arrangements for porting phone numbers when customers

EXECUTIVE SUMMARY

Ofcom has proposed that a common database (CDB) solution be implemented and populated by 31 December 2008.

BT agrees with Ofcom that near instantaneous porting of mobile numbers cannot be achieved without a CDB. However, BT does not believe that this should be a regulatory imperative, based on the evidence presented to date. BT is also unconvinced that a recipient led porting process would be in consumers’ best interests until it can be demonstrated that satisfactory measures exist to prevent slamming, porting fraud and phone number theft through porting.

BT considers that setting a date now for shortening the mobile porting lead time in the manner suggested would be premature. The forthcoming wider consultation on transfers and migrations in a converging world, of which mobile number porting is but one element, will provide a better opportunity to consider this issue in the round including the identification of appropriate implementation dates.

BT believes that even with the most optimistic view of timelines, the dates for implementing a CDB proposed by Ofcom are unachievable. As a comparison, progress has been slow in relation to the more straightforward modification required to expand the existing Mobile Number Portability (MNP) system for new entrants Therefore, even if there were consensus on how to approach meeting Ofcom’s challenge, the task being set is simply too demanding. As yet, there is no decision making body, and inadequate consideration has been given to governance arrangements, potential commercial models and dispute resolution mechanisms. Additionally, the technical thinking, which is perhaps the most advanced, is not universally accepted. Consequently, BT believes that Ofcom should set a series of achievable milestones to reach its objective, setting a later end date, contingent upon hitting appropriate intermediate milestones.

Given the above uncertainties and the early stage of NGN migration, BT strongly believes that it is far too early to set a date in General Condition 18 for the mandatory use of the CDB, especially for fixed numbers. BT suggests that Ofcom consults, say at one or two year intervals, setting dates only once a satisfactory cost benefit analysis has been completed and once NGN roll out and CDB implementation plans are further advanced.

BT would be supportive in engaging with industry and Ofcom on a suitable rollout plan and the setting of intermediate milestones. The number portability regime has come a long way and the impact of this next very important step must be fully analysed prior to implementation. The outcome ultimately affects the consumer experience and we should collectively seek to ensure that the optimum solution is fit for purpose.

Introduction

1. In the UK, Number Portability (NP) is provided using the onward routing (OR) technical solution. Ofcom consulted in November 2006 on whether this should be progressively replaced by a common database solution (CDB). Following that consultation, Ofcom has decided that a common database should be developed, and, in this combined statement and consultation, Ofcom is consulting on time scales.

2. The current lead time for mobile NP (MNP) is five working days. In the November consultation, Ofcom consulted on shortening the lead time. In this statement and consultation, Ofcom is consulting on the modification of General Condition 18 regarding the time scale for reducing the lead time to two business days. It is also consulting on further reducing the MNP time scale to less than two hours following the introduction of direct routing.


Ofcom’s decision to implement a CDB and reduce the MNP time scale

3. BT notes Ofcom’s decision that a CDB solution for NP is required in the UK. Whilst BT understands why Ofcom is keen to ensure that an operator failure is handled more effectively for customers than the Atlantic failure, BT is of the view that the decision has been taken prematurely, before a cost benefit analysis justifies it, in particular for fixed numbers.

4. BT notes that Ofcom has introduced but not consulted upon additional market research and sensitivity analysis. It has also mandated the solution without explaining how good governance will be ensured in the timescales proposed and how the costs of a CDB will be apportioned, in particular given the uncertainty of benefits and, where they exist, their distribution (between customers, different industry sectors and specific companies within those sectors). BT believes that Ofcom has not sufficiently considered the overall commercial framework under which its proposals would be delivered, or the complex migration issues involved. Annex B sets out BT’s thinking on this in more detail. However, for the most part, this response will focus on the dates and issues on which Ofcom is primarily consulting.

Transition to a CDB solution

5. Whilst BT did not reach the same conclusions as Ofcom that the customer or industry benefits justified significant early expenditure on a CDB solution, and believed that investment would deliver greater consumer benefits elsewhere, particularly for the fixed industry, BT acknowledges that in a next generation network (NGN) world, this solution could have some advantages in both fixed and mobile networks.

