Updated: Feb 22
Emotion is a big word! It’s also one that causes some confusion. For instance, many CX and AI professionals (working in CX) believe that you must design for and manage emotions since:
1. Emotions are separate from cognition.
2. Emotions are irrational.
3. Emotions ‘drive’ customer behaviour.
4. More positive emotion is a good thing. And more negative emotion a bad thing.
5. The functional is not emotional.
This leads them to overweight CX programmes towards designing for delight or ease; believing that those who are passive are of limited value and that we must embed emotional clues in journey maps and seek to measure emotions effect.
To be fair, this emphasis on emotion is not entirely incorrect.
After all, emotions are critical to decision-making. They tell us what is important ‘to us’ and enable fast decision-making when faced with a complex world of limited information.
As Antonio Damasio, the famous neuroscientist said: “Without the filtering provided by emotions and their somatic markers, the data sets for any given decision — whether it’s what to get for lunch or whom to marry — would be overwhelming. The working memory can only juggle so many objects at once. To make the right call, you need to feel your way — or at least part of your way — there.”
In effect, emotions are for learning, they alert us to what is important: however, what is often missed by CX professionals is that when we have learnt we don't need 'to feel' so strongly.
To ‘manage’ emotion we must therefore not fall into folklore and mythology.
Myth 1: emotions are separate from cognition
To see emotion as the only influencer on consumer decision-making is wrong. Cognition is pretty important too!
Indeed, in psychology the current approach to emotion is that it is based not on some emotional cue that sits outside what we experience, but on how we first appraise a situation in light of our well-being and goals.
Let me explain.
Imagine a football match. You have invited your friend along to see the game in which your favourite team is playing. Your friend doesn’t much like football but is coming along anyway. Halfway through the game your team scores a goal. What’s your emotional reaction? Happiness, of course. What’s your friend’s emotional reaction? Well, he might be happy for you, but couldn’t really care less about the game.
Here, the same game was being watched but the ‘emotional’ reaction was different because the meaningfulness of the event to each person was different.
As the famous psychologist Lazarus put it: “Before emotion occurs, people make an automatic, often unconscious, assessment of what is happening and what it may mean for them or those they care about. From that perspective, emotion becomes not just rational but a necessary component of survival.”
For Lazarus, different emotions are elicited when situations are evaluated according to three categories:
1. Relational aspect: the relationship between a person and the environment.
2. Motivational aspect: an evaluation of a situation in relation to the achievement of one’s goals.
3. Cognitive aspect: an appraisal or an evaluation of how relevant and significant a situation is to one's life.
Even Lisa Barrett Feldman - who makes a strong case for emotions and cognition arising together - states that emotions are socially constructed and that FMRI studies FAIL to find any brain system that is singularly the 'organ' for a specific emotion.
Emotions are therefore a downstream metric and should be treated as such. I summarise this by saying: ‘It’s not the emotion that matters, but what it means’.
And this extends to its value too.
Sometimes, frequently even, we don’t need emotion to be loyal.
I go to Morrisons every week, have done for over 10 years. Do I need to float on a cloud of emotion each time? Indeed, I can ‘feel’ negatively because I care: ‘why have they changed the store around, I preferred it as it was’. Are we to say, therefore, that negative emotions drive loyalty as well? Or, since I usually go shopping without being heavily engaged with the experience, that passive feelings are in fact the strongest loyalty drivers?
But I won't be too churlish. Automatic emotional response, for instance. can be important and this involves less conscious cognitive engagement. But in our fleeting response to brand imagery UI , or UX writing are we saying emotions appear from nothing or is there a reaction based on a primed more unconscious level of cognition that leads to an emotional response?
Implications for CX professionals
Firstly, define emotional moments since these flag up important and salient customer experiences. But be aware that is far from saying that feeling an emotion is a judgement of value! See my Morrisons example. This is also an area where inference from big data can help in situations that are simple and require less cognitive understanding i.e., we saw you looking confused on a website through our breadcrumbs data or the threshold for web download speed was breached you must be dissatisfied.
Secondly, understand the emotional context, since emotions are based on appraisals. You cannot set emotion outside of what it relates to.
Thirdly, since business is based on goal directed behaviour, the most important emotions are frequently those grounded in cognitive reflection. This is good news since it means that surveys can usefully collate big ticket influences. However, this is sometimes less good news for AI which by its very nature strips out self-reflection (the cognitive reflection) in favour of a general algorithm.
And the question remains that what we measure in surveys may be pretty much the same as attitude anyway.
Fourthly, we still have the impact of fleeting and often unconsciously felt emotions; reactions to brand for instance. To understand this requires more ethnographic approaches and ones that deliver a closer appreciation of the situation.
Finally, be deliberate and consistent in collating emotional data using the following approaches:
Ethnography helps determine emotional effects.
