Issued by the Director General of Telecommunications

30 July 2001

Contents

Summary

Chapter 1 Introduction
Background
Objectives and scope
Market definition
The regulatory environment

Chapter 2 Setting the scene
Introduction
Internet penetration
Means of Internet access
Significance of dial-up
Summary

Chapter 3 Retail Internet service provision market
Introduction
Effective competition indicators
Consumer outcomes
Consumer behaviour
Supplier behaviour
Market structure
Conclusions
Implications for regulation

Chapter 4 Wholesale call origination market
Introduction
Effective competition indicators
Market structure
Consumer outcomes and behaviour
Conclusions
Implications for regulation

Chapter 5 Wholesale Internet call termination market
Introduction
Effective competition indicators
Consumer outcomes and behaviour
Supplier behaviour
Market structure
Conclusions
Implications for regulation

Chapter 6 Next steps
Details of the consultation process
Summary of questions

Annex A Market definition

Annex B FRIACO

Annex C Effective competition indicators

Annex D Glossary

Summary

S.1 This consultation document presents the findings of Oftel’s review of effective competition in dial-up Internet access. We have assessed the extent to which competition is providing UK consumers with the best deal in quality, choice and value for money. Our aim is a regulatory regime that is appropriate to the extent of competition in the market.

S.2 This review looks at one retail and two wholesale markets. The retail market is dial-up PSTN Internet access (unmetered and metered), including ISDN. Two wholesale products needed by Internet Service Providers (ISPs) to provide retail Internet packages are call origination and Internet call termination.

S.3 Residential and business dial-up Internet consumers are getting a good deal. Oftel considers that the retail market is effectively competitive because:

• UK prices compare with the best in the world;

• there is a wide range of packages;

• the UK is the first market in Europe to offer consumers competitive unmetered packages;

• UK consumers are generally satisfied with the quality of service they receive; and

• low barriers to entry, the plethora of tariff structure and options and the large number of suppliers are all indicative of an absence of supernormal profits and pricing above the competitive level.

S.4 Oftel considers that BT is dominant in wholesale call origination:

• the wholesale call origination market is not effectively competitive. BT has a large market share, estimated at over 80%; and

• there is some competition in infrastructure but continued network charge controls and the availability of quality wholesale products such as FRIACO (see note 1), are essential to maintain competition. Oftel proposes in this consultation that no additional regulation is required.

S.5 Wholesale Internet call termination appears to be an effectively competitive market. However the nature of the market is changing and this has the potential to affect effective competition:

• wholesale Internet call termination appears to be a more diverse market than call origination in terms of market shares;

• barriers to entry are less than for call origination and wholesale prices for a wide range of Internet call termination products appear to be at a competitive level. However, the cost of providing Internet call termination is changing:

• demand for unmetered Internet access is leading to a renewed requirement for extensive network infrastructure;

• the use of IP networks and associated wholesale dial-up Internet services introduced by BT (BT’s ‘dial IP’ products) may suggest that the nature of the wholesale Internet call termination market is changing;

• an IP interconnection product might be needed to enable other operators to compete with BT’s ‘dial IP’ products and achieve greater levels of network efficiency.

This consultation seeks views on these findings.

S.6 If, as a result of this consultation, Oftel concludes that effective competition exists in one or more of the three markets, Oftel may remove any specific regulation. If Oftel concludes that one or more of the markets is not effectively competitive, Oftel may consider whether further regulation is necessary to promote competition.

S.7 Chapter 1 of this document sets out the objectives and scope of this review and defines one retail and two wholesale markets.

S.8 Chapter 2 overviews the UK retail Internet market. It considers levels of Internet penetration, the extent of Internet traffic and the significance of dial-up Internet access in the UK.

S.9 Chapters 3, 4 and 5 measure effective competition in each of the three markets against the four groups of effective competition indicators: Consumer Outcomes, Consumer Behaviour, Supplier Behaviour and Market Structure.

S.10 Chapter 6 contains details of the consultation process and a summary of the questions Oftel is asking stakeholders.

S.11 The annexes to the main consultation document contain further details of the market definition, Oftel’s effective competition indicators and a glossary of terms.

S.12 A further wholesale product needed by ISPs is Internet connectivity. Oftel is undertaking a separate review of the Internet connectivity market and will be consulting on its initial findings later this year.

Note to summary

1 Flat Rate Internet Access Call Origination (see Annex B)

Chapter 1

Introduction

Background

1.1 Oftel’s strategy is to match the level of regulation to the level of competition in the market, removing formal regulation as competition develops, while continuing to protect and inform consumers where the market cannot do so. Oftel carries out market reviews to assess the level of competition in a particular sector and to ensure that regulation is appropriate.

1.2 In October 2000, Oftel issued a kick-off statement, setting out the objectives and scope for the review. This document invited stakeholders to submit data and views that would allow Oftel a fuller understanding of the Internet market. Oftel has since discussed views on the market with various industry and consumer bodies, representing a wide spectrum of stakeholders.

Objectives and Scope

1.3 This review considers effective competition in dial-up Internet access and the benefits enjoyed by residential and business consumers in the UK.