6. BT recognises that a CDB solution is perhaps more pressing and more achievable in shorter time scales for 07 numbers than for other numbers. However, it is widely agreed that with increasing convergence, there is an associated probability that at some time in the future it is likely that a CDB should replace OR for fixed NP. The UK approach should allow for a single solution as opposed to one which militates against it.

7. BT would also wish to avoid a solution developed by the mobile industry alone that would not meet the fixed industry’s future needs. As such, BT welcomes Ofcom’s compromise that requires both industries to implement and populate a database at the same, or more or less the same, time. Requiring the fixed and mobile industries to collaborate on the solution allows for a common implementation but with a later mandate for use of the CDB by the fixed industry .

Time scale to implement and populate the database

8. BT understands that Ofcom is keen for MNP to be CDB-based by September 2009. As such, a milestone for implementing and populating a CDB of 31 December 2008 has been proposed, to be included in General Condition 18 (GC18). BT believes that this is too optimistic. Even with everything running as smoothly and efficiently as possible, BT does not believe it could be achieved before 31 March 2009 at the very earliest. Indeed, BT’s experience of seeking MNP on the existing solution where it has already taken more than two years would suggest that even this is very over-optimistic. BT suggests that Ofcom considers setting a series of achievable milestones that industry can support rather than an over-ambitious end date, as a number of stages are contingent upon the completion of a previous stage.

9. In order to facilitate the loading of mobile data in a timely fashion, Ofcom might consider setting different dates for populating the database with different number ranges, given for example that the CDB is unlikely to be widely used by the fixed industry as soon as by the mobile industry. This could take the form of a mandate to populate 07 numbers by Date A, with fixed numbers to be added by Date A + n months.

10. In terms of the initial proposed implementation of a CDB for MNP, BT understands from this consultation that this would only apply to 07 numbers and only implemented by Mobile Operators. Fixed network operators will only have the capability to directly route calls on the basis of porting data contained in the CDB once they have migrated to a NGN, even for calls from fixed numbers to mobile numbers. This will be coincident with their ability to implement for all fixed numbers. Therefore the earlier date would not be achievable for most fixed operators.

Governance

11. BT agrees with Ofcom that a cross-industry team is required to deliver the industry-wide database in the shortest possible time. Establishment of a clearly mandated body to progress the task is a critical activity, and all subsequent milestones should be predicated on this task completing, with dates being set relative to that completion date. Working-level teams need to be established by that body as soon as possible after it is established. These would need to include technical, governance, commercial framework, processes and transfer engineering. The whole of industry would need to be involved to agree a common approach and BT would be keen to be involved on all sub-teams, to help make the transition a success for customers and commercially equitable. It will be easier to set realistic time scales for completion once the initial thinking around all these areas has commenced and a degree of consensus emerges.

Mandating the use of the CDB for direct routing

12. Ofcom has explained why it wishes to mandate the use of the CDB for direct routing. However, the timing of any requirement is key. Ofcom recognises that in the fixed industry the requirement is inextricably linked to the migration to NGNs. Ofcom proposes that a CDB should be used to enable direct routing of all calls to ported numbers by December 2012. BT believes that it is too early in the NGN roll out to set a date with confidence for the mandatory use of the CDB, with different companies moving to NGNs at different speeds. Fixed NP should reflect NGN roll out rather than drive it. NGN rollout will be phased over a number of years, and enforced use of the CDB for routing before NGN rollout is complete would incur significant overheads.

13. BT believes that the task is far more complex than is implied in this consultation, as many factors interact. For BT, these include the gradual roll-out of 21CN exchanges (and similarly for other companies’ networks), process changes arising from Ofcom’s review of the wider migrations process, and how these might interact with home mover processes. It is imperative that the customer experience does not suffer as a result of companies trying to do too

many things at once, especially given how thinly resource would need to be spread to deliver these initiatives alongside business as usual initiatives, delivering better and cheaper services to customers, whilst maintaining and improving customer service.

14. As such, BT is of the view that Ofcom should review time scales on a regular basis, perhaps every one or two years, setting dates only when there is more certainty around their achievability. As such, BT does not support Ofcom’s Option A.

