5. Summary of the Evidentiary Record

Below we consider the record created in the course of nine days of hearing. We describe the explosion in Cingular's California customer base beginning in early 2000 and the related capacity and coverage problems on its overburdened network, which Cingular failed to disclose to customers. We review the evidence on Cingular's intensive marketing during this period, its limited sales disclosures, and the content of its advertising. Finally we examine seven different data sources, which represent (or purport to represent) customer dissatisfaction with Cingular. In Section 6.1 of today's decision we discuss how this evidence, on balance, establishes two violations of law and supports the penalties and reparations we discuss in Section 6.2.

5.1 Customer Growth and Network Development

The record reveals that Cingular's California customer base numbered approximately 3 million customers by late 2001/early 2002. This represents nearly a doubling of customers over a two-year timeframe, since, as summarized in the prepared testimony of witness Michael W. Bennett (Bennett), Cingular's Executive Director - External Affairs, Cingular added approximately 1.5 million customers from January 2000 to the end of 2001. During this same period, the average minutes of use per customer more than doubled as well, from 258 minutes a month to 546 minutes a month. The rebuttal testimony of witness James Jacot (Jacot), Cingular's Regional Vice President, Network Operations for the West Region, documents that throughout most of 2001 much of Cingular's California service area failed to meet three internal performance targets: service denied (also referred to as blocked calls); lost calls (defined as customer perceived dropped calls); and switch congestion.

Cingular generally concedes that growth during this period led to network problems, particularly during 2001. The rebuttal testimony of witness Kathleen M. Lee (Lee), Cingular's Network Sales and Network Issues Manager for the West Region states: "Cingular acknowledges that it has had its share of growing pains from the unexpectedly large growth in customer base and minutes of use, during a relatively short period of time." (Ex. 402, p. 6.) Bennett's rebuttal testimony, focusing on usage, admits: "This extreme growth in minutes of usage did impact the sufficiency of Cingular's network." (Ex. 407, p. 4.)

Bennett's rebuttal contends, however: "Cingular did not sit idly by and ignore the impact this explosion of customer use had on its network." (Ibid.) Cingular spent over $1.6 billion in California on network upgrades during this two-year period, adding 1700 cell sites and increasing its coverage area from 39,334 square miles to 61,099 square miles. According to Cingular's Jacot, about 50% of that sum was spent in 2002, about 30% in 2001 and about 20% in 2000.⁸

Jacot's rebuttal testimony admits that Cingular did not anticipate - or budget -- for the minutes of use growth that it actually experienced. His rebuttal states: "Although a company can estimate what the results of a marketing plan will produce, it is not always predictable or certain how popular a calling plan will turn out." (Ex. 401, p. 11.) By example, Jacot's rebuttal refers to "the college campus phenomenon," explaining that:

...during the Fall of 2001, Cingular experienced a spike in MOUs [minutes of use] in non-peak hours (after 9:00 p.m.) on college campuses in California. We did not predict that college students living on campus would make multitudes of long distance phone calls after 9:00 p.m. using their wireless phones because it was cheaper than using their landline phone service. (Ibid.)

However, email correspondence within Cingular's engineering department establishes that at least as early as February 2001 some employees there were concerned about the network's ability to perform adequately in response to a marketing campaign referred to as "Spring Promotion," which would offer unlimited nights and weekends calling to rate plans costing $29.99 per month or higher. A February 9, 2001 response to the marketing proposal states:

...the short answer is that we have NO excess capacity. We have had nights and weekend rate plans for a while, but not with that kind of take rate. Our highest blocking is currently on Saturday. Increasing sales of this would simply make an existing problem worse." (Ex. 202, Attach. 3.)

The same author's further response anticipates switch congestion problems and also states:

On the radio side, in some areas weekend traffic is already a problem. This promotion will cause a need of additional equipment. We are so far behind now in funding, that trying to estimate the amount and cost of this is not a good use of time. We are focusing on trying to catch up with the current situation..." (Ibid.)

A subsequent response from a different author, dated February 13, 2001, concludes that if traffic increases 5% in April and May and returns to normal in June, Cingular may "survive this promotion".⁹ (Ex. 18, Attach. 21; Ex. 202, Attach. 4.)

Nonetheless, the record reflects that Cingular continued to advertise heavily in 2001. In fact, a moratorium on sales was out of the question, according to Jacot, who has been with Cingular since May 15, 2001. He testified:

I don't believe that they have ever taken a moratorium on sales. They have certainly, in the time I've been here, moved to less aggressive sales plans, less aggressive promotions in order not to overburden the network. But the difficulty of putting a moratorium on sales is your salespeople - agents and employees - all need to go somewhere where they can make sales and get commissions. You don't recover from that." (Tr. p. 948.)

Cingular's witness Ricardo Cruz (Cruz), currently President of Ronin Technology Partners, a wireless engineering and management-consulting firm,¹⁰ examined network performance data from 2000 to 2002. His rebuttal testimony concludes that the capacity problem attributable to the growth in minutes of use "was largely confined to 2001 ... and since that time, has continued to improve." (Ex. 400, p. 27.) Cruz's rebuttal quotes from the deposition transcript of CPSD's witness Robert Zicker (Zicker), an independent consultant in the telecommunications field.¹¹ In the cited portion of the transcript, Zicker acknowledges Cingular's performance improvements from 2001 to 2002.

Much of the extensive debate between these two expert witnesses, while interesting, is not material to decision of the issues raised by this OII. CPSD's witness Zicker and Cingular's witness Cruz focus on the technology of wireless communications, including the increasingly more powerful signal strengths necessary for outdoor, in-vehicle and in-building coverage, on the reasons for wireless communications failures, on differences between GSM and other wireless technologies and upon the details of Cingular's system, which was not originally designed for in-building coverage. They also address propagation maps used by radio frequency engineers and the usefulness of such maps to most customers. We discuss this latter issue in Section 5.3 of today's decision, together with other evidence on the kinds of measurement information generally available within the wireless industry.

In many respects Zicker and Cruz agree. For example, they agree that Cingular's system has various kinds of coverage holes, as do all wireless systems.¹² They differ as to the import of such coverage holes. At hearing, however, Zicker clarified that he was not contending that Cingular's network is inferior. He also testified: "I don't believe that I've ever stated any wrongdoing by Cingular except its lack of notification to its customers about the existence of these coverage holes." (Tr. p. 736.)

Zicker reiterated this point in the course of questioning about in-building coverage.

Q. [Cingular counsel] Has Cingular ever denied that it didn't - that it didn't design, originally design, its own network to provide in-building coverage in the context of - in any of the testimony that you read in this proceeding by Cingular witnesses?

A. [Zicker] In the testimony I've seen, no, Cingular has not denied, I have not seen any customer brochures that even addresses the issue of in-building coverage. (Tr. p. 746-7.)

Attachment 19 to the rebuttal testimony of UCAN's Shames includes a list of 19 company-owned and agent stores in which Cingular used signal enhancers at some time during 2001 or 2002. UCAN's opening brief argues that existing signal strength at these store locations was inