The tools and techniques for retail analysis of
customer-level data are the same as the ones used for consumer analysis. The focus however
shifts from brands to outlets.
Analytics Suite
The range of commonly used analysis for customer-level
retail analytics includes:
- Customer (shopper) profile analysis
- Loyalty and propensity
- Assortment analysis
- Overlap analysis
- Outlet group analysis
- Outlet repertoire analysis
- Gain–loss analysis
- Trial and repeat visit analysis for new outlets
- Penetration and repeat rate
- Sales forecasting (relevant for new outlets)
In consumer analytics the focus is on brands. For instance, profile analysis
in consumer analytics pertains to consumer profile of brand users, whereas profile analysis
in retail analytics pertains to customer (shopper) profile at outlets, where depending on the
industry, outlets could be retail stores or bank branches or petrol stations etc.
Gain–loss in retail analytics examines outlet switching, whereas gain–loss in
consumer analytics examines brand switching.
Similarly, in retail analytics, loyalty is in relation to banners, or stores or
branches, and not brands.
Since the above-listed techniques have already been covered in Chapter
Consumer Panels and Consumer Analytics, this chapter
will not dwell on their methodology.
The case example,
Evaluation of Opening of New Petrol Station
, which comes later in this chapter, illustrates the application of some of these analyses.
Loyalty Panel Data and Consumer Panel Data
Data confined only to the retailer’s own transactions, as is the case with loyalty
panels, curtails the scope of the above-listed techniques. Analysis of such data is restricted in
context to the retailer’s customers and their transactions at the retailer’s outlets.
For example, penetration would be measured as the proportion of loyalty card
holders. Analysis of gain–loss will not capture gains from and losses to competing chains.
Forecasts would be confined within the boundaries of the retail chain. And metrics such as loyalty
and propensity cannot be computed. Both these measures require an assessment of customers’
transaction across the entire market.
Consumer panels which are prevalent in the FMCG sector, provide a holistic view of
the market and do not suffer the limitations of loyalty panels or CRM transaction data. They are
however expensive to set up and maintain, and their sample size would tend to be relatively small
compared to sales transaction data or loyalty panel data. So, while consumer panels can provide
for better diagnostics, their samples sizes may be inadequate for analysis at the granular level.
