with brands has intensified beyond all expectations and has taken many forms.

As a result, a number of online customer engagement behaviours should be added to the previous list: customers blogging about you, talking about you in forums and social networking sites, writing product reviews on your products on your own and other websites, creating online fan or anti-fan clubs, participating more directly in new product development or redevelopment, creating a wiki customer service or FAQ, creating a product hacking or troubleshooting site, etc.

The reason why frequency and recency of purchase is not in the list, however, is that even a combination of high frequency and recency of purchase is not a telltale sign of customer engagement. Repeat buying might be the result of other factors such as location, lack of competitors, or low price etc. A repeat customer may even be unsatisfied and anxious to defect.

Given the range of all these activities and the value they offer to the organisation, it is indeed very useful to talk about, measure, and try to influence the degree and kind of your target customer’s engagement with your brand.

Secondary value: economy of communication

Of course, we don’t have to use the term ‘customer engagement’. Just like the concept of a ‘car’ is utterly superfluous. We can refer to a car as a human-steered, mechanical-engine-powered, four-wheeled vehicle, but for reasons of economy we call it a car. Is it a car if it has three wheels? Is it a car if the dream of computer-automated steering becomes a reality?

It is important to understand that answering these questions does not require us to identify what the essence of the “car” is. Answering these questions, the business of concept-creation, is based on pragmatist considerations human speakers make in everyday life.

Similarly we don’t have to talk about engagement. We can talk about number of blog posts, forum posts, customer reviews, WOM advocacy, etc. instead, and when your manager asks you about the success of one of your actions in terms of outcomes other than immediate financial impact, you could answer in around as many words as it took to write this post.

Primary value: conceptual value

At this point I can imagine Neil arguing that he agrees with most of the above but objects to calling or, rather, in his own words, ‘dressing up’ these valuable behaviours as ‘customer engagement’. In his article he says:

If we believe there are sets of valuable behaviour that lead to beneficial outcomes, why dress it up and call it ‘engagement’? Why not just have a valuable behaviour index (VBI) instead of an engagement index? At least then we’d know what it is and what it means”.


As I hope I have showed that engagement is not an elusive concept as Neil seems to think, the rest of my argument will focus on why I think that:

  1. Using the ‘valuable behaviour index’ to refer to the aforementioned set of behaviours is problematic to say the least.
  2. Aggregating the aforementioned types of behaviour under a single umbrella term is useful.

I will use the same arguments to prove both points:

  • Neil’s concept of a valuable behaviour index is in some cases too general, while in others too restrictive and therefore fails to capture the dimension we seek to capture regarding the phenomenon of customer engagement. We are looking for a very particular kind of valuable or, in the case of negative engagement, damaging behaviour — behaviour that has a strong emotional as opposed to any other dimension. According to Neil’s notion of ‘valuable behaviour index’, for example, repeat purchase or frequent purchase must definitely be admitted as valuable behaviours. This is not so with the definition of ‘customer engagement’ cScape proposes. Our definition filters out this kind of certainly valuable behaviour. At the same time Neil’s notion is restrictive enough to be unable to capture negative engagement.
  • Emphasising the behavioural outcomes of customer engagement is actually potentially dangerous. The danger lies in that by giving it such a name we are prone to fetishise the behaviours. Although the ultimate objective of marketing efforts that aim at influencing, stimulating or facilitating customer engagement is to influence behaviour, it is not behaviour that is the immediate target of these efforts. These target behaviours are the indirect outcomes of marketing efforts to engage customers. It is psychological states that are the primary target of customer engagement marketing. Although facilitating the target behaviours is important, merely doing so, say by sticking ‘a recommend to a friend’ option, is utterly useless.
  • The repercussions of the argument from economy reach a lot further. Imagine you ran two campaigns, each of which effected an increase in each of the aforementioned behavioural components of engagement in your target customers. How would you decide which campaign was more successful? As soon as you start weighting each behaviour differentially you are on your way to create a synthetic and holistic engagement metric. In short: if you want to be able to compare the outcomes of your marketing efforts along the aforementioned dimensions you immediately need an umbrella-concept (and metric) of customer engagement. This is not to say that it is possible to standardise the components that constitute a customer’s engagement with an object. They will, of course, vary along the dimensions of object, kind, content and their behavioural manifestations. The possibility of having a concept with different components — another frequently encountered criticism of customer engagement — has been discussed elsewhere.

I hope I have dispelled two arguments commonly leveled against the concept of customer engagement, proven its conceptual value, and supported my dual conclusions that:

1) Customer engagement is not elusive, and therefore does not have to be a nebulous concept.

2) Customer engagement is a useful concept for marketers’ ability to: achieve economy of communication; capture a customer’s emotional relation with a company, brand, etc.; and compare.


2 thoughts on “Guest Post: The Value Of Customer Engagement

  1. Theo,

    I was with you for about half of the post; a quite elegant discussion of the notion of engagement and the emotional component. I agree Neil’s argument is largely semantic.

    At the same time, how do I take action on the approach you propose? Like Neil’s VBI, your approach to the analysis of engagement is not likely to be portable across channels or platforms, and may easily cost more to implement than the value of the information gained.