15. It appears that the requirement relates equally to number ranges such as 03, 05, 08 and 09. There are different issues with the implementation of these services, some of which are typically IN based. A detailed structured plan for different potential approaches for different number ranges would be worth considering as there may be some scope for earlier implementation for some non-geographic numbers (for example 055 which is largely used currently in corporately focussed NGN implementations).

16. Because a CDB solution may be achievable for MNP sooner than it is for fixed numbers, BT believes that it may be appropriate for a date to be set following this consultation. However, without confidence in the first milestone (31 December 2008), it is difficult to be confident in the September 2009 time scale. BT would suggest that if Ofcom retains a date for mandating the use of a CDB for MNP, that it should be framed by reference to a 9 month date from 31 December 2008, as/if revised.

17. Ofcom is silent on portability of Personal Numbers. BT is assuming that it was Ofcom’s intention that these should be treated in the same way as mobile numbers, but invites Ofcom to clarify this in its Statement.


Ported numbers or all numbers?

18. Ofcom has indicated that it is not mandating all numbers in use to be entered on to the database. BT notes however that there may be benefits in holding all numbers in the CDB, and that NICC is leaning this way. BT believes it would have been helpful had Ofcom set out and considered the merits of each approach. BT does not believe that the additional costs associated with including all numbers has yet been justified and agrees that mandating the population only of ported numbers on the CDB is the correct approach pending further consultation. This approach is consistent with Ofcom’s policy objective to only regulate to the extent necessary.

Shortened MNP lead times, and the proposed move to a recipient led process

19. Ofcom is consulting on changing MNP so that the lead time is reduced to less than two hours, and the process becomes recipient led.

20. BT has reviewed Ofcom’s fresh market research around MNP lead times and does not believe it supports the priority Ofcom is giving to reducing them. BT is not opposed to shortening lead times, but has two particular observations. Whilst Ofcom quotes shorter time scales in other countries, there is no supporting detail. In particular, there is no explanation of what activities each country uses to define the time line. In other words, when does the starting whistle blow? Have some activities been completed beforehand, whereas, within the UK, they form part of the activity within the current lead time, for example checks around identity verification? Before mandating a reduced porting lead time, or a move to a recipient led process, BT would urge Ofcom to commission an international comparison of best practice, to guide the UK, and to implement subsequently only those elements that would improve matters for UK consumers.

21. The two potential changes of practice should be considered independently as either on its own might be workable, but when they are combined, significant risks arise for consumers who might find their service ported or number “stolen” without their being aware. BT believes that shortened lead times may be achievable if the process remains donor led, as they have better information to check quickly to verify the porting request. Equally, a longer lead time might give a recipient more time to conduct proper checks.

22. BT’s key concern is that the UK should avoid customers and industry suffering from porting frauds and scams of differing types, for example number theft and slamming which would have a significantly greater negative impact on the consumer experience than the assumed positive impact that shortened porting periods would have. The UK has endured more than its fair share of malpractice in other product areas, such as with CPS and WLR, and BT would anticipate it spreading to porting if the opportunities were not excluded from the outset.

23. Due to the shorter time frame for the introduction of a CDB for MNP, BT believes that it may be appropriate to consider when a “near-instant” lead time could be introduced. However, in our view, since the evidence does not suggest that this should be a priority, BT would advocate reducing the risk of things going wrong for consumers by avoiding tying these proposals too tightly to the introduction of a CDB solution. The timing of introducing the CDB and revising the porting process should be staggered, so that if there is a problem with introducing the database, then this may be resolved before the MNP transaction is shortened and/or its basis changed, albeit that this delay may only need to be a small number of months. This lower risk approach appears proportional to any benefits likely to be achieved, and also allows lessons to be learnt from practice in other regimes.

Proposed modifications to GC 18

24. Given that Ofcom’s position as stated in the consultation, and at subsequent meetings, is that the regulatory requirement of a CDB is that it contain ported numbers, BT is surprised to see that that the definition of “Common Database” set out under the various options amending GC18 in Annex 8 refers to the Database “containing [information] in relation to each Telephone Number in active use in the UK ….”. BT assumes that this was simply a typographical omission by Ofcom and would suggest that this is modified to make it clear that the Database would contain details of every ported number from the UK National Numbering Scheme.