1.4 It measures competition in three distinct markets:

• retail Internet service provision;

• wholesale call origination; and

• wholesale Internet call termination.

1.5 Oftel sets out proposals on the assessment of competition in each of the three markets and considers the appropriate level of regulation.

1.6 Oftel invites views on its assessment of effective competition and levels of regulation.

Effective competition indicators

1.7 The review measures effective competition against a range of indicators under four headings:

• consumer outcomes;

• consumer benefits;

• supplier behaviour; and

• market structure.

1.8 These indicators for effective competition are detailed in Annex B. Information on Oftel’s approach can be found in Implementing Oftel’s Strategy: Effective Competition Review Guidelines.

Market definition

1.9 The review is concerned with dial-up access to the Internet, also known as narrowband access. For the purposes of this review, dial-up/narrowband Internet access is defined as being up to, and including, 128kbit/s (see Annex A). Most residential consumers and SMEs in the UK use a PSTN dial-up connection (Public Switched Telephony Network) or an ISDN dial-up connection (Integrated Services Digital Network) to access the Internet.

1.10 A detailed analysis of market definitions is provided in Annex A of this report.

Arriving at the definition

1.11 Competition analysis identifies a retail market and considers the hypothesis that in a market served by a sole provider, it could be profitable to raise the price to the monopolistic level, well above the competitive level. The exercise explores the extent to which demand and supply-side substitution would occur were the monopolist to raise the price. If consumers and/or producers are able to switch to alternative products in response to the price increase, these products would be substitutes and provide a constraint on the price of the former product. The market will be broadened accordingly.

1.12 The market definitions in this review should not be interpreted rigidly. The Internet industry is changing rapidly. Definitions must be flexible and although the analysis is about the present, likely future developments are considered. All definitions should be read as being valid only at present. Defining the market enables the assessment of the state of competition in the relevant market, and should not be regarded as an aim in itself.

Three markets

1.13 This review looks at one retail and two wholesale markets. A further wholesale product, needed by Internet Service Providers (ISPs) to provide retail Internet access packages, is Internet connectivity. Oftel is undertaking a separate review of the Internet connectivity market and will be consulting on its initial findings later this year.

Retail market (Chapter 3)

1.14 The retail market is the market for dial-up PSTN Internet access (metered and unmetered), including ISDN. There are two main components to the ‘Internet access product’:

• network service – access to the network and the conveyance of calls by a telecommunications company.

• Internet service provision – linking the customer to the Internet and services such as email, content and billing. Service provision can be offered free or through a subscription, normally a monthly fee.

1.15 These two parts of the retail market can be sold separately or together. The consumer must purchase both to obtain Internet access.

Figure 1.1: Three markets considered by this review: two wholesale and one retail.

Source: Oftel

Wholesale markets

1.16 Two wholesale products needed by Internet Service Providers (ISPs) to provide retail Internet access calls on the retail market are:

• call origination; and

• Internet call termination.

Call origination (see Chapter 4)

1.17 Call origination is provided by the network operator with access to the customer. Originating operators provide customers with access to the network for fixed telephony and Internet access. Telephony and Internet call origination markets are, therefore, the same.

Internet Call termination (see Chapter 5)

1.18 Internet call termination is the link between the originating network and the ISP. This normally includes a combination of PSTN and IP networks. ISPs generally buy this part of the call from the terminating network operator, who buys the originating part of the call from the originating network operator.

The regulatory environment

1.19 Oftel’s goal is the best deal for consumers in terms of quality, choice and value for money by promoting effective and sustainable competition.

1.20 Oftel’s strategy provides a framework to ensure appropriate regulation is achieved, adjusted to the level of competition in the market and focused on the area of concern. Too much, or too little regulation can work against consumers’ interests by deterring investment and innovation.

1.21 Oftel’s concern is access to the Internet not the Internet itself. Oftel’s role is to remove barriers to competition in the Internet access market.

1.22 Every firm running a telecommunications system needs a licence, including some ISPs. Some firms are also subject to rules of interconnection. A number of self-regulatory initiatives are in place to manage certain aspects of Internet access and service provision. The following sections describe the regulatory regime.

Retail Internet service provision

1.23 Oftel has taken a number of steps to promote competition in wholesale services, notably on Number Translation Services (NTS) and FRIACO. Both have facilitated greater competition in retail Internet service provision. NTS allows a non-geographic number to be translated into a geographic one and routed to a geographic location. The NTS revenue model enables ISPs, terminating operators and originating operators to share revenues. Oftel’s directions on Flat Rate Internet Access Call Origination (FRIACO) are discussed below in the context of wholesale call origination (1.25).

1.24 A number of organisations promote self-regulation over formal regulation:

• The Internet Service Providers Association – ISPA

ISPA represents Internet Service Providers in the UK. ISPA has around 100 members, including around 90% of the dial-up ISP market. Members commit to ISPA’s Code of Practice, which covers areas such as fair trading, terms and conditions, illegal material on the Internet, data protection and privacy, and complaints procedures. ISPA’s objective is to "promote competition within the UK ISP industry, with the aim of bringing long-term benefits to consumers."

