Consumer Analytics, Loyalty Analytics
Click on any of the following solutions for demonstrations of the wide range of analysis that fall under consumer analytics.
- % Buyers
- Consumption: Volume per buyer
- Demographic profile
- Loyalty profile
- Solus/core/non-core
- Basket
- Cross basket analysis
- Gain-loss
- Trial
- Purchase Frequency
- Repeat decay
- Overlap analysis
- Forecasting: Market Share = Trial × RBR × Buying Index
Consumer analytics is one of the dazzling facets of marketing science. Pertaining to the analysis of continuous, individual/household level (customer level) behavioural data, it yields insights that are particularly useful in identifying and addressing business issues.
The data may be sourced from online or offline consumer transactions or interactions, or from managed sources such as consumer panels or loyalty panels.
These structured data sources are ideal for diagnosing the buying behaviour of products and services, especially where repeat purchasing is the norm, through a wide repertoire of analytic metrics and techniques including width and depth of purchase, buyer groups, profile analysis, behavioural brand loyalty, trial and repeat purchase, overlap analysis, basket analysis, gain–loss and forecasting of new product sales. The metrics use analytic algorithms that filter and drill into the data, to yield insights pertaining to consumers buying behaviour.
DBS Black - Launch Evaluation
A fictitious case example assessing the launch of the DBS Black Card using transaction data.
Penetration
DBS Black achieves a 21% trial rate in its first year. This is a reflection of the new card’s success at its first moment of truth. An important measure for success, the trial rate must be viewed in conjunction with repeat usage rate and usage index to accurately forecast the contribution of the new card.
Overlap
The new card boosts the penetration of DBS Cards. 13.5% of the new card’s users are duplicators, i.e., they also use other DBS cards. The remaining 7.5% are new DBS card customers.
Repeat Usage Rate
Repeat Usage Rate, the second moment of truth, measures the propensity of trialists to continue using the card. It is the parameter that usually has the greatest bearing on the success of the new product.
Share Prediction
Based on penetration, repeat usage rate and use index, the model predicts that DBS Black Card will contribute about 18% of the usage (in $ terms) of DBS cards.
Gain-Loss: Source of Growth
Gain loss reveals that DBS Black has resulted in the growth of the DBS Cards portfolio.
A little over half the source of growth of DBS Black is due to increase in spend by these customer. A further 20.7% of DBS Black card holders’ spend is emanating from new or lapsed DBS card customers.
Gain-Loss: Cannibalization
Most of DBS Black’s gains through cannibalization are emanating from DBS Altitude and DBS Platinum Card.
DBS Black - Launch Evaluation
DBS Black is expected to contribute about 18% of the usage, in terms of dollars spent, of DBS cards.
It is driving the growth of the DBS Cards portfolio. A little over half the source of growth of DBS Black is due to increase in spend by these customer. A further 20.7% of DBS Black card holders’ spend is emanating from new or lapsed DBS card customers.
Example (disguised): DBS Black.
The fictitious DBS Black case example exhibited above, illustrates a range of analytic tools used for the analysis and validation of new products. Metrics such as trial, repeat usage rate and buying index are some key indicators of the success of a new product, based on which analysts are able to forecast market share. In this example, since the data source is assumed to be the bank’s transactional data, it provides for an estimate of the contribution the new card is likely to achieve within DBS’ portfolio of credit and debit cards.
Of considerable interest too is the new product’s impact on the bank’s total business, which requires answers to questions such as the ones listed here:
- How much incremental business is the new card likely to generate?
- To what extent does it cannibalize other DBS card products?
The gain-loss analysis, which reveals the amount of business each product is gaining from its competitors, is particularly useful in providing answers to the above questions. In addition, techniques such as overlap analysis provide a measure of the level of duplication within a group of products.
Brand managers are ultimately interested in enhancing their brand’s performance. They would be keen to execute actions that get more customers to try their new offerings, and get existing customers to use them more often. The behavioural diagnostics derived from transactional data analysis, combine well with information from other sources on attitudes, perceptions and beliefs to yield the insights that these managers are seeking.
For a more detailed demonstration of the diagnostic capabilities of consumer analytics, refer to the techniques, examples and cases covered in the chapter Consumer Analytics and Consumer Panels from the Marketing Analytics guidebook.
Analytics reveals what consumer do, which may indeed be indicative of what they think. However to clearly comprehend their thoughts, we need to rely on other forms of research, and in particular, qualitative and quantitative research.
Consumer behaviour may also be inferred or diagnosed by means of machine learning techniques, data mining, crowd sourcing, optimization techniques, visualization techniques and sophisticated cognitive systems. These techniques work well with unstructured data sets that cannot be processed or analysed by means of the conventional methods that apply to structured data. Such data sets which are collectively referred to today as big data, require new ways of data handling and analysis. Prominent among these technologies are Hadoop and Hadoop-related projects, cloud computing, and cognitive systems.
In case you are interested, these analytic techniques are covered in the chapter Consumer Analytics and Big Data from the Marketing Analytics eGuide.
Benefits
Benefits
Customer behavioural data is the marketer’s gold mine that offers numerous advantages:
- The potency of the data is largely due to its complexion — continuous, disaggregate. Disaggregate data is far more revealing than aggregate data; therefore better suited for diagnosing consumer behaviour.
- Size. These databases tend to be large enabling the data analyst to drill down into proportionately small customer segments. In particular, transaction data and big data allow for micro segments, or even personalized marketing (e.g. targeted promotions and communications).
- Though not always representative of total market, the data is highly actionable. It allows for the execution of tactics at the customer segment level and at location — i.e. store or bank or website, etc.
- Provides deeper understanding of consumers’ buying behaviour. Whereas aggregate data for instance, can compute metrics like market share, disaggregate data goes a step further revealing breadth and depth of consumption, brand loyalty, switching behaviour, purchase baskets and so on.
- Reveals not only current and historical behaviour, it can also predict future behaviour. For example, techniques based on disaggregate data can accurately forecast market share of a new product soon after its launch.
- Estimates growth as well as the sources of growth. (Disaggregate data reveals switching patterns through gain–loss analysis).
- Uncovers the impact of marketing activities in the current time period as well as in the future. Reveals the long term impact of some of these activities.
- Reveals most valuable customers.
- The union of people’s physical and the digital world via data integration, yields valuable insights that help better target consumers.
- Accurate and complete information provides for greater transparency and accountability.
Contact
Ashok Charan,
Marketing Analytics Practitioner’s Guide,
Digital Marketing,
Marketing Mind,
Destiny Marketing Simulator,
ashok.charan@gmail.com
