ADVANCED ADDRESS QUALITY DIMENSIONS

1
ADVANCED ADDRESS QUALITY DIMENSIONS

• Joe Lubenow
• Lubenow and Associates
• Presentation to UPU DMAB
• 24 April 2007

2
ADDRESS QUALITY IN A NUTSHELL

• Phases and Stages
• Address Accuracy
• Move Update
• Address Presentation
• Address Standardization
• Chronology of Key Events
• Current and Upcoming Requirements
• Actual and Potential Incentives

3
PHASES OF ADDRESS QUALITY

• Address Acquisition
• Parsing Address Elements
• Address Standardization
• Address Matching
• Database Storage
• Address Data Transmission

4
Major USPS Systems and Products

• CASS (Coding Accuracy Support System)
• DPV (Delivery Point Validation)
• LACSlink (Locatable Address Conversion System)
• ACS (Address Change Service)
• OneCode ACS (using Intelligent Mail Barcode
  (IMB))
• NCOAlink (National Change of Address)
• MASS (Multiline Accuracy Support System)
• PARS (Postal Automated Redirection System)
• AEC I and II (Address Element Correction)
• SuiteLink (validate secondary identifiers for
  businesses)

5
Address Accuracy

• How well does address represent a delivery point
  description?
• Currently ZIP4 level gets best rates
• But an address can have multiple ZIP4 codes of
  varying precision
• This also means different barcodes
• Only a complete address can have full depth of
  code
• Default ZIP4 codes lack full depth of code
• Addresses with ZIP4 can be incomplete

6
Move Update

• Is the addressee located at the address?
• USPS has two types of systems to check this
• NCOAlink an example of pre-mailing system
• ACS an example of post-mailing system
• In First-Class a move update process is required
  since 1997
• Issue of database updating contentious
• In Standard Mail no move update process required
• This leads to unnecessary waste
• USPS to extend move update requirements to other
  classes

7
Address Presentation

• Current requirements only for automation rates
  (ZIP4)
• Only sufficiently complete address required
  
• Non-automation pieces have no such requirement
• Missing information may prevent or delay delivery
• Missing information may lead to inconsistent
  delivery
• USPS has recently proposed All Required Elements
  Present (AREP) as a higher standard
• By obtaining complete and correct addresses,
  which are also current in move update, this would
  resolve issue
• But how would USPS validate the process?

8
Address Standardization

• Address elements presented in proper form with
  standardized elements
• Together with AREP and proper element order, can
• mean addresses presented in canonical form
• AREP can be assured through DPV confirmation
• Proper element order can be assured by an address
  template
• USPS has a standard template for valid addresses
• This is part of the USPS work on international
  address
• standardization

9
1998 Cost Study

• Done by Price Waterhouse Coopers
• Updated in 2000 by extrapolation
• 17 of population moves each year
• Came up with 1.5 billion in Undeliverable As
  Addressed (UAA) costs
• This later updated to 1.8 billion
• Gives volumes and costs by mail class

10
MTAC Work Group 97

• Convened by Mailers Technical Advisory Committee
  (MTAC) in 2005
• First work group on AQ methodology
• Address by address, not just list by list
  approach
• Try for ready to go address hygiene
• Evaluate electronic address validation
• Perform AQ activities just in time before
  mailing

11
MTAC WORK GROUP 97BEST PRACTICES IN ADDRESS
QUALITY

• Report is available on USPS Web site
• Recommends a new approach to address quality
• Not line by line, but element by element
• lt1401 N Main Stgt has four address elements
• Assemble elements to create address block
• Store dates of address hygiene performance
• Vendors have yet to fully support this technology
• See Address Data Interchange Specification (ADIS)
  from IDEAlliance (www.idealliance.org) for an
  example
• This approach is used in UPU S42 international
  addressing standard

12
R2006-1 UAA AND PKR COST STUDY

• By Christensen Associates
• First study since 1998
• Studies Undeliverable As Addressed (UAA) Mail
• Studies Personal Knowledge Required (PKR) Mail
• Outgrowth of USPS/industry Product Redesign
• Based on 2004 (pre-PARS) environment
• Postal Automated Redirection System (PARS)
  currently applies to automated letter mail

13
R2006-1 UAA COST STUDY KEY POINTS

• USPS PMG Jack Potter in 2005 called for cutting
  UAA in half by 2010
• UAA is 10 billion pieces and 1.8 billion
• R2006-1 starts with 2004 pre-PARS environment
• PARS rollout calculated through 2008
• 2008 is R2006-1 test year
• 2008 is also date of full PARS implementation for
  letters

14
R2006-1 UAA COST STUDY KEY NUMBERS

• 2004
• Forwarding cost 21.3 cents
• Return cost 51.3 cents
• Wasted mail cost 4.4 cents
• 2008 (projected)
• Forwarding cost 18.2 cents
• Return cost 50.6 cents
• Wasted mail cost 5.3 cents

15
R2006-1 UAA COST STUDY BREAKDOWNOF COSTS

• 2004 PRE-PARS 2004 WITH PARS 2008 WITH PARS
• FORWARD 422 M 319 M 326 M
• RETURN 822 M 697 M 727 M
• WASTED 270 M 290 M 365 M
• CORRECTION 318 M 200 M 221 M
• SUBTOTAL 1832 M 1506 M 1639 M
• ADMIN/SUPP 24 M not est. not est.