• The Internet Watch Foundation – IWF

The IWF was established in 1996 by UK ISPs to tackle illegal and harmful material on the Internet. The IWF advises consumers on how they can restrict access to harmful and offensive content on the Internet. All ISPA members co-operate with the scheme.

• Nominet UK

The huge growth in Internet and e-commerce has led to increasing interest in domain names (web addresses). Nominet UK is the national registry for all Internet domain names ending in .uk. It derives its authority from the UK Internet industry and is a not-for-profit company, charging only for its running costs.

Nominet is working on a Code of Practice for the voluntary self-regulation of Domain Name registration practices amongst UK Internet Service Providers to enable consumers to make clear, informed choices over the provision of Domain Name registration services.

Wholesale call origination

1.25 The majority of UK customers continue to access telecommunications services from their BT line. Oftel regulations promote competition and protect consumers in this market, including wholesale and retail price controls. Reflecting this, in May 2000 and February 2001, Oftel made Directions to require BT to offer wholesale Flat Rate Internet Access Call Origination (FRIACO). This enables other operators to offer flat rate products in competition with BT, increasing the range of wholesale products available to network operators and boosting competition in the provision of retail unmetered access. As a result, consumers can get unmetered Internet access more cheaply and easily than before. Some aspects of the origination market are pursued in a co-regulatory way – industry discussions on FRIACO issues, facilitated by Oftel, have included both network operators and ISPs. However, issues of market power make significant self-regulatory initiatives unlikely.

Wholesale Internet call termination

1.26 There are few measures in place to promote competition in Internet call termination, although measures in wholesale call origination, such as FRIACO, have an impact on the market. Certain obligations apply to BT, particularly an obligation to publish charges and not unduly to discriminate in its provision of wholesale call termination services. BT argues that the obligations in its licence should be relaxed in the light of increasing competitiveness in wholesale call termination. This issue is described further in Chapter 5 and views are sought on the appropriateness of such measures.

Do you agree with Oftel’s definition of the three markets making up dial-up Internet access provision: retail Internet service provision; wholesale call origination; and wholesale Internet call termination? (See Market Definition, Annex A)

Chapter 2

Setting the scene – overview of the UK retail dial-up Internet market

Introduction

2.1 This chapter considers the UK retail Internet market and the context in which competition will be measured. Throughout this review, small businesses are defined as firms with 1-50 employees and medium businesses as firms with 51-500 employees.

Internet penetration

2.2 The UK Internet market is growing rapidly. 40% of households and 60% of all small businesses are now connected to the Internet. The amount of Internet traffic continues to climb steadily. Internet traffic now accounts for a third of all call minutes originating on the PSTN.

Figure 2.1 UK Internet Growth

Source: National Statistics, Oftel

Residential

2.3 Oftel’s residential survey shows that 40% of UK households were online in May 2001. This accounts for 10 million Internet households, an increase of 3.75 million in a year.

Figure 2.2: Percentage of UK households online

Source: Oftel residential surveys

2.4 Figure 2.3 shows that the proportion of the population with Internet access in the UK is higher than in most of Europe (though still behind Scandinavia and the U.S.).

Figure 2.3 Internet access in Europe – percentage of adult population with Internet access

Source: Proactive International: Pan European Internet Monitor Q1 2001

2.5 Higher income and younger households are far more likely to be connected to the Internet. However, 19% of older consumers (55+) claimed to be online at home in May 2001.

Figure 2.4: Make up of Internet households

Source: Oftel residential surveys

2.6 In February 2001, a significant proportion of residential consumers said they remained unlikely to get connected to the Internet. Cheaper calls, subscription and equipment costs, along with better provision of information about the potential uses and benefits of the Internet, would encourage 15% to get access sooner. A further 36% of homes said they did not want or need the Internet at home. This is comparable to the US where 41% of households were online at the end of 2000 (see note 2), and 30% have said they have no desire to connect (see note 3).

Business

2.7 In February 2001, 60% of all SMEs were connected to the Internet. 92% of medium businesses were connected and a further 5% were in the process of connecting or likely to connect in the next 12 months – the number having almost reached a ceiling. The proportion of small businesses online was lower at 59%, but with a further 15% likely to connect soon.

Figure 2.5: % UK SMEs with Internet access

Source: Oftel quarterly business survey

2.8 20% of UK businesses said they were unlikely to connect to the Internet within the next twelve months. This included 25% of small businesses.

Figure 2.6: Internet penetration amongst UK businesses

Source: Oftel Q4 business survey, February 2001

Means of Internet access

Residential

2.9 Figure 2.6 shows that the majority of online households uses traditional PSTN dial-up access (81%). A further 10% of residential consumers access the Internet via ISDN.

Figure 2.7: Methods used by consumers to access the Internet at home

Source: Oftel Q4 residential survey, May 2001

2.10 A growing number of households claim to be accessing the Internet via digital TV (4%) and a further 2% say they use ADSL.

Business

2.11 For small businesses, PSTN dial-up remains the primary mechanism for access to the Internet (64%) with 25% using ISDN dial-up.