    You said: “the ultimate objective of marketing efforts that aim at influencing, stimulating or facilitating customer engagement is to influence behaviour”. I’m with you. Given any action, the likelihood to repeat that action has to be a measure of engagement with the action. Likely to repeat = engaged by the action. Less likely to repeat = less engaged with the action.

    In the end, I don’t need to know the “psychological states” of why the action is repeating; if there is value in the action and if I can predict the likelihood of the action to take place, I have a uniform “scale” I can use to rank enagagement on any action across any platform.

    Given the Recency of an action predicts the liklihood of it to repeat, I’m not sure why I would need to go any further than that.

    I’m perfectly satisfied to have a measure of engagement that gets me 90% of the way there at 1% or less of the cost suggested by your approach.

    Not that anything is wrong with your approach, but I would make sure that last 10% of the knowledge you want is actionable and worth 100 or perhaps even 1000 times the cost of using a simple Recency metric to get most of the way there.

    Keep up the good work.

  2. Jim

    First of all I want to excuse myself for delaying my response. I just wanted to make sure I can respond carefully and constructively to your objections to the way I approach ‘customer engagement’.

    If I understand you correctly, you are making two points about my approach. The first one – that my analysis of customer engagement “is not likely to be portable across channels or platforms” – I do not see how you arrive at and would like to ask you to tell me a bit more about. I have addressed this point in another guest post in Ron Shevlin’s blog ‘More thoughts on Customer Engagement’ where I tried to show why the notion of customer engagement I discuss is in fact portable.

    I will therefore focus my response to your second criticism, namely, that my approach to engagement ‘may easily cost more to implement than the value of the information gained.’ This critique you level because you have an alternative to my approach in mind.

    According to your approach “Given any action, the likelihood to repeat that action has to be a measure of engagement with the action. Likely to repeat = engaged by the action.” I don’t agree with this because I think the equation you propose is incorrect. Likely to repeat can equal a host of other things other than engagement. For example: Likely to repeat = compulsion = habit = satisfaction = lack of choice etc. Unless of course you are prepared to equate all these things with ‘engagement’ and suggest that ‘engagement’ is equal (=) to all and every motive due to which a customer may be likely to repeat a certain action or behaviour.

    If this is what you propose I don’t think that your proposal does any justice to the value I along with other proponents of customer engagement qua emotional investment find in that concept. If this is the case then I think we are talking past each other. Let me explain why.

    You believe that there is no distinction between Retention and Engagement. This you explicitly suggest in your post ‘Recency Defines Engagement’. You also believe that there is no distinction between Recency and Engagement. This you explicitly suggest in both the aforementioned post and in your comment to my post.

    On the other hand, I have explicitly expressed in my post the belief that there is a useful, for marketing purposes, distinction between engagement and recency/retention. It is easy to find an example where what I call ‘engagement’ and what every marketer calls ‘retention’ or what you call ‘recency’ can come apart. It is a case where the customer performs the target action the marketer wants him to perform in a way that reveals a low – according to my definition of engagement – degree of engagement, namely, apathy, or even negative engagement (to whatever degree). A case of apathy is when you buy an alright cup coffee, at an alright price, from a barrista you have almost no feelings towards. She wasn’t too friendly, fake polite or outright rude. In fact you don’t even remember her. You still go back, to the same place, perhaps out of habit or maybe out of lack of choice, but you have no emotional investment in that place. Another exemplary case is one where a repeat buyer actually badmouths the company whenever he/she gets the chance to, but continues to provide his/her custom simply because he/she has no other choice.

    Neither recency nor retention are notions that are fine grained enough to be able to discriminate between the aforementioned scenarios. Neither captures degree of engagement. Neither notion captures kind of engagement either. They therefore put apathetic or negatively-engaged repeat customers in the same basket as every other repeat or recent customer, irrespective of the reason they repeatedly or recently performed the target action. Finally they do not capture negatively engaged non-customers at all, despite their, possibly, high degree of engagement. These are all jobs the definition of customer engagement I have proposed can carry out.

    We are talking about two different things. You say ‘Recent activity is indicative of engagement’ and ‘the more recently that activity occurred, the more ‘engaged’ the customer’. In my terminology, Recent activity is only indicative of recency relatedness. This is the crux of our semantic disagreement. You frame ‘engagement’ as synonymous to ‘relation’ and irrespective of degree, content, character and kind. The way I and others use the concept of ‘customer engagement’ is to refer to a particular kind of relation (see the aforementioned definition of engagement) between the customer and the company/brand.

    The way I define engagement, retention and recency, just as much as loyalty and profitability are indirect outcomes of a customer’s high degree of positive engagement with a brand. Recency is not a telltale sign of an engaged customer. Not every correlation is a two-directional causality. Although a positively engaged customer must be a recent customer, a recent customer does not have to be an engaged customer, and certainly not necessarily a positively engaged customer. S/he may be dis-engaged or even negatively engaged.

    Of course I find your Recency chart and the Lifetime value graph it supports very useful. But it only measures the recency of a customer’s relation with a brand which as you have shown in ‘Recency defines Engagement’ is useful in itself.

    Understanding why customers engage emotionally with a brand, understanding the customers’ desire to engage with a brand and through or about that brand with others, helps provide with additional insights that your approach I feel is unable to capture and which to my experience are very valuable.

    Thanks for your comment.

Comments are closed.