BellSouths Performance

1
BellSouths Performance Assurance Plan
Review Tennessee Workshop September 14, 2004
2
Outline

• Market Results of Performance Under Current
  Performance Assurance Plan.
• Aggregate Performance under Current SQM and SEEM
  plans.
• Performance by Domain (e. g. Provisioning,
  Maintenance and Repair) under Current SQM and
  SEEM plans.
• Detailed discussion on performance follows
  Overview

3
Overview

• Principal purpose of Performance Plan is to
  Facilitate Competition
• Monitoring
• Enforcement
• Monitoring- Current SQM
• Measures and Sub-metrics Defined
• CLEC specific and Statewide Data and Comparisons
  Produced
• Enforcement- Current SEEM
• FCC Complaints
• TRA Complaints
• Contract Dispute Resolution
• Courts
• SEEM Automatic Payments to deter backsliding
• Both plans based on what little was known in 1999
  to 2000

4
Overview

• FCC Defined Characteristics of an Effective
  Performance Assurance Plan
• Potential liability that provides a meaningful
  and significant incentive to comply with the
  designated performance standards
• Clearly-articulated, pre-determined measures and
  standards, which encompass a comprehensive range
  of carrier-to-carrier performance
• A reasonable structure that is designed to detect
  and sanction poor performance when it occurs
• A self-executing mechanism that does not leave
  the door open unreasonably to litigation and
  appeal
• Reasonable assurances that the reported data is
  accurate

5
Overview

• Robust Competition
• In BellSouths territory, 23 of all lines
  served by CLECs in TN, compared to 16
  Nationwide
• In BellSouths territory, 43 of business lines
  served by CLECs in TN
• Results due to
• TRA Initiatives
• BellSouth Performance
• No material service issues

6
CLEC Market Share in BellSouth territory in
Tennessee
7
Aggregate SQM Plan Performance

• Overall performance continues to be good.
• For period Sept 2003 through June 2004,
  performance standards met per current plan
  averaged 84.
• Lengthy discussion of performance issues confined
  to remaining 16.
• Performance level consistent with when 271
  approval granted.
• When BellSouth applied to FCC for 271 authority,
  data period May July 2002, performance
  standards met based on current plan averaged
  83.8.
• Performance in Tennessee is Consistent with
  levels in other BellSouth States.

8
Development of SQM and SEEM

• Current Plan, which is same as Florida, was
  agreed upon by BellSouth and CLECs and adopted by
  TRA in 2002.
• Current plan was reasonable approach given the
  information available at the time
• Had little experience and no frame of reference
  for what was needed
• Plan was actually based on what we knew in 1999
  and 2000
• Uncertainty and lack of knowledge created bias
  for granularity in measures and submetrics
• SEEM created when volumes low - so designed to
  produce high penalties at low volume
• Measure based structure untested
• Now facts and later information show that SQM and
  SEEM can and should be improved.

9
Evolution of SQM and SEEM

• Weve had several iterations for Florida SQM in
  terms of an effectiveness review (the model for
  the TN Plan)
• 1. Monitored SQM results in FL since 1999 (using
  Georgia template)
• 2. Commission approved interim SQM plan for 3PT
• 3. Had evidence and hearing on second FL plan in
  April 2001
• 4. Had periodic review of that plan in 2002-2003
• 5. Implemented changes resulting from that
  periodic review in 2003
• 6. Are now embarking on periodic review
• SEEM has not been revisited.

10
Performance Across Region Essentially The Same
Have Regional Processes Small Differences
driven principally by difference in plans
11
Aggregate SQM Plan Performance

• Data was not adjusted for several high profile
  FCC Triennial Review (TRO) areas such as
• No Pre-ordering bulk migration data
• No Non-coordinated hot-cut timeliness data
• No Bulk migration ordering data
• Line sharing still reflected in data
• No flexibility to address in current plan