Figure 2.8: Access to the Internet by SMEs

Source: Oftel Q4 business survey, February 2001

2.12 A slightly higher proportion of medium businesses was using ISDN (53%) than PSTN/dial-up (38%). ISDN gives businesses a dedicated line with speeds of up to 128 kbit/s or twice 64 kbit/s. Substitutability of dial-up and ISDN seems to be high and for the purposes of this review they are considered to be in the same market.

2.13 Another 18% of medium businesses use leased lines to access the Internet. This provides them with a fixed, permanent connection to the Internet supplying high data rates. Leased lines are not considered to be in the same market as dial-up PSTN and ISDN, as the price of leased lines is considerably higher and as such, leased lines are used only when the rate of usage is high enough or when there is a need for particular applications that require a permanent connection.

Significance of dial-up

2.14 The importance of dial-up Internet access is crucial. Analysts and market research widely predict that dial-up access will remain the dominant method of connecting to the Internet among residential consumer and small businesses for the foreseeable future. Broadband access will be attractive for some users and some applications. However, for the average Internet consumer, dial-up access may well continue to be sufficient for managing an email account and surfing the Internet for a few hours a week. Unmetered packages make the continuing role of dial-up more attractive for relatively high users of the Internet, and for occasional users, pay-as-you-go (pence per minute/metered) packages will remain the best option. Notwithstanding developments in broadband access, therefore, dial-up consumers will continue to make up a substantial share of the Internet service provision market.

Summary

2.15 Internet traffic is growing. There is a slight shift towards connection among lower income groups in the UK and smaller businesses are steadily catching up with levels of connectivity among larger SMEs. On the other hand, a significant proportion of residential consumers and small businesses has no interest in getting connected to the Internet.

2.16 More households and businesses continue to use dial-up Internet access. The majority of connections to the Internet is currently via dial-up and is likely to stay so for some time.

Notes to chapter 2

2 Trends in Telephone Service, Federal Communications Commission, December 2000
3 Cahners In-Stat Group 18 April 2001

Chapter 3

Retail Internet service provision market

Introduction

3.1 As described in Annex A, the retail dial-up Internet access product consists of:

• network service – access to the network and the conveyance of calls by a communications company and

• Internet service provision – linking the customer to the Internet and services such as email, content and billing. Service provision can be offered free or through subscription, normally consisting of a monthly fee.

3.2 These two parts of the retail market can be sold separately or together. However, the consumer must purchase both in order to obtain Internet access.

Effective competition indicators

3.3 This chapter considers the provision of retail dial-up Internet services to consumers and how effectively competitive the market is, using the four group of indicators shown in Annex B.

Consumer outcomes

Price

3.4 The impact of effective competition is that UK consumers enjoy the best or near best deal in comparison with consumers in similar economies. Oftel’s international benchmarking of dial-up Internet access covers the UK, France, Germany, Sweden, and the US states of Ohio and California.

Residential

3.5 The June 2001 benchmarking results found that for residential users, the price of up to 40 hours per month dial-up Internet access in the UK at both peak and off-peak times is the cheapest of the five countries considered.

3.6 For residential unmetered access, prices in the UK were similar to the US, but much cheaper than those in Germany. France and Sweden do not have any unmetered packages making prices for high users in these countries much more expensive.

Figure 3.1: Residential Dial-Up Internet Access Prices

Source: Oftel benchmarking, May 2001 (prices as at February 2001)

Business

3.7 For business consumers of PSTN dial-up access, Oftel’s international benchmarking study showed there was little difference in the overall prices for metered packages, though Germany and Sweden were slightly cheaper than in the UK. Prices for business unmetered packages in the UK were significantly lower than in Germany and the US. Unmetered packages were not available to business users in France and Sweden.

3.8 These results are for PSTN dial-up access only. While prices for business users have fallen in the UK and are now broadly comparable to competitor countries, there are fewer PSTN dial-up packages designed especially for businesses, particularly outside the UK. This may reflect marketing strategies to encourage greater take-up of alternative technologies such as ISDN dial-up.

3.9 A significant proportion of UK businesses is using ISDN dial-up. The latest Oftel study shows that ISDN prices for business in the UK were significantly higher than in Europe. However, Oftel research shows that may small and medium businesses find that access packages designed primarily for residential users, are adequately meeting their needs.

Figure 3.2: Business Dial-Up Internet Access Prices

Source: Oftel benchmarking, May 2001 (prices as at February 2001)

Range of services

3.10 A wide range of services available to consumers is indicative of a competitive market.

Choice of ISP package

3.11 There are four main types of ISP packages available to consumers in the UK:

• Metered

• (‘pay-as-you-go’) 1) Subscription + cost of calls whilst online 2) Free subscription + costs of calls whilst online

• Partly unmetered 3) Monthly/annual subscription + free usage at certain times/for a certain amount of time (eg off-peak or 12 hours a month) + call costs at other times

• Fully unmetered 4) Subscription + unlimited free usage + no call costs

Residential

3.12 The number of residential users subscribing to unmetered access packages has grown substantially over the past year. In May 2001, 35% of Internet homes used some type of unmetered access, up from 25% in August and 16% in May 2000, accounting for over 3 million unmetered subscribers in the UK. Use of fully unmetered packages has made up most of the recent growth, reflecting the increase in availability and awareness of 24/7 flat rate offers.

