The Timings of a Voice of the Customer Intervention
Does it matter at what point you collect the Voice of the Customer?
In my last post I discussed the relevance of systematically gathering and putting into context the Voice of the Customer. In this post, I want to discuss whether it matters when (in relation to an experience) you gather customer feedback. In our view – it does.
The most advanced (and systematic) Voice of the Customer programmes employ two types of studies:
Transactional studies are typically employed shortly or immediately after an interaction. The purpose of these (when employed appropriately) is a focus on tangible, direct and thus actionable feedback directly to the front line, which enables both service recovery if required and operational improvements.
Relational studies are targeted towards a representative sample of your customer base (again, if employed appropriately). Here, focus is not on recent experiences, but on overall perceptions of an organisation, which is formed by the sum of experiences with a provider, direct and indirect communications and perceptions of product and price.
It is immediately obvious that the content to be covered in these surveys will be different. More detailed and in-depth around a specific experience in transactional, more focus on broader experiences and perceptions in the other.
Nevertheless, as researchers and customer experience professionals in client organisations alike, we are often faced with the dilemma of seeing different scores (or trends) for similar metrics across various studies.
The typical (and often correct) explanations of this phenomenon are:
- Difference in sample make up (e.g. representative customer sample vs. customers who contact)
- Differences in methodologies (e.g. CATI vs online vs mobile)
- Framing effects (e.g. where within the survey flow is a question positioned)
- The impact of time between an experience and the survey (memory fade).
All of these reasons are functions of research methodology. As such, the existence of different scores and trends is often attributed to limitations of the research, which – consciously or subconsciously – tends to undermine stakeholders’ trust in customer experience measurement outcomes.
We believe, however, the differences we measure in different surveys are more than just research artefacts – they are a true reflection of how customers remember. Failure to appreciate this may limit the value derived from Voice of the Customer programmes.
The two concepts that help explain this human behaviour are the ‘Peak End Rule’ and the ‘Experiencing self vs the Remembering self’.*
The concept of experiencing vs remembering self essentially says that human perceptions of a situation differ depending on the time passed between the experience and the time of asking. The experiencing self provides feedback on a situation based on ‘right now’ which means that experiences impact on what customers tell us in an immediate manner.
The remembering self, one the other hand, forms perceptions on the basis of an ‘average’ memory of an experience. This ‘average’, in turn, is influenced by two elements of an experience: The peak (positive or negative) and the end (again, positive or negative).
This all sounds very complicated, so let’s try and make it more relevant with an example: You are on a call with a service provider. You were on hold for five minutes (which you don’t think is that bad in some sectors) until you spoke to an advisor. This advisor tells you that some of the information provided to you previously was incorrect, which does annoy you quite a bit. They then tell you that you need to speak to a colleague to get the issue resolved, which means you end up in another queue. The new person you speak to resolves your problem and apologises for the inconvenience caused. They are very understanding and really take their time to get everything sorted.
Collecting feedback from the remembering self at different stages during this experience, would likely result in a graph a little like this:
It is clear that feedback is provided immediately, without taking into account elements of the experience that happened before, or that will happen afterward.
The remembering self, in contrast, won’t consider the call experience as a succession of individual experiences. It will think about the call as a whole, and form its perceptions on that basis – but with the important limitation that it will likely weigh the end of the call and the highest peak or trough more strongly than other experiential elements.
There are some clear implications for any Voice of the Customer programme:
Feedback collected from the experiencing self via transactional studies provides greater detail on specific experiences, and thus enables detailed root cause analysis at customer touchpoints. It also enables identification of negative experiences, which can feed into service recovery, thus creating a positive end to an experience. This explains the service paradox that sometimes, when things go wrong but are handled well, customers end up more positive than if nothing had gone wrong in the first place.
We believe that it is the remembering self that influences future behaviour. As such, relationship surveys are designed to speak to this part of ourselves, as this will allow us to understand (and statistically model) the impact of remembered experiences on claimed future behaviour. Transactional surveys help us to manage out the extreme experiential peaks and troughs.
The consequence from this should be to analyse feedback gathered through different surveys in their own right. This is as much a stakeholder management task as it is a programme design task. We believe that through a deeper understanding of their customers, as outlined in this post, the credibility of customer experience measurement programmes and thus their impact can be maximised throughout an organisation.
*These concepts are explained in more detail in Kahneman’s book “Thinking, Fast and Slow” – a foundational book for the area of behavioural economics.
For more customer experience insight visit the KPMG Nunwood CEM blog.