Mass Personalization

1
Lecture 3
Mass Personalization
Dongsoo Han ICU
Electronic Commerce Fall 2000
2
What is Personalization?

• Addressing customers by name and remembering
  their preferences
• Empowering the customer. Examples Lands End,
  llbean
• Showing customers specific content based on who
  they are and their past behavior
• Product tailoring. Example dell.com
• Connecting to a human being when necessary
• Allowing visitors to customize a site for their
  specific purposes
• Users are 20-25 more likely to return to a site
  that they tailored (Jupiter Communications, Inc.)

3
Web Personalization

• Creates relationship
• Human tendency to show loyalty to the familiar
  (McKinsey)
• Stickiness
• Tendency of customers to return
• Tendency to stay longer (shelf space in the
  brain)
• Competitive necessity
• Effective use of customers time
• Increased cleverness convenience

4
Empowering the User

• Users have tremendous leverage on the Web
• Leave your site freely
• Compare prices cheaply
• Exchange information quickly
• Gómez.com, epinions.com
• How can business recover the advantage?
• Personalize! If everyone is treated differently,
  power cannot be concentrated

5
The Secret Know the User

• IP address, e.g. 192.151.11.40. Look it up.
• Anonymous, but I might know your employer
• Domain name, e.g. hp.com
• I probably know your employer
• Name, address, phone no.
• A good start
• Social security number
• I know everything

6
Prime Personalization Candidates

• Companies with
• Many products/services
• Complex products/services
• Many customers
• Competitive environment
• Industries
• Newspapers/Magazines/Research
• Catalogs/Retail
• High Tech
• Financial Services

7
Cookies

• Scratchpad memory for the web (typically 4KB)
• Small files maintained on users hard disk,
  readable only by the site that created them (up
  to 20 per site)
• Internet Explorer keeps them in \windows\Cookies
• Netscape keeps them in a file cookies.txt in the
  Netscape directory
• Used for
• website tracking, online ordering, targeted
  adverts
• Can be disabled
• Visit Cookie Central
• We have no privacy left anyway. See Anonymizer

8
DoubleClicks Cookie
on my laptop!
idd75ae834doubleclick.net/014689387523158341357021
488029320845
9
How DoubleClick Works
Merchant Cookie
Client
1. Client requests a page
Merchant Server e.g. Altavista
DoubleClick Cookie
2. Server sends a page with a DoubleClick URL
3. Text is displayed
4. Client requests the DoubleClick page
Web Page
5. DoubleClick reads its cookie
DoubleClick Server
If you choose to give u personal information
via the Internet that we or our business partners
may need -- to correspond with you, process an
order or provide you with a subscription, for
example -- it is our intent to let you know how
we will use such information. If you tell us that
you do not wish to have this information used as
a basis for further contact with you, we will
respect your wishes. We do keep track of the
domains from which people visit us. We analyze
this data for trends and statistics, and then we
discard it.
6. DoubleClick decides which ads to send
10
Personalization Techniques

• Content targeting rules
• Show all announcements whose subject contains
  mortgage to anyone whose DwellingStatus is
  HomeOwner
• User segmentation rules (cluster the audience)
• Include anyone whose PastPurchases in the last
  12 months is above 10,000 in group BigSpenders"
• Behavior-based profiling rules
• When a person views Home Equity Loan Information
  set DwellingStatus to HomeOwner

11
Filtering Techniques

• Rule-based filtering
• Ask user questions to elicit preferences,
  adaptive sequencing
• Personalogic
• Learning agents (nonintrusive personalization)
• implicit profiling
• webgroove.com
• Collaborative filtering
• base decisions on preferences of like-minded
  users
• moviecritic.com
• amazon.com

12
Clickstream Analysis

• Determine distinct visitors
• Determine repeat visits
• Effectiveness of marketing campaigns
• Path to revenue generation
• Popularity of different sections of the site
• Understand when and where people leave the site
• ROI on marketing and advertising expenditures

13
Clickstream Analysis Examples

• MatchLogic www.matchlogic.com
• Andromedia www.andromedia.com
• E.piphany www.epiphany.com
• Broadvision www.broadvision.com
• Personify www.personify.com
• net.Genesis www.netgen.com
• Accrue Software www.accrue.com

14
Server Log Analysis

• Servers maintain logs of all resource requests
• remotehost name authuser date "request" status
  bytes
• gateway.iso.com - - 10/MAY/1999001030 "GET
  /class.html HTTP/1.1" 200 10000
• Referrer logs
• 08/02/99, 120235,http//ink.yahoo.com/bin/quer
  y?p"samplelogfile"b21hc0hs0,
  130.132.232.48,
  biomed.med.yale.edu
• Analog (Cambridge)

DATE
REFERRING QUERY
REQUESTING IP ADDRESS
REQUESTING DOMAIN
15
Broadvision
SOURCE BROADVISION
16
Personalization Roadblocks
SOURCE FORRESTER RESEARCH (12/98)
17
Personalization Pitfalls

• Only ask for information you need
• Never ask for information before you need it
• Respect the privacy of your customers
• Do not underestimate response
• Be prepared for sales credit issues
• Be aware of scalability issues

18
Personalization Tools

• Blue Martini www.bluemartini.com
• Art Technology Group www.atg.com
• e.piphany www.epiphany.com
• HNC www.ehnc.com
• GuestTrack www.guesttrack.com
• Net Perceptions www.netperceptions.com
• Manna (FrontMinds) www.mannainc.com
• Engage www.engage.com