Figure 3.3: Type of package used at home, 2000

Source: Oftel quarterly residential survey

3.13 The take-up of unmetered access has been at the expense of the more traditional, pay-as-you-go means of access. Use of subscription-free access has fallen most significantly within the last 12 months (from 66% to 44%).

Business

3.14 Use of unmetered access among SMEs has remained fairly consistent over recent months with a quarter using either a fully or partly unmetered package. Movement in businesses’ use of subscription-free packages is most likely the result of the different profiles of the businesses interviewed, in terms of how long they have been online and the package they have chosen.

3.15 Indeed a larger proportion of medium businesses (the more mature users) said they used some form of unmetered access in February 2001 (40% compared to 25% of small businesses). Subscription-free ISPs were the most popular packages for small businesses.

Figure 3.4: Main type of Internet package used by businesses

Source: Oftel quarterly business survey

Consumer choice of ISP

3.16 ISPA estimates more than 400 ISPs operate in the UK.

Residential consumers

3.17 UK consumers use a wide range of ISPs and packages. One in five Internet homes use more than one ISP (1 in 3 in May 2000). This is likely to be the result of consumers of moving to unmetered access with less flexible contracts. It is also indicative of a more mature market.

Business consumers

3.18 In February 2001, 25% of SMEs used more than one ISP. Businesses take into account cost, speed of access and content when making decisions about which ISP to use.

Consumer satisfaction

3.19 Consumer satisfaction with quality of service is one indicator of a competitive market.

Residential

3.20 In May 2001, 90% Internet users were satisfied with the overall quality of their home Internet service – a figure broadly in line with the fixed telephone and mobile markets.

3.21 Customer satisfaction with various aspects of Internet service in November 2000 is shown below. 90% of consumers were satisfied with quality of service offered by ISPs and 88% were satisfied with their subscription charges. 77% were satisfied with the cost of calls and the customer care service offered by ISPs. The lowest number, only 65% were satisfied with the speed of service.

Figure 3.5: Customer satisfaction with Internet service

Source: Oftel Q3 residential survey, November 2000

3.22 Satisfaction with quality of service, speed of service and customer care were slightly higher amongst subscription-paying customers than those using subscription-free or unmetered services. Satisfaction with call costs and subscription charges was highest among those on unmetered packages. Younger consumers and those in larger households were least satisfied with the speed of service.

3.23 77% of consumers were satisfied with the number of ISPs available. 70% were satisfied with the range of quality of service and 68% were satisfied with the types of connection available. 63% were satisfied with the choice of packages on offer and the range of customer care. Between 12% – 19% did not know how satisfied they were.

3.24 Satisfaction was lowest among residential Internet users for choice of speeds. Which?Online’s Annual Internet Survey found that, when asking consumers if they could improve the Internet in one way what would they choose, the top three choices were to make searches simpler (10%), to reduce costs (12%) and to improve speed (31%).

Figure 3.6: Satisfaction with level of choice in the Internet market

Source: Oftel Q3 residential survey, November 2000

Business

3.25 The majority of SMEs connected to the Internet were satisfied with various aspects of their Internet service – in particular, subscription charges (92%) and quality of service (90%).

Figure 3.7: Satisfaction with Internet service

Source: Oftel Q3 business survey, February 2001

3.26 Small businesses were significantly less satisfied with speed of access than medium businesses (77% and 88% respectively). This reflects their higher usage of PSTN dial-up access compared to medium businesses, who are more likely to use faster means of access such as ISDN.

3.27 Satisfaction with customer care remained lowest, and had fallen from 78% to 73% since November last year. Small businesses, who were least satisfied, were also less likely than medium businesses to have switched ISP. This suggests there is a potential for them to make greater use of market opportunities.

3.28 On the other hand, between November 2000 and February 2001, satisfaction with subscription charges rose 17% (75% to 92%).

3.29 In November 2000, Oftel also asked SMEs how satisfied they were with the choices of services on offer. This survey found that at least three in four SMEs were satisfied with the amount of choice available – in particular the number of ISPs available and the quality of service offered by providers.

Figure 3.8: Satisfaction with choice in Internet service

Source: Oftel Q3 business survey, November 2000

3.30 Satisfaction was lowest with the choice of customer care. By business size, the only major difference in satisfaction was for choice in access speeds – small businesses were less satisfied (78%) than medium businesses (88%).

Oftel’s Consumer Representation Section

3.31 Oftel’s Consumer Representation Section (CRS) receives around 100,000 complaints and enquiries each year from consumers. Oftel has received a gowing number of Internet related complaints over the last 12 months. Data shows, however, that between June 2000 to May 2001, the number of complaints fell relative to the growth in Internet households. In June 2000, 6 million UK homes were connected to the Internet. In that month, Oftel’s CRS received 323 complaints about ISPs. In May 2001, the number of Internet households rose to 10 million and the number of complaints increased to 391. As such, complaints grew over that period by 21%, yet the number of Internet households grew by 66%. This suggests that growth has been managed without significant consumer dissatisfaction.

