Network Reinvention Flashcards
What factors have generated retail turbulence and market opportunity?
Diversification - e.g. supermarket banks Consumer power Deregulation Competition Globalisation Disintermediation New entrants - start-ups competing head to head with established players New delivery channels/tech.
What are the 3 main limitations to SIMs?
- Consumption factors
- Provision factors
- Technology factors
What consumption factors do SIMs ignore?
Demographic segmentation Work-based trips Leisure trips Low value and convenience purchases Elastic demand
What provision factors fo SIMs ignore?
Product quality Price Brands Customer loyalty Scale economies Footfall and micro-location
What technology factors do SIMs ignore?
Internet competition e.g. Blockbuster vs. Netflix
Distribution channels e.g. e-commerce
Home delivery
What are the alternatives to SIMs?
- Analogue models
- Regression models - in retail R2 can’t be
- Network segmentation
- Site rating models
What is network segmentation?
Grouping of stores based on similar characteristics - inspired by geodemographics
What characteristics can stores be grouped by in network segmentation?
Location
Competition
Demographic factors
Site factors
What was Birkin et al.’s (2002) scorecard rating system?
Chose variables based on network segmentation analysis: Urbanisation Attractiveness Retail centre quality Demand
If a store achieved a score > average segment score then OVERACHIEVING
If a store achieved a score below average segment score then UNDERACHIEVING
What is the equation for the predicted scorecard result?
Segment average volume x (Branch rating/Average segment rating)
What are Site Rating Models?
Similar to network segmentation, but used in things other than food retail e.g. petrol stations
Scorecard model based on range of factors - location and facility components
e.g. ‘ATM Ratings model’ used: demographics, retail and business activity, workplace data etc. to derive a score for ATM locations - best score = best location
What are Analogue models?
Attempts to forecast potential sales by drawing comparisons with other stores that are similar in physical, locational and trade-area circumstances
I.e. draws ‘analogies’ between an existing and the proposed store
What is Regression modelling in retail analysis?
Defines a dependent variable e.g. store turnover, and attempts to correlate this with a set of explanatory variables e.g. store size
What are the limitations of regression models in retail?
- Evaluates stores in isolation - doesn’t consider impacts of competition or store’s existing network
- Problem of ‘heterogeneity of sample stores’ - how easy is it to find stores sharing similar characteristics?
- Assumes X variables are independent and uncorrelated - in reality floorspace and parking are likely to be correlated
- Fails to handle spatial interactions - doesn’t model the process that generates flows of revenue between demand zones and stores
When are Site Rating Models preferred to SIMs?
Non-food retail
E.g. petrol stations, ATMs, post offices etc. because they don’t work on the same principles of distance and attractiveness
Used in highly complex markets where SIMs are likely to break down
What are the advantages to Site Rating Models over SIMs?
Simpler, more robust and cheaper
BUT
Less accurate and less transparent
When are Analogue models used?
Most common - used in store location and store formatting
Store formatting: deciding which type of store is most suitable based on levels of competition, size of market etc. e.g. Tesco Express vs. Tesco Extra
What are the limitations to Analogue models?
- Success is very dependent on the analysts’ ability to choose a store that is similar enough to draw analogies
- Performance of store from which analogies are being drawn may be inaccurate e.g. what if it is over or under-performing?
- Difficult to evaluate greenfield sites using this method as catchment may be distorted by local transport networks and hard to import revenue predictions from other towns and cities
When are Regression models used?
Single store analysis and to assess entire store networks
What are the strengths of Regression models?
Simple
Robust
More sophisticated than basic analogue models
Why is planning the location of ATMs difficult? Because of this, what is the best location method to use?
Complicated because they are frequently visited but irregularly used by a variety of customers at a variety of locations
Site Ratings Model
