Marketing planning with true customer insight
Insight is successful when it enables your business to make better decisions. We start the insight development process by learning from you how and where a stronger insight knowledge base will yield greatest value
By focusing on how the insight deliverables can be used inside your business, we aim to deliver what you need, not what we think you might want
Typical insight deliverables include:
- Sales performance models that take account of the full on-line and off-line ecosystem
- Market segmentations that provide the basis for longer term planning and benchmarking
- Individual risk models for the insurance industry
- Propensity models for response and value
- Customer value analysis to explain where value is coming from, and how best to develop it
Sales performance models
The purpose of a sales performance model is to explain, in the context of all your marketing activities, why sales vary. Additionally they explain the causality of sales through your on-line channel, much of which is driven by off-line activity, as well as the impact of off-line on your conventional channels.
Our models take account of your media spend activity, your pricing, and where available competitive behaviour.
Market segmentation is a key planning tool when you want to differentiate your products, propositions, and communications to different sectors of an overall market.
It is also of use when calculating market share, and allocating resource such as sales or communications.
There is no set methodology for building a market segmentation; our starting point is the factors that are important to your business in understanding how consumers or businesses behave. These can be the traditional consumer life-stage and affluence dimensions, but they may also be how businesses adopt to new technology, or what they need in their particular industrial environment.
We have developed segmentations for financial services companies, brewers, hotel chains, bookstores, fast food outlets, and soft drinks manufacturers.
Individual risk models
These are normally developed for insurance companies wanting to forecast risk at an individual level based on an external data source; for instance understanding the risk represented by individual bank customers.
The development approach usually partitions customers that have or have not claimed, or had different underwriting results, into separate groups, and then looks at how the underlying data, whether it is lifestyle, demographic, or transactional, differentiates them.
The output is normally a ‘gains chart’ which shows the expected risk performance for different groups.
Propensity models for response and value
A response or value propensity model is built to rank a customer base, or prospect list, by their likelihood to respond to an offer, or their expected value if they do.
Response models can be built at many different levels from product level propensity to purchase, to likely response to a particular marketing campaign.
In every case the model has a target variable, e.g. responders, or customer value, and predictor variables, obtained from an underlying customer or prospect base.
The role of the model is to use the predictor variables to give the best estimate of an individual’s likely behaviour.
Customer value analysis
Many organisations hold extensive transactional data without going the next step to analyse trends in customer value.
A customer value analysis can be used to identify:
– whom to recruit
– when to cross-sell
– what to cross sell
– when to undertake anti-attrition exercise
At Berry Thompson we have developed customer value analyses for furniture retailers, hotel chains, insurance companies and a fast food outlets; in every case we have found great differences in individual customer value and hence opportunities to enhance it.
See some of our case studies