3.32 In May 2001, 30% of complaints concerning ISPs were about quality of customer service – chart 3.9 shows a consistently high proportion each month. These complaints are about difficulty in contacting an ISP, being ignored or kept uninformed, letters not being answered, the supply of incorrect information and being passed around the organisation.

Chart 3.9: Number of complaints received by Oftel CRS about ISPs

Source: Oftel CRS

3.33 Complaints about charges are also a large proportion of the total complaints. About one in three related to being charged by an ISP for customer services.

Figure 3.10: % of complaints made to Oftel CRS about ISPs.

Source: Oftel CRS

3.34 Complaints about ISPs make up only a 5% of the overall total.

Other consumer concerns

3.35 Oftel has received comments from consumers and consumer groups about access, provision of services and terms and conditions. Some examples of these concerns are paraphrased below:

• Access

"When an unmetered ISP starts taking on customers, the quality of access is satisfactory. However, after a while it can become saturated by new subscribers and cannot cope with increased demand. Consumers complain, which in turn results in the ISP expelling heavy users to relieve the burden on the service."

"If an ISP is facing serious service problems then they should stop new subscriptions until they are able to upgrade their service."

• Provision of services

"Complaints about pricing, though far from having ceased, are now taking second place to complaints about provision"

"Consumers are angry when they are not able to contact their ISP or feel that they do not have access to relevant information."

• Terms and conditions

"ISPs have been known to change their terms and conditions without giving consumers sufficient notice. This has particularly affected unmetered products where they actually start removing services and initiating ‘time-outs’, even on supposedly 24/7 access (see above)."

"Terms such as ‘free’ and ‘unlimited’ no longer mean exactly that. Over-subscription leads to some heavy users being removed from a service altogether, whilst others are ‘timed-out’ after just two hours."

3.36 The Consumers’ Association (CA) has received consumer complaints relating to ISPs, many of which involve unfair terms and conditions. Which? Online (September 2000) found that consumer confidence in the ISP market is lessened by unclear or misleading contract terms. CA called for ISPs to make it clearer to their customers what terms and conditions they will be bound by when they sign up to an unmetered Internet access service. It said "Rather than luring consumers in and then kicking them off schemes, what is needed on their part is better planning, realistic projections of customer usage and clearer advertising for new schemes."

3.37 ISPs need to develop accurate and easy to understand terms and conditions and to be open about their activities – such as the introduction of new packages or modifications to existing ones. Oftel has powers to determine the suitability of standard contract terms under both the Unfair Contract Terms in Consumer Contracts Regulations 1999 and the Telecommunications (Open Network Provision) (Voice Telephony) Regulations, 1998. Oftel handles complaints that terms in telecom service contracts are contrary to the requirements of these regulations. These Regulations fit with Oftel's strategy to protect and empower consumers, by setting out a framework of what would or would not be acceptable standard terms in telecom service contracts and enabling consumers to take better-informed decisions.

Consumer behaviour

Access to information

3.38 The extent to which consumers are able to make informed and effective choices is dependent on the availability of adequate, comparable information on service providers. Lack of information and awareness may lead to consumers missing out on the best deal.

Residential consumers

3.39 Oftel’s quarterly survey of residential Internet users in November 2000 asked consumers where they got their information from about ISPs, packages, equipment and costs when getting connected to the Internet at home. The survey also asked consumers how satisfied they were with the information available to them.

3.40 Over one third of respondents to Oftel’s survey of residential users said their information came from family, friends or other acquaintances. This reflects consumer behaviour in other telecoms markets such as the mobile market. However, around 47% of those using friends and family for information also said they used other sources of information. Furthermore, those using friends and family for information were more satisfied than average with both their ISP quality of service, and with the information available to assist them choose an Internet package.

Figure 3.11: Sources of information used by consumers when connecting to the Internet

Source: Oftel Q3 residential survey, November 2000

3.41 About one in six Internet homes said they looked to advertising and company literature for information about going online and a similar number referred to computer magazines. Other sources included newspapers, web sites, computer shops and free CD ROMs. One in five consumers could not remember what information they had used, or had used none at all when connecting to the Internet.

3.42 Organisations that provide consumers with information include traditional consumer advice organisations such as Trading Standards, Citizens Advice Bureaux and the Consumers Association, and Government-backed initiatives such as the Consumer Gateway. Of these, only Consumers Association/Which? Magazine was explicitly mentioned by residential consumers in Oftel’s survey, though 6% did refer to independent web sites.

3.43 More recently, online forums such as www.ISPreview.co.uk and www.Net4Nowt.co.uk have provided Internet consumers with information about ISPs’ offerings and quality of service and a chance for them to air their opinions on service providers and compare experiences.

Business users

3.44 The survey of SMEs found that the most common source of information used by businesses getting connected to the Internet was advice from family/friends/colleagues – particularly for small businesses. Other sources (less than 2% of SMEs connected) included TV and radio programmes, advertising, trade associations and training courses.