ANNEX A – ANSWERS TO OFCOM’S SPECIFIC CONSULTATION QUESTIONS

Question 1.

Ofcom has decided to require fixed and mobile providers to implement and populate a common database to enable direct routing of calls to ported numbers. Do you agree that providers should be required to achieve this by 31 December 2008?

BT does not share Ofcom’s view that there is an imperative for a CDB to be introduced for NP in the UK. The market justification for one at best is extremely marginal.

Given that Ofcom has taken the decision to proceed nonetheless, BT wishes to see it introduced in a manner that reduces uncertainty, is effective and delivers benefits to customers as soon as possible.

BT favours a milestone based approach and highlights the critical dependency on establishing clear governance arrangements without which no meaningful progress can be made by industry. Taking likely milestones into account, BT does not believe that the 31 December 2008 is achievable.

Question 2

When setting the deadline for implementing and populating the database, should Ofcom simultaneously set deadlines for using the database to deliver Direct Routing of calls to ported numbers? If so, would it be appropriate to require mobile operators to achieve Direct Routing of calls to ported mobile numbers by 1 September 2009 and require mobile and fixed operators to ensure Direct Routing of all other calls by 31 December 2012? Could this be done any earlier?

BT believes that insofar as there may be merit in setting a date for using the database for 07 numbers, Ofcom should desist from doing the same for other numbers. There is a greater degree of certainty around projected dates 2-3 years out than 5-6 years out.

Should Ofcom set a date for mandatory use of the database for MNP, BT believes that it would make more sense to accommodate the uncertainty around the date for implementing and populating the database by 31 December
2008 (Date A). BT would suggest that the date for mandating the use of the CDB for MNP should be set by reference to Date A, in the event it is either changed or drifts. BT suggests that the target for MNP should be Date A + 9 months.

The situation for fixed NP is very different. The key dependency is the rate of roll out of NGNs across the UK. BT believes that there is a considerable degree of uncertainty around this. As such, a date should only be set once roll out has advanced and that an achievable date is no more than 2-3 years away. Nonetheless, BT favours Ofcom’s approach that requires the fixed industry to implement and populate the database by an appropriate date, but that the date for mandatory use of the database should remain subject to further periodic consultations, say once every one or two years.

In relation to fixed numbers, BT suggests that Ofcom consider treating geographic and non-geographic numbers separately. The platforms supporting 01/02 (and 03) are very different from those supporting 05, which in turn are different to those supporting 08 and 09, and the implementation issues are likely to be different for each case.

Question 3

If you believe Ofcom should not set a deadline for deploying the database to deliver Direct Routing at this stage but should, instead, consult again during 2008, how could Ofcom and industry ensure that appropriate momentum is maintained such that Direct Routing is achieved at the earliest practicable date?

BT believes that the approach outlined in answer to Q2 above would maintain the momentum, as the fixed industry would be fully involved technically and commercially in the implementation and population of the database.

Very specific milestones can be used to maintain momentum, each referenced from the previous one. This is likely to be more effective than a blanket “Fixed and mobile” mandate which does not allow industry to prioritise and focus effort appropriately.

Question 4

Do you agree that, where a common database is in place and supporting Direct Routing of calls to ported numbers, changes could be implemented enabling (i) recipient led and (ii) near-instant (not longer than two hours) porting of mobile numbers at modest incremental cost proportionate to the benefits? Ofcom would welcome detailed views on the additional costs involved, including whether any additional costs would be incurred in ensuring that the database itself can support near-instant (not longer than two hours) recipient led mobile porting.

BT believes that a MNP lead time of less than two hours could not be achieved without a CDB. Ofcom seems to acknowledge this too.

In suggesting a two hour lead time, Ofcom has not been clear in clarifying which activities potentially fall within the two hours and which need to be completed before the two hours start. Without understanding better the challenge, it is not clear that a fit for purpose porting process with a two hour lead time is achievable, even with a CDB.

Ofcom also does not address how potential NP fraud, scams and mischief are avoided in other countries, and whether such techniques can be built-in within the two hours in the UK. BT believes that such problems are more likely to manifest themselves in recipient led porting, as less information will be available to the recipient, whose relationship with the customer may be minutes long, whereas, more information will be available to a donor, which will have provided service to, and billed, the customer for somewhat longer.