Customers and employees are not good at relating back the full emotional experience. However, by using qualitative methods such as ethnography we can shine a light on micro-moments that are often fleeting.
Example: UX writing
Our UX insights identify points of confusion on a website. For instance, a word such as ‘prognosis’ that is or could be misunderstood. No customer reports it, but nonetheless we see it and can appreciate how the experience of doing business with us could be improved by changing the word.
By doing so we engage in marginal benefits. Focusing on small changes that added together over time improve the whole flow of the experience. So while on its own, there is no root-cause effect on value, as a micro-moment it remains impactful as it puts our customers at a higher risk of feeling ‘less disposed’ to engage with us. For more on dispositional over root cause effects see Cynefin.
I have to also say, I never understood why UX research is considered different from CX. As a qualitative method ethnography picks up those micro-moments of importance to CX as well.
We can also as ethnographers bootstrap ourselves into the world of the customer and say, 'what would engage the customer more'. For instance, that UI looks bad, we can do better; if we have boring experience why should we just focus on the current drivers of that experience after all.
And since innovation can come from anywhere, we need to treat its collection a little differently to backward looking service quality insights i.e., co-creation work groups and digital platforms, the use of questions that engage in conversation, engagement with a diverse set of thinkers.
When dealing with subjective data on what we could do we need expert judgement to build hypotheses for action.
Use narrative not scales to measure emotional content.
The reason for this goes back to anthropology: in interpreting our experiences, customers use stories. For instance, no-one walks out of a store and says that was a great 8.5 out of 10 experience; and no-one evaluates each and every detail of their experience to come up with an algorithm of ease or delight usable by AI.
I signpost SenseMaker and make the point that in the KPI 'more stories like this, fewer stories like that' there is indeed quantification.
Since emotions are grounded in appraisal (goals, wellbeing), just going with a set of emotions or sentiment classifications is not enough. Only by understanding the appraisal basis to emotion can we reveal ‘our true motivations’.
Quantifying appraisals and emotions can be done through applying algorithms to text. On the basis of say a 60% level of accuracy, this can be sufficient for mass data feeds such as social media and where effects are linear If you are surprised by this acceptance, don't be. I follow the Cynefin approach, if your business is pretty simple, simple measures are fine.
However, my favoured method at least with smaller samples or complex situations is to get customers or employees to quantify their own narratives for emotional content and attendant appraisals.
How customers interpret their own narrative and emotional response adds another layer of meaning and has to come from the respondent (self-quantification). No AI can layer on subjective meaning, it will always strip out context by using general patterns that miss at best nuance and at worst the point entirely.
My argument is: only once we understand appraisal do we move to personas, since in appraisal-directed behaviour it’s not the ‘personality’ dynamics of the role or some other criteria that is important but an understanding of the appraisals themselves.
Examples: Tarmac and Ericsson
Tarmac wanted to build a set of journeys’ around their small-medium and large customer base. On review this missed the similarity between segments and an understanding of where their clients were heading i.e., this was more about your a small business, we don't want to serve you, so we will put you all on self-serve digital platforms. There is no doubt some value in this approach, but it misses out the potential to find new opportunities that are emerging and where these might cut across segments.
By engaging in Journey based selling it was found that many new opportunities had been missed simply because no one had got into the life of the customer beyond selling what Tarmac was pushing. Journey based selling uncovered new opportunities to help optimize the clients experience based on their emotional needs, wants and expectations.
For instance, this surfaced:
Demand for large trucks based in Holland that could deliver cement and aggregate to Midlands firms working in warehouse construction.
Micro-surfacing with SME firms in the North of England
Journey based selling mapped out the emotional needs, wants and expectations of the client based of course on deeper levels of conversation and then figured out where Tarmac could add value.
This uses a consultative sales technique but against the client journey, not necessarily the journey with Tarmac. The ideas surfaced could then be held and shared in a CRM database.
Critically, this had nothing to do with profiling the persona or size of business, but had everything to do with deep discussion on their situational needs which could then be profiled against any size or type of business. The kicker for such discussions frequently being due to a new IT infrastructure being put in place or a new digital platform.
With Ericsson: the desire of BTEC (Bahamas Telecom) to have a system built that would track dropped calls off network by their key Cayman Island banker clients. Likewise the need to define use cases for their new Service Operations Centre.
Myth 2: Emotions are irrational
Here is a quote from Oatley and Johnson-Laird on the irrationality of emotions: “A crucial distinction, however, is whether an emotion is incidental to a reasoning task or emerges naturally from it. When an emotion arises from the task, a recent hypothesis is that reasoners are more motivated and more likely to consider possibilities that they would otherwise neglect.”
Hence, emotions are not only grounded in ‘the event’ but can act to enhance a rational response.