Figure 3.12: Sources of information used when getting

Source: Oftel Q3 business survey, November 2000

3.45 More medium-sized businesses (16%) used independent web sites to get information than small businesses (4%). This is relevant to the availability of sites such as Telecomsadvice (see below) and highlights the need to generate greater awareness amongst SMEs of these sources of impartial information.

Satisfaction with availability of information

Residential

3.46 Oftel research showed that, in November 2000, 59% of residential Internet users were satisfied with the information available to help them choose an Internet package and 22% were not satisfied. Satisfaction seems relatively low.

3.47 Indeed satisfaction is lower than in the mobile telephone market where 78% of consumers were satisfied with the range and quality of information available to assist their selection and just 8% said they were not satisfied.

Business

3.48 Satisfaction with existing information about connecting to the Internet remains high among business users (86-88% in February 2001). Businesses commonly accepted that a shortage of reliable independent information had led them to use these more informal sources of advice, such as friends and family.

Figure 3.13: % UK businesses satisfied with information available on the Internet market

Source: Oftel Q3 business omnibus survey, February 2001

3.49 Oftel has set up focus groups among small businesses to investigate their use of Internet and ecommerce. These focus groups discussed the problems and barriers businesses face getting online. Small businesses in particular felt there was a shortage of useful information available to them on getting online. Instead, and as confirmed in figure 3.12 above, small businesses were relying on informal sources of advice such as friends and family.

3.50 Initiatives already exist that aim to improve businesses’ awareness and understanding of the benefits of the Internet. These include Telecomsadvice – www.telecomsadvice.org.uk – an industry-sponsored, independent advice site for small businesses (established in response to a recommendation by Oftel’s Small Business Taskforce) and www.ukonlineforbusiness.gov.uk – an industry/Government partnership helping UK SMEs to make the most of new technologies.

Switching ISPs

Residential consumers

3.51 Around 30% of Internet homes had switched ISP at least once. In May 2000, the main reasons given for using more than one ISP were "in case one isn’t available/can’t be accessed" and "for differing amounts/allowances of unmetered access". Other reasons included the advantages of using several ISPs for "different content", and in order to satisfy "household members’ preferences". On the other hand, some consumers said they used a number of ISPs as a result of failing to cancel old contracts.

3.52 In addition to switching, a number of residential consumers use more than one ISP. In May 2001, around one in five said they used more than one ISP.

Figure 3.14: Number of ISPs used by Internet households

Source: GB homes with Internet (518 – 9% don’t knows have been excluded) – Oftel Feb 2001

3.53 However, multi-ISP usage has fallen as consumers are deciding on one preferred ISP or are increasingly moving towards unmetered packages. Consumers using fully unmetered packages do not use different ISPs for cheaper access at different times of the day, unlike many of those using a variety of partly unmetered and pay-as-you-go types of access.

Business consumers

3.54 Almost a quarter of businesses use more than one ISP. The main reason why SMEs switched ISP was to get a cheaper service. Faster access speeds and better content were also incentives to switching. 13% of medium businesses had switched ISP in order to receive better levels of customer service.

3.55 83% of businesses had found switching easy. One fifth would be likely to cancel their subscription to an ISP and switch to a new one in the future.

Supplier behaviour

Active competition in terms of price, quality and innovation

3.56 ISPs have been innovative in terms of price, quality and services. The pay-as-you-go model, which has been copied by many other countries, was established in the UK. Internet access is available at prices below local-rate voice calls. Some packages include a monthly subscription fee in addition to the cost of calls, whereas others are ‘subscription-free’.

3.57 The introduction of subscription-free services in 1999 had a significant effect on the number of UK households connecting to the Internet. Freeserve, the pioneer of this model, continues to be the market leader for UK residential and SME dial-up access (see market shares below).

3.58 Price competition among subscription-free ISPs forced down Internet call charges. The first partly unmetered ISP packages offered off-peak unmetered access via an 0800 number, but these were limited by time-out periods and were available to only a limited number of subscribers.

3.59 In January 2000, Oftel’s update to the Number Translation Service enabled terminating operators to set their own retail tariffs rather than being restricted by the originating operator’s tariffs (normally local rate). This enabled ISPs to offer consumers packages based on monthly subscription fees plus call charges below the price of local rate voice calls at around 1p and 2p a minute, for off-peak and peak access.

3.60 Unmetered access was the primary focus in 2000, with the introduction of flat rate access for both peak and off-peak access. Many of the 0800 services failed, having underestimated consumer demand and cost. Some of these packages were withdrawn from the market and others were never launched, although the cable companies introduced their own flat-rate access packages successfully.

3.61 In June 2000, BT introduced its retail product SurfTime at £5.99 per month for unlimited off-peak access and £19.99 per month for anytime access. SurfTime provides network service only and must be bought by consumers in conjunction with a SurfTime-participating ISP. For example, Lineone Surftime offers unmetered off-peak access for a total cost of £9.99 a month, £5.99 of which goes to BT.

3.62 Following a complaint by Worldcom, Oftel required BT to offer a wholesale flat-rate interconnect model in May 2000 (FRIACO, see Annex B). AOL launched the first FRIACO-based anytime unmetered package in September 2000 at £14.99 a month. Since then a number of ISPs, including BT and Freeserve, have launched similar products for between £12.99 and £14.99 a month. These are prices that compare favourably with the United States.