12
Aggregate SQM Plan Performance

• 78,000 LSRs processed in typical month for TN.
• 54,600 Service orders installed in typical month
  in TN
• More than 650,000 CLEC lines in Tennessee
• 1902 sub-metrics means Data is Not Focused
• 401 Diagnostic or parity by design not
  meaningful
• Based on a review of data for January 2004
• 771 No activity not meaningful
• 730 Non-diagnostic submetrics with some activity,
  but
• 338 very low volume (1 30 transactions) - not
  meaningful
• Only 392 of the 1902 sub-metrics could
  potentially be useful to monitor performance
• At the time plans developed, uncertainty and lack
  of knowledge created bias toward large number of
  measures and sub-metrics
• About 20 of submetrics are potentially useful at
  the statewide level
• Fewer than 3 of submetrics are potentially
  useful at the CLEC level
• On 3rd revised version, but still based largely
  on conjecture

13
Aggregate SEEM Plan Performance

• No Material Service Issues
• No FCC complaints
• No TRA complaints
• No Contract Dispute Resolution
• No Court Cases
• SEEM payments are inconsistent with level of
  performance
• Paid 15.3M in TN over last 12 months however,
• Nondiscriminatory Service Has Been Consistently
  Delivered

14
Aggregate SEEM Plan Performance

• Competition is thriving 43 CLEC Business
  Market Share 84 of Sub-metrics met.
• No performance backsliding 83 submetrics met
  at 271 approval deemed non discriminatory, 84
  currently however,
• Experience has shown that
• FL/TN plan generates much higher payments for
  essentially the same performance than plans in
  the other BellSouth states.
• Amounts paid to CLECs for only small disparities
  can amount to several years of free service due
  to
• SEEM created when volumes were low - so the plan
  was designed to produce high penalties at low
  volumes
• SEEM designed to generate high payments even if
  non-discriminatory performance provided

15
Performance Across Region Essentially The Same
Have Regional Processes Small Differences
driven principally by difference in plans
16
SEEM plan payments under current plan for current
performance per 1000 CLEC lines in Service.
September 2003 - February 2004
SEEM Payments much higher in TN than any other
state despite essentially same performance
17
Performance is not the principal driver of
payments

• Performance across Region essentially the same.
• TN and FL payments much higher than other states
  due to measure-based SEEM structure.
• Transaction-based SEEM plans in other states
  produce more rational SEEM payments because
• Naturally scalable
• Penalty applies only to number of transactions
  that are out of parity
• Penalty inherently varies due to extent of
  failure
• Greater incentive to incrementally improve
  performance to avoid additional penalties
• Currently, used in at least 40 states, including
  FL for Verizon
• Greater incentive to do a good job on every task
  every time
• Penalty is directly proportional each month to
  how close or how far each transaction drives
  BellSouth to the performance standard
• SEEM still in its original version, but SQM on
  3rd revision

18
Performance by Domain Provisioning

• Measures the installation processes.
• 1148 SQM submetrics 378 SEEM submetrics
• UNE Fee schedule from 4750 to 14,250. (Monthly
  escalations over 6 months)
• Met timeliness standard on 99.8 of orders for
  Jan-Jul 2004
• Met accuracy standard on 97 of orders for Jan
  Jul 2004
• Achieved performance standard on 82 of SQM
  submetrics that had activity during Jan-Jul 2004
• Penalties of 5.9M during period July 2003 June
  2004.
• Percent Missed Appointment (PMI)
• 214,000 in UNE Tier-1 SEEM payments, even though
  less than 1 of the installation appointments
  were missed (i.e., over 99 of the appointments
  were met as noted above)

19
Performance by Domain Provisioning (contd)

• Provisioning Troubles within 30 days of Svc Order
  Completion (PT30)
• 800,000 paid for service installations that had
  trouble rates of 4 or less. (gt96 of orders
  installed without a trouble report).
• PT30 CLECs received payments for just one
  trouble reported in a given month for all
  circuits that were installed in the previous 30
  days.
• Under current plan, 1 missed appointment or 1
  provisioning trouble report can trigger a penalty
  equivalent to installation charges for 216 UNE-P
  or 63 UNE-L.
• Non-Recurring charge for UNE-P is 22 vs. 4750
  penalty
• Non-Recurring charge for UNE L (SL2) is 75 vs.
  4750 penalty
• Charge for service is the typical basis of
  payments under commercial arrangements.

