Retail 2.0 - A new paradigm for location specialists   Steven Feldman KnowWhere
Location still matters – or does it? “  The dream of electronic commerce has become a reality in today’s wired world,  where physical geography no longer presents a barrier to retailers or consumers  – and shopping has never been easier.  ” Nielsen
Agenda Web 2.0 Generations Impact of online shopping Bricks ‘n Clicks New demographics Impact on retail location strategies How will we find our customers of the future? Future data creation and analytical techniques
Web 2.0 Rich user experience User participation Dynamic content  Openness, freedom Collective intelligence “ creating network effects through an " architecture of participation ," and going beyond the page metaphor of Web 1.0 to deliver  rich user experience ” Tim O’Reilly
Generations
Why Web 2.0 matters to Retailers 16m UK households have internet access Online sales £55bn in last 12 months! Sales growth 25-30% 30-50  year olds lead online spending Generation Y  are the high value customers of the future Main impact  today  is in comparison goods but  tomorrow  ……?
Categories – who is buying online and where? Books & music Losing share to online Higher in remote rural areas Clothing & footwear Catalogues go online and upmarket Mix of rural areas and prosperous suburbs, urban professionals and areas of high ethnicity PCs/Games consoles & software Losing share rapidly 15-30s boys Asian Communities Household Goods & Appliances Frequently need delivery Online winning where large retailer not close Source CACI
Where is hurting? Source CACI A mixture of affluent suburbs, some prosperous urban areas and an affluent rural hinterland
OnLine vs InStore OnLine Price Convenience  Wider Range  Comparison Customer Reviews Product information InStore Want to evaluate Immediacy No shipping costs or waiting at home for delivery Trusted Brand Knowledgeable Sales people
OnLine vs InStore
Bricks ‘n clicks can beat online only Combining an online presence with physical outlets may offer the best option Leverage core competencies Supplier and distribution networks Brand equity Consumer trust Economies of scale
Bricks ‘n clicks can beat online only
Finding the e-consumer Target 30-50s May come from minority groups E-literate
 
Location still matters
RETAIL 2.0 A New Paradigm for Location Specialists Andy Thompson
The World of the Customer
Impact on Retail Location Strategies Here are three possible future states.... Retail catchment areas will be much smaller in size Customers will reject brand loyalty in preference to price/convenience trade-off Purchase behaviour will become more complex and transaction patterns less predictable We need to find or create new data and analytical techniques to understand the customers and markets of the future
WHERE 2.0 Electronic generation of mapping and location technology Location enabled mobile devices will become the norm in the future By 2012 at least 18% of  phones will be smartphones .
Crowd Sourced Data Been around for centuries! C19 th  Oxford English Dictionary  WEB 2.0 provides the structure for collaboration e.g. openstreetmap.org Debate around coverage, currency and  quality What about crowd sourced retail data ?
Pervasive Location Path Intelligence Tracking individuals using mobile phone signals Footpath TM  Technology Retail applications Rental valuations and tenancy mix Shopper flow management In centre marketing Behaviour analysis Copyright 2007 Path Intelligence Ltd
 
Pervasive Location Geo-location Services Yahoo Fire Eagle Mozilla Geode Google Geolocation API Don’t need a GPS service Detection of “where” people are interacting with the web Location intelligent Web 2.0 services Sales and marketing offers.... not new anymore.... Security (location verification) True “location-tailored” content Intelligent Near Field Communication services
Historical Data Analysis? “ When you rely on historical data, its much harder to tease out causation”  Ian Ayres, Super Crunchers - “ How Anything Can Be Predicted” “ Marketing is concerned with recent past and near future.... and human beings are remarkably consistent in their behaviour over short periods”  Sean Kelly, Customer Intelligence -  “From Data to Dialogue”
Random Sampling Sir Ronald Fisher  wikipedia.org/wiki/Ronald_Fisher Randomisation test Study of errors in data with a non-normal distribution Gold standard in medical research for testing success of treatments Why aren’t we using randomisation? We have the resources We can create new data
Randomisation in Action Divide your customers randomly into groups, treat them differently and measure the responses Sample size is key If the sample is large enough we can assume they are statistically identical Our intervention can then be measured purely - the  “treatment effect” Capital One – proactively intervenes in the market through randomised experiments Today  we  have data which has random characteristics but are we treating it as random data?
New Techniques in Predictive Modelling
Generalised Linear Model
Generalised Additive Model
GAMs in Performance Modelling
We need to act now! “ Statistics are like a lamp-post to a drunken man.... more for leaning on than illumination” Lets make data analytics the currency of our organisations and the first activity of decision making not the last!
Location still matters
Thank You Steven Feldman [email_address] http:// GIScussions .blogspot.com Slides at  http://www.slideshare.net/stevenfeldman   Andy Thompson [email_address] www.thewendovergroup.co.uk Slides at