Given that in BT’s view the market research provided by Ofcom does not support the shortening of lead times as a priority for customers, BT would suggest that given the imminent consultation on transfers and migrations in a converging world, Ofcom wraps up its decision on this element of this consultation with the outcome of that one.

BT is providing Ofcom such cost information that it has in relation to reducing the MNP lead time separately, as requested by Ofcom.

Question 5

Do you support Ofcom’s approach to achieve industry agreement on effective governance of the new proposed number portability solution, as set out in paragraphs 4.74 to 4.78?”

BT agrees that a legal entity established by the industry would be required to bring a CDB to fruition. However, it is disappointing that Ofcom appears to have given so little steer on how such a body might be established, its remit, how it would be paid for - “paid for by the industry” (para 4.77) is too vague – and how disputes and non-co-operation would be handled. This issue is likely to be non-trivial, with fixed and mobile industries moving at different paces, and with different market incentives and network topologies within each grouping. BT believes that such a body would need to include the active involvement of a senior Ofcom manager throughout and that it would not be appropriate for Ofcom to set up a group and then disengage.

As stated previously, the creation of such a body and establishment of the commercial framework and relationships are critical to the success of the CDB proposal.

ANNEX B – ASSESSMENT OF COST BENEFIT ANALYSIS

1. This Annex assesses Annexes 4-6 of Ofcom’s Arrangements for porting numbers when customers switch supplier – A review of General Condition 18. These cover its Impact Assessment, Cost-Benefit Sensitivity Analysis and Further Consumer Research.

2. These annexes are designed to provide evidence to support Ofcom’s preferred approach, Option A. This requires industry to construct a central database by 31 December 2008 which is capable of supporting Direct Routing of calls to fixed and mobile ported numbers. It has specific dates for call routing and porting for mobile and fixed numbers between that date and 31 December 2012.

3. In its review of the cost-benefit analysis included in Ofcom’s November 2006 Consultation, BT highlighted five areas of concern with Option A. These were that costs were understated; benefits were likely to be lower than expected; the distribution of costs and benefits between fixed and mobile operators was highly unequal; funding of the central database could be difficult; and that, strategically, there was little evidence that a central database would improve consumer switching.

4. In BT’s view, the three annexes still do not address our concerns and therefore still provide no evidence to justify Ofcom’s preferred approach of Option A. Almost all the evidence they provide relates to mobile networks.

Level of Costs

5. In “Summary of Sensitivities” (p78), Ofcom presents six sensitivities of its preferred Option A. We do not feel that any of these address our specific cost concerns from the November consultation. Sensitivities 1 and 2 have the highest NPV, of £238m-£274m, but use a DCC of 0.8ppm, even though Sagentia stated in November 2006 that

“The Donor Conveyance Charge (DCC) is the mobile equivalent of the

APCC, the payment made to the donor operator for forwarding a call…The current rate used by the industry for settlement is 0.8p/min; however this figure has not changed in 6 years…We believe this figure is inappropriate to use when considering the industry as a whole, as it does not reflect the current costs of OR. Based on the capital equipment and operating costs figures we have received fromoperators, we have estimated a cost figure for the DCC of 0.1p per minute.”

6. In our view, the use of such a high DCC invalidates the resulting high NPVs.

7. Three other sensitivities (3, 4 and 6) have NPVs of + / - £11m. Over a discount period to 2018, extended to include more benefits, this is too close to zero / negative to make a compelling case for implementing a £75m capital cost investment.

8. Whilst Sensitivity 5 has a NPV of £68m, it is achieved only by extending the discount period by two years to allow more benefits, increasing the mobile porting rate from 2% to 5%, doubling the DCC to 0.2ppm to reflect the fact that the DCC (of 0.1ppm) represents 50% of the underlying costs of conveyance. The fact the discount rate has been increased from 7% to 12% has only a marginal negative impact.

