|
| |
Measuring Customer Affinity - Part 2
By Peter Lavers
Design and Development Principles for Customer Affinity Measures
In my first Blog on this subject I proposed a model for measuring ‘Customer Affinity’, as follows:
-
Attitudinal Affinity
-
Level of positive sentiment or bonding to the brand/ organisation; usually assessed via external customer research
-
Behavioural Affinity
-
Level of positive interactions/ transactions demonstrated by the customer; achieved from internal data analysis/mining; dependent on the quality of the information and how well joined-up it is
-
Affinity Value
-
A measure of how valuable the customer is, however the organisation defines ‘value’ (not just monetary)
I suggested that modelling this would be a very useful tool for businesses seeking to optimise Customer Experience.
Here are some Design and Development principles to create such a measurement tool.
Design Principles:
-
Each dimension will probably consist of several factors, which will need to be either computed, attributed or inferred
-
The model works for B2C, B2B and indeed for the Public and Voluntary sectors – the factors that constitute the dimensions will vary significantly
-
Empirical values can be used in each factor, but rules-based mechanisms to derive ‘scores’ are preferable (and simpler to populate)
-
The model also lends itself to be included in segmentation (high score in all three = ‘best’ customers) – or indeed to drive it!
-
Behavioural and Value scoring is dependent on good internal data
-
Value factors could include revenue, profitability, life-time value, share of available spend/giving, but need not necessarily be purely monetary value (e.g. volunteering)
-
Behavioural factors could include frequency and recency of interaction/transaction, plus propensity if that is modelled
-
Recently researched Attitudinal data is unlikely to be available for every customer
-
Attitudinal factors can be sourced/supplemented from internal data (e.g. sentiment recording, permissioning, responsiveness) and from social media if monitored
-
Avoid unstable Attitudinal factors e.g. satisfaction, which can change quickly whereas net promoter (NPS) is more stable
-
Beware inferring Attitudinal data from Behavioural data – continuity does not equal loyalty!
-
Geodem (market) segments can be overlaid to help targeting
Development Principles;
-
Create a practical “working definition” of each dimension and factor
-
Identify what source data is immediately available; what needs consolidating & cleaning; what can be sourced; and what is currently unavailable
-
Populate for a statistically valid sample group to test & tune the scoring algorithms; drop insignificant/unreliable factors
-
Commission further research or data mining if required
-
Finalise and fully populate the model
-
Enable scores for each dimension to be recorded on your database
-
Map scoring into geographic and geodemographic databases to identify trends/gaps
-
Operationalise the scoring – incorporate in segmentation and develop propositions and treatments to grow each dimension
-
Sensibly develop and improve factors and dimensions over time
|
Peter Lavers, 03/04/2013 |
|
Blogs from the Archive
|
|
| | Measuring Customer Affinity – Part 1 | | There are lots of terms used in Marketing, Research and Customer Experience circles that try and describe just how much customers and prospects like or dislike our companies and brands. For the purposes of this blog I’m referring to this as ‘Customer Affi
More ...
| | Peter Lavers |
|
| CRM is not just about technology | | Many organisations both large and small will have a Customer Relationship Management (CRM) system of sorts, whether it’s an excel spread sheet or a fully blown Microsoft Dynamics system
More ...
| | Ben Tresham |
|
| Using the Power of Insight | | In the Business to Consumer (B2C) world many companies don’t have clear sight of their customers, but focus more on product sales, revenue, market share, productivity, quality, etc.
More ...
| | Ben Tresham |
|
| | |
|
| |
| | |