Retail 2.0 A New Paradigm For Location Analysts

  • 1.
    Retail 2.0 -A new paradigm for location specialists Steven Feldman KnowWhere
  • 2.
    Location still matters– or does it? “ The dream of electronic commerce has become a reality in today’s wired world, where physical geography no longer presents a barrier to retailers or consumers – and shopping has never been easier.  ” Nielsen
  • 3.
    Agenda Web 2.0Generations Impact of online shopping Bricks ‘n Clicks New demographics Impact on retail location strategies How will we find our customers of the future? Future data creation and analytical techniques
  • 4.
    Web 2.0 Richuser experience User participation Dynamic content  Openness, freedom Collective intelligence “ creating network effects through an " architecture of participation ," and going beyond the page metaphor of Web 1.0 to deliver rich user experience ” Tim O’Reilly
  • 5.
  • 6.
    Why Web 2.0matters to Retailers 16m UK households have internet access Online sales £55bn in last 12 months! Sales growth 25-30% 30-50 year olds lead online spending Generation Y are the high value customers of the future Main impact today is in comparison goods but tomorrow ……?
  • 7.
    Categories – whois buying online and where? Books & music Losing share to online Higher in remote rural areas Clothing & footwear Catalogues go online and upmarket Mix of rural areas and prosperous suburbs, urban professionals and areas of high ethnicity PCs/Games consoles & software Losing share rapidly 15-30s boys Asian Communities Household Goods & Appliances Frequently need delivery Online winning where large retailer not close Source CACI
  • 8.
    Where is hurting?Source CACI A mixture of affluent suburbs, some prosperous urban areas and an affluent rural hinterland
  • 9.
    OnLine vs InStoreOnLine Price Convenience Wider Range Comparison Customer Reviews Product information InStore Want to evaluate Immediacy No shipping costs or waiting at home for delivery Trusted Brand Knowledgeable Sales people
  • 10.
  • 11.
    Bricks ‘n clickscan beat online only Combining an online presence with physical outlets may offer the best option Leverage core competencies Supplier and distribution networks Brand equity Consumer trust Economies of scale
  • 12.
    Bricks ‘n clickscan beat online only
  • 13.
    Finding the e-consumerTarget 30-50s May come from minority groups E-literate
  • 14.
  • 15.
  • 16.
    RETAIL 2.0 ANew Paradigm for Location Specialists Andy Thompson
  • 17.
    The World ofthe Customer
  • 18.
    Impact on RetailLocation Strategies Here are three possible future states.... Retail catchment areas will be much smaller in size Customers will reject brand loyalty in preference to price/convenience trade-off Purchase behaviour will become more complex and transaction patterns less predictable We need to find or create new data and analytical techniques to understand the customers and markets of the future
  • 19.
    WHERE 2.0 Electronicgeneration of mapping and location technology Location enabled mobile devices will become the norm in the future By 2012 at least 18% of phones will be smartphones .
  • 20.
    Crowd Sourced DataBeen around for centuries! C19 th Oxford English Dictionary WEB 2.0 provides the structure for collaboration e.g. openstreetmap.org Debate around coverage, currency and quality What about crowd sourced retail data ?
  • 21.
    Pervasive Location PathIntelligence Tracking individuals using mobile phone signals Footpath TM Technology Retail applications Rental valuations and tenancy mix Shopper flow management In centre marketing Behaviour analysis Copyright 2007 Path Intelligence Ltd
  • 22.
  • 23.
    Pervasive Location Geo-locationServices Yahoo Fire Eagle Mozilla Geode Google Geolocation API Don’t need a GPS service Detection of “where” people are interacting with the web Location intelligent Web 2.0 services Sales and marketing offers.... not new anymore.... Security (location verification) True “location-tailored” content Intelligent Near Field Communication services
  • 24.
    Historical Data Analysis?“ When you rely on historical data, its much harder to tease out causation” Ian Ayres, Super Crunchers - “ How Anything Can Be Predicted” “ Marketing is concerned with recent past and near future.... and human beings are remarkably consistent in their behaviour over short periods” Sean Kelly, Customer Intelligence - “From Data to Dialogue”
  • 25.
    Random Sampling SirRonald Fisher wikipedia.org/wiki/Ronald_Fisher Randomisation test Study of errors in data with a non-normal distribution Gold standard in medical research for testing success of treatments Why aren’t we using randomisation? We have the resources We can create new data
  • 26.
    Randomisation in ActionDivide your customers randomly into groups, treat them differently and measure the responses Sample size is key If the sample is large enough we can assume they are statistically identical Our intervention can then be measured purely - the “treatment effect” Capital One – proactively intervenes in the market through randomised experiments Today we have data which has random characteristics but are we treating it as random data?
  • 27.
    New Techniques inPredictive Modelling
  • 28.
  • 29.
  • 30.
  • 31.
    We need toact now! “ Statistics are like a lamp-post to a drunken man.... more for leaning on than illumination” Lets make data analytics the currency of our organisations and the first activity of decision making not the last!
  • 32.
  • 33.
    Thank You StevenFeldman [email_address] http:// GIScussions .blogspot.com Slides at http://www.slideshare.net/stevenfeldman Andy Thompson [email_address] www.thewendovergroup.co.uk Slides at