Level of Benefits

9. Benefits arise from two broad sources – savings in routing costs and benefits from porting. Ofcom’s cost-benefit includes only the efficiency benefits from savings in routing costs. These are clearly very significant, as the “ceteris paribus” difference in NPV between Sensitivity 2 and Sensitivity 3 – where the only change is a reduction in the DCC from 0.8ppm to 0.1ppm – is circa £250m.

10. The benefits from porting to consumers (supposedly £296 in 2005 prices per ported customer) are not included in Ofcom’s cost-benefit. BT has previously disagreed with this level of benefits – nowadays there are so many low cost, low effort ways of notifying a number change, it is unlikely that the benefits of protecting against network failure are as great as the £296 per ported customer calculated by Sagentia. Also, by excluding such benefits from the cost benefit calculation that is used to help justify its policy proposal, Ofcom does not provide an empirical link with its “Further Consumer Research” – it is impossible to say how much the results from such research are worth in terms of discounted cash benefits.

11. By omitting to value such consumer benefits, Ofcom’s cost benefit appears to make the key justification for the policy proposal savings in routing costs, which could be internalised by network operators rather than benefit consumers.

Distribution of Costs and Benefits

12. The annexes do not deal with the issue of the unequal distribution of costs and benefits between fixed and mobile networks. According to the November 2006 figures, fixed networks would accrue discounted benefits worth £15.3m, compared to £193m for mobile networks; capital expenditure costs would be £61.5m for fixed networks, but only £12m for mobile networks.

13. The gross discounted benefit to fixed networks is therefore just 8% that of mobile networks. With three of the six sensitivity tests having NPVs of + / - £11m, and another test with a NPV of £68m, the implication over an already extended discount period is that fixed networks will experience no significant gross benefits and, when their much higher capital expenditure costs are taken into account (circa 84% of the total), very likely negative net benefits.

Funding

14. BT believes the fairest approach to funding a CDB would be that those who receive higher benefits make a proportionally higher contribution toward costs. However, at the capital expenditure level, fixed operators are expected to pay more than five times as much as mobile operators, but receive just 8% of the gross benefits.

Strategic Issues – Consumer Switching and Porting

15. Ofcom maintains that an investment in a CDB is justified because it will facilitate easy and swift consumer switching and enable the costs of operator failure to be avoided. Previously, BT suggested that neither of these claims had particularly strong empirical support – Ofcom’s consumer research indicated that only 10% of those who have switched fixed calls providers and 3% of mobile subscribers who have not switched supplier were dissatisfied with porting times.

16. The “Further Consumer Research” that forms Annex 6 does not provide any more compelling evidence that shorter porting times will increase switching. Ofcom finds that circa 70% of mobile subscribers have either never switched network or have done so more than four years ago, and that “there is little concern about portability amongst non-switchers”. According to its November 2006 research, “even when prompted with porting times as a reason for not switching, only 1 in 20 non-switchers cite this as a factor”. That same research also found that 8 in 10 of those who have ported their number were satisfied with the time the process took. We also note that Ofcom now reports that mobile porting has increased from 2% to 5% notwithstanding existing 5 day porting times – which again indicates that such times are not a barrier to switching.

17. Shorter porting times are expected to increase competition. However, with large numbers of both fixed and mobile network operators, MVNOs, WLR and CPS providers, and LLU operators it is questionable what the marginal value add will be on competition.

Conclusions

18. Ofcom justifies the need to implement Option A by the savings on routing costs and benefits in terms of portability. BT believes neither justification is well supported by the evidence in Annexes 4-6 of the current consultation, nor by the evidence in the November 2006 consultation. NPVs are essentially zero or negative under most realistic scenarios, particularly for fixed network operators. The other issues we highlighted in our response to the November 2006 consultation in terms of funding the central database when costs and benefits fall on different parties, and the weak support for shorter port lead times provided by consumer research all remain.

19. We continue to believe that telecoms will be better served by focused investment in fewer areas, like NGNs and NGAs, where the dynamic effects of innovation will lead to far higher returns than in a central database for number portability, the benefits for which are largely static (eg enabling firms to avoid the cost of changing stationery) and focused on a limited number of stakeholders (mainly mobile network operators).

BT would welcome any comments on the contents of this document. Comments should be addressed to Howard Erdunast by e-mail to howard.erdunast@bt.com.