For sure we can act irrationally ‘with emotion’ but it is a step too far to say emotions are irrational. Especially when we simply would not understand what is rational for us without them.
An example of this comes from the work of the great neuroscientist Antonio Damasio. When dealing with a patient (Elliot) who could not feel emotion but still had unimpaired cognition, Professor Damasio found his decision-making was impaired. In Descartes Error he writes:
“What was even more confounding is that Elliot could think up lots of options for a decision. When given assignments of assessing ethics (like whether or not to steal something for his family, Les Miserables–style), business (like whether to buy or sell a stock), or social goals (like making friends in a new neighborhood), he did great. But, even with all the idea generation, he could not choose effectively, or choose at all. “I began to think that the cold-bloodedness of Elliot’s reasoning prevented him from assigning different values to different options,” Damasio writes, “and made his decision-making landscape hopelessly flat.”
Reduction in emotion may constitute an equally important source of irrational behavior!
Implications for CX professionals
Firstly, do not consider emotions as irrational and something to avoid. When dealing with human systems we cannot and should not ignore them. We equally should not try to ‘take out emotion’ through processes such as zero UX or considering customers only as cost-benefit calculators.
For instance, if I cannot appraise you as different from the competition, my only appraisal would be related to convenience and low price. And this appraisal of difference comes from my emotional understanding of you: the intent behind the eyes.
Secondly, do not assume that we can ‘create’ irrationality in the customer by ‘making’ them emotional so they ‘irrationally’ stick with us for the long-term or spend more.
Myth 3: Emotions ‘drive’ customer behaviour
For sure, sometimes an emotional response does represent a driver. If a customer service rep is rude to me on the phone, I appraise it as a bad thing. I feel annoyed and never use you again.
However, many times the fleeting moments that emotion represents are reflections of something else. Their dispositional not driving effect, here are some examples:
· Weak signalling: Small emotional clues (weak signals that I attend to) influence how I am disposed. For instance, consider how Luton Airport ask me to pay £1 for a security bag – where Heathrow do not. This is something I dislike and impacts on my emerging and changing sense of the quality of customer service (the appraisal) at the airport. Quality of customer service in essence is a gross signal, a big ticket item but one defined by the many actions and interactions that are experienced by consumers within the airport - the weak signals.
Hence, during my stay in the Leela hotel I felt disappointed that they had started turning off the air conditioning to rooms and removing fruit displays. In this case, the experience is not enough to turn us off from using the service again, but they are enough to degrade our engagement.
Of course, we need to know when a degradation hits a tipping point and we lower our spend or just leave. The key thing is, frequently we cannot know! All we can hope to do is be aware of it, and respond in kind.
Likewise, we can become aware of how to use weak signals as part of a Nudge approach. Moving people towards a disposition rather than yanking them to new set of behaviours.
· Emotion is indirectly impactful: How emotion shapes behavior: feedback, anticipation, and reflection, rather than direct causation (source: Baumeister, Vohs, DeWall, Zhang):
Fear causes fleeing and thereby saves lives: this exemplifies a popular and common sense but increasingly untenable view that the direct causation of behavior is the primary function of emotion. Instead, the authors develop a theory of emotion as a feedback system whose influence on behavior is typically indirect. By providing feedback and stimulating retrospective appraisal of actions, conscious emotional states can promote learning and alter guidelines for future behavior. Behavior may also be chosen to pursue (or avoid) anticipated emotional outcomes. Rapid, automatic affective responses, in contrast to the full-blown conscious emotions, may inform cognition and behavioral choice and thereby help guide current behavior. The automatic affective responses may also remind the person of past emotional outcomes and provide useful guides as to what emotional outcomes may be anticipated in the present. To justify replacing the direct causation model with the feedback model, the authors review a large body of empirical findings.
Implications for CX professionals
Firstly, use real-time narrative collection and vector methods to alerts us to which emotional narratives we need to amplify and dampen (the KPI: more stories like this, fewer like that).
Secondly, use the Cynefin framework to understand better whether appraisals and their emotional response fit into a complicated mechanistic approach or a more dispositional approach.
Thirdly, enable human judgement and interpretation whenever viewing a data driven change. Heighten our sense of understanding by being better observers, SenseMakers and co-creators. That way we will get a better feel from the indirect effect of emotion.
We will never get there if we outsource this understanding to AI alone (except again in linear and simple situations).
Myth 4: More positive emotion is always a good thing
If emotions are for learning, then it stands to reason that they are not required if we have learnt: except perhaps in a muted way.
If, for instance, we went to our local supermarket for the first time we might find ourselves feeling pleased because of its convenience and value for money. On the tenth occasion, our emotions are muted to the level of the category.
We can still be tremendously loyal of course, but we don’t need to feel it!
So, customer value creation is critical but that is not the same thing as ‘feeling’ positive. In fact, we can, for instance, ‘love’ a hotel: which is why we notice and feel strongly about any negative things that happen.