20
Performance by Domain Maintenance Repair

• Measures the trouble clearing and maintenance
  processes.
• 255 SQM submetrics 180 SEEM submetrics
• Fee schedule from 4750 to 14,250
• 98 of lines did not experience a trouble.
  (Monthly escalations for 6 months.)
• For the 2 of lines that did have a trouble
  report (not always an actual trouble)
• Met timeliness standard for 98 of the reports
• Met quality standard for 89 of the reports
• Achieved performance standard on 87 of SQM
  submetrics
• Statistics too sensitive to small performance
  differences
• Penalties of 8.4M during period July 2003 to Jun
  2004

21
Performance summary Mtce Repair (contd)

• Customer Trouble Report Rate (CTRR)
• 1.06 million paid for UNEs Overall average
  CTRR was 2 (i.e., 98 trouble free service).
• CLECs received payments for just one trouble
  reported in a given month for all of their
  in-service base of circuits for a particular
  product. The amount paid was a minimum of 4,750
• Maintenance Average Duration (MAD)
• Over 158,000 in payments even though 93 of the
  MAD measurements indicate that BellSouth cleared
  the CLECs troubles more quickly than the
  comparable retail service.
• Under current plan, 1 trouble report can trigger
  a penalty equivalent to 17 years of the recurring
  rate for one UNE-P or 16 years for one UNE-L.
• Recurring charge for UNE-P is 23 vs. 4750
  penalty
• Recurring charge for UNE L (SL2) is 25 vs. 4750
  penalty
• Charge for service is the typical basis of
  payments under commercial arrangements.

22
Summary

• Robust competition
• Nondiscriminatory service levels
• Current plan is the best that all of us could do
  with information known at the time, but we now
  know more and can improve
• No material service quality issues, and
• No performance backsliding
• Level of SEEM payments is inconsistent with this
  good performance principally due to
• SEEM Structure,and
• Fee schedule
• Only small percentage of sub-metrics in SQM at
  statewide level or for an individual CLEC have
  sufficient activity to be potentially useful

23
Performance Data for Additional Domains
24
Performance by Domain Ordering

• Measures BellSouths processes for handling CLEC
  LSRs
• 358 SQM submetrics 255 SEEM submetrics
• Met timeliness standard on 98 of LSRs. (2 not
  met)
• 98 of LSRs without errors received FOC within
  required timeframe
• 98 of LSRs with errors were returned within
  required timeframe.
• Achieved performance standard on 75 of SQM
  submetrics
• The other 25 of the submetrics are primarily low
  volume measurements capturing the 2 of the LSRs
  where timeliness standard was not met.
• Penalties of 936,000 during July 2003 June
  2004

25
Performance by Domain Billing

• Measures BellSouths processes for CLEC billing
• 25 SQM submetrics 7 SEEM submetrics
• Met timeliness standard 100 met
• Billing accuracy of 99.7
• Achieved performance standard on 86 of SQM
  submetrics
• Missing invoice accuracy metrics with very small
  dollar values.
• Penalties of 125,000 during period July 2003
  June 2004

26
Performance by Domain OSS / Pre-Ordering

• Measures availability and compares performance of
  BellSouths OSS systems that serve CLECs.
• Measures OSS response timeliness
• 67 SQM submetrics 37 SEEM Submetrics (incl. Tier
  2)
• Availability during Jan 2004 July 2004 99.9
• 5.9 million responses. Average response time 3
  seconds
• Achieved performance standard on 82 of
  submetrics
• Remaining 18 missed, but differences in response
  time for BellSouth versus CLEC were typically
  only a few seconds or less.
• However one response type has difference of 10-20
  seconds, but less than 1 tenth of 1 of responses
  are of this type.
• Penalties of 22,000 during July 2003 June 2004

27
Performance by Domain All Other
Several of these domains and metrics are
information only. Change management misses
primarily due to low activity, where only
perfection will achieve the standard, and
tardiness in correcting cosmetic software errors.