3.63 The growing importance of the SME market has had a significant effect on the Internet access market. A number of ISPs offer packages specifically for small business users. ISPs such as Demon and Globalnet tailor their services to business users, whilst larger ISPs have established distinct brands, for example AOL’s Compuserve.

Retail prices

3.64 Oftel’s six-monthly international benchmarking studies show that, between February 2000 and February 2001, retail prices have, in general, fallen.

3.65 For residential consumers, prices for metered access have fallen by 33% and 37% for off-peak and peak access respectively, between February 2000 and February 2001. This compares reasonably well with price falls seen in other countries. For unmetered access, residential prices have fallen 14% for fully unmetered access, and have risen slightly (4%) for partly unmetered access, between August 2000 and February 2001.

3.66 For business metered access, prices have fallen by 31% between February 2000 and February 2001 – faster than everywhere else apart from Germany. For unmetered access, prices have stayed about the same between August 2000 and February 2001.

3.67 These figures do not capture the full impact of unmetered packages. In other words, before the widespread availability of unmetered packages, a heavier user would have to pay by the minute (metered) to access the Internet in the same way an unmetered user could. The availability of unmetered products now allows businesses and residential consumers to access the Internet at anytime based on a flat rate, bringing prices down significantly.

Recent entry

3.68 The number of ISPs entering the dial-up market has slowed down as the market has matured and business models have evolved.

3.69 However the wide number of new dial-up access packages, based on unmetered and pay-as-you-go models, being offered to both SMEs and residential consumers, indicates a highly competitive market in which the consumer benefits from a wide choice of prices and packages.

3.70 Recent activity in the ISP market does indicate a degree of consolidation. This is particularly noticeable at a European level with large operators buying up a number of European ISPs. Italian operator Tiscali began a particularly noticeable campaign, taking over several ISPs during the last 12 months including the UK’s World Online and Liberty Surf. The UK’s largest ISP Freeserve, was acquired by the French ISP Wanadoo SA in January 2001.

Competitive pricing

3.71 Competition puts pressure on firms to set prices close to cost. In a competitive market, therefore, one would not expect to see persistent excess profits, although temporary excess profits may result from innovation for example.

3.72 Persistent excessive profits are an indication that prices are above the competitive level and consequently that an element of market power may exist. In the case of Internet service provision, the plethora of available tariff structures for ISPs and the large number of suppliers operating in the market (ISPA estimates 400+ ISPs) makes it difficult to observe directly whether prices are in general close to cost or whether excess profits are being made. It should, however, be recognised that the existence of this variety is itself a sign that the market is competitive.

3.73 Moreover, two further pieces of evidence suggest that it is unlikely that Internet service providers could make persistent excess profits. Firstly, pricing information points to the availability of a number of different tariff options designed to encourage customer usage and meet consumer needs. Some of these deals are subscription-free, and some others entail a fixed price for the purchase of unlimited Internet access.

3.74 Oftel considers that this information would seem to indicate the presence of active competition in the market. More importantly, this information can be considered in the context of one of the main characteristics of the ISP market, namely, the lack of significant barriers to entry. This would suggest that any supernormal profit margin (ie a margin which is linked to prices being set above the competitive level) would be competed away rapidly, as entry to the market is likely to occur quickly and without incurring significant sunk costs. Notwithstanding the absence of detailed information on prices and profitability at this stage, these elements (evidence of current pricing behaviour of firms, the large number of participants and low barriers to entry to Internet service provision) point to the fact that the market is effectively competitive.

Market structure

Market share

3.75 Internet service provision market shares are difficult to calculate. Those estimates that do exist are made less easy to interpret by the large number of users who register with more than one ISP.

3.76 Oftel asked major ISPs to provide estimates of active subscribers – ie those registered subscribers who actually use the service. Oftel received few responses from the ISPs and so the data cannot be used to make any firm assessment. This document therefore uses estimates from Oftel’s existing market research, which asked existing residential and business users which ISPs they use.

Residential

3.77 The following table is based on what ISPs residential consumers said they were using over the last 12 months. Only ISPs with a share of more than 2% are shown – those with less than this amount are included in the ‘other’ category.

Figure 3.15: Estimate of Internet Service Provider Subscriber Share

Source: Oftel quarterly residential survey

3.78 Residential consumers are making use of a wide range of ISPs. Whilst Freeserve has continued to be the most popular residential ISP over the last 12 months (18%), AOL’s market share has grown considerably from 8% to 17% (May 2000 – 2001). This means that AOL is now almost equal to the market leader. BT Internet and more recently, NTL, also enjoy significant market presence (8-12%). There are many smaller ISPs representing 28% of the market who, at under 2% each, are not listed in the table above. This is indicative of the large number of ISPs in the UK market.

Business

3.79 The table below shows the individual ISP shares of the SME market based on what ISPs businesses said they were using (some businesses are using more than 1 ISP.) Again, only ISPs with a share of 2% or more are shown.

Figure 3.16: ISP share of SME business Internet access