The journey mapping rule of trying to emotionalise every action is incorrect and can just lead to exhaustion or annoyance. Personally, I don’t want every touchpoint on a customer journey to be something I evaluate as ‘positive’. And many times no experience is the best experience, especially when dealing with a hygienic and transactional one.
This is also true of employee experience as well. We should not overrule our natural human ways of interacting with a demand for fake smiles.
I say whither the Kano model. Heading to delight (dependent on category) is not always a desired or stable condition.
And if you are still unsure about the importance of this emotional regulation and by implication its criticism of always going for high positive states, I quote Baumeister:
“There is evidence that people spontaneously regulate their emotions (Forgas & Ciarrochi, 2002). Immediately after an emotional event, people in both happy and sad moods experience more mood-congruent than mood–incongruent thoughts. With time, however, the content of people’s thoughts moves toward the opposite valence. That is, after a few minutes, participants induced to feel sad were having happy thoughts, whereas those put into a happy mood had relatively more sad thoughts. This homeostatic emotion regulation fits nicely with the current analysis: mood-congruent thoughts help people learn the lessons of their previous behavior, but adaptive future behavior requires that emotion regulation take place.”
Implications for CX professionals
Firstly, be relevant. Understand where and why you are seeking to create more of an emotional pull.
Secondly, do not create journey maps where every touchpoint is measured and required to score highly on an emotional criteria.
In measurement, do not seek to push for the highest scores (except if you are selling a Faberge egg or other hedonic experience). Instead, focus on the well-satisfied consumer, one that is engaged and comfortable with the experience, that maybe would score an 8 out of 10 but sees the cost-to-benefit dimension as positive and differentiating.
Hence, I focus on customer narrative as a measure. After all, if we looked at the supporting narratives, an 8 out of 10 at Aldi, is not the same as an 8 out of 10 at Waitrose and will have a different correlate to value.
Thirdly, imperfections are not necessarily a bad thing. In an encounter between two people we are establishing the ground rules of a relationship. This means shifts and turns that cannot be directed by the brand and sometimes require negative or no emotion as part of their enactment.
Myth 5: The functional is not emotional
Finally, I end with a view on functional actions.
Sure, most of the time downloading a web page, picking up a cup of coffee or making a phone call is a functional item. But we must always be aware that when the functional goes wrong what is hygienic can quickly turn into a moment of pain and stress. If a functional activity like getting a mobile signal fails when I’m in a rush, sure I’ll get emotional!
In addition, a functional activity may have potential to be invested with more emotional potentiality. For instance, look at how Amazon transformed a functional activity such as opening a box.
If it can be made meaningfully different and of value to the customer then it can be emotional.
Implications for CX professionals
Firstly, ensure that loss aversion is adequately covered and responded to.
Secondly, consider functional experiences in the light of their potential to be innovated for emotional response.
Link your CEM efforts to creating value for the customer. Which is not the same as emotional response.
However also search for the emotional need we obtain or could obtain 'from' the service or product received. After all:
Football entertains - more entertainment please
Car insurance is a pain - avoid the pain please
Retail could entertain in the right conditions - more wow please
But this still starts with what is practically meaningful to you in the value exchange not some abstract notion of emotion.
My view on AI
Where we need to interpret human 'emotion', we need to ingest and understand subjectivity. Multiple human judgers are best for this. After all. would you seriously ask a complex court case to be determined by AI?
The capability to interpret data, bootstrap yourself into the context of the defendant and apply subjective and objective tests can only be done by a trained legal representative: and note I deliberately use the word complex.
Millions of years of evolution has trained humans to understand the antecedents of emotion. Whereas most of 'the AI can measure emotion pitch’ is overhyped and dangerous.
This way will lead to catastrophic failure, simply because you strip out context.
So I am a contact centre rep being tasked to ignore my own capability in favour of a machine algorithm? If so I would warn against this.
In short, AI helps in certain instances: so define those instances.
Once again, we are left with 'context is king'.
Steven Walden's new book, Customer Experience Management Rebooted – Are You An Experience Brand Or An Efficiency Brand? is available now.
A sample customer review:“We are at a moment in time when technologies are at last enabling businesses to make the profound shift from doing things ‘at customers’ to doing them ‘with customers’. This doesn’t just impact Marketing and Sales but, rather, every function, real and virtual, in the demand chain. All of which makes Steven Walden’s book timely and important. Important because it breathes new life into the CX argument at a time when it was becoming fossilized. Important because it clearly identifies and explains the subjective nature of CX: “Customer experience is the experience the customer has”. Important because it clarifies and articulates the intimate connection between Customer Experience and Customer Value, and the management of both of those domains. It's densely written in places, but I think this book is terrific.” David Pinder