16
R2006-1 UAA COST STUDY PKR MAIL

• What about PKR mail?
• We may think of this as incomplete addresses
• It is a likely cause of inconsistent delivery
• Mailing Industry Task Force (MITF) studies showed
  this as a serious problem
• Also it is not affected by PARS in its current
  form
• So the 2004 figures are not reduced by PARS
• The 2008 increased costs are not estimated

17
R2006-1 UAA COST STUDY PKR MAIL

• 2004 PKR pieces are 2,332 M
• 2004 PKR costs are 160 M, or 6.9 cents per piece
• This is based on sortation rejects, redirects
  within route and redirects within station
• But this may leave out some factors
• What about carrier delivery slowdowns?
• PKR mail will have different delivery point
  barcodes for high rise apartments and suites
• PKR mail may not delivery point sequence (DPS) to
  a contiguous location in the tray
• This could add to the PKR cost totals

18
APARTMENT AND SUITE NUMBERS

• What can the USPS do about missing apartment and
  suite numbers?
• SuiteLink can help for suites
• ApartmentLink is technically feasible
• ApartmentLink may cause privacy concerns
• So what can the USPS do about missing apartment
  numbers?
• We have contended that there are four main
  approaches to this issue
• Inform
• Invent
• Infer
• Incent

19
STRATEGIES FOR APARTMENT NUMBERS

• The USPS cannot be expected to inform mailers
  about apartment numbers due to privacy concerns
• The USPS cannot be expected to invent substitute
  virtual apartment numbers due to cost and
  complexity factors (remember the ill-fated Kelly
  plan in NYC in the 90s)
• USPS equipment could infer some apartment
  numbers, but not when name and address formats
  vary or when the apartment number is not in
  private databases
• Therefore it is time for the USPS to incent
  mailers to obtain and share available apartment
  numbers
• This can best be done by de-averaging the rates
  for complete and incomplete addresses

20
R2006-1 ADDRESS CORRECTIONS

• Manual corrections
• Now 75 cents, will be 50 cents
• Electronic corrections
• Now 21 cents
• Will be 6 cents (First-Class), 25 cents (other)
• Automated corrections (letters only)
• First-Class, first two per address free, after
  that 5 cents
• Standard Mail, first two 2 cents, then 15 cents
• IMB required for OneCode ACS

21
ADDRESS QUALITY AND THE IMB

• The Intelligent Mail Barcode is now available
• Used for OneCode ACS and OneCode Confirm, or just
  for sorting mail for delivery
• Combines Postnet and Planet code functions
• 31 characters of information in 65 bars
• Can identify mailer, mailpiece, and address
  uniquely
• Already being used for move update notification
• Could be used to receive address element
  corrections

22
AUGUST 2007 CASS REQUIREMENT

• CASS currently has a Delivery Point Validation
  (DPV) option
• DPV is becoming mandatory in August 2007
• National average 95 ZIP4, 91 DPV according to
  USPS
• Full DPV match requires complete and correct
  addresses
• But August 2007 requirement is only for correct
  addresses in terms of primary numbers, such as
  101 in lt101 Main Streetgt
• So incomplete addresses with missing apartment
  and suite numbers are still allowed

23
Future Requirements

• 2009 Requirement from Federal Register
• CASS for all mail classes
• Correct ZIP4
• Potential Secondary Identifier Requirement
• Move Update for Other Classes

24
List Certification

• MTAC certified list made up of certified
  addresses
• Need freshness dates on ZIP codes and move update
• Non-certified addresses could still be mailed,
  though not necessarily with the same results
• Minimum processing CASS, DPV, NCOAlink, LACS
• AREP required, standardization recommended
• AEC I and II used offline to resolve problem
  addresses
• How to get updates? OneCode ACS, Corrective
  Action Required List (CARL)
• All certified mailers could receive posted
  corrections by matching CARL

25
Address Validation

• Proposal use Secure Hash Algorithm (SHA-1) to
  code name/address in compressed form
• Includes freshness dates
• Digital signature would be generated
• Could use 2D barcode or electronic address file
• Unlike electronic mailing summary files, do not
  send, keep in escrow
• Question not whether address currently correct,
  but instead whether prior AQ hygiene performance
  was timely and carried through on the mailpiece

26
Address Quality IncentivesNSA, Niche, Discount
and Surcharge

• Negotiated Service Agreement (NSA) with USPS
• Requires approval of Postal Regulatory
  Commission (PRC)
• Provides discounts with provision for later
  extension to others similarly situated
• Niche classification allows multiple early
  adopters
• Discount and surcharge approach amounts to
  de-averaging of rates
• NSA and niche classification allow for monitoring
  of performance
• NSA and niche classification can foster
  innovation
• NSA and niche classification can be used to phase
  in new methods
• Discount and surcharge approach is suitable for
  activities that all mailers can perform
• Combination of approaches maximizes depth and
  breadth of impact

27
CONCLUDING PERSPECTIVE

• The highest levels of address quality needed to
  support the maximum system efficiency cannot be
  achieved through USPS technology alone
• The highest levels of address quality needed to
  support the maximum system efficiency cannot be
  achieved through mandated address quality
  regulation alone or in conjunction with USPS
  technology
• The highest levels of address quality needed to
  support the maximum system efficiency can only be
  achieved by combining technology and regulation
  with the cooperation of industry and the USPS
  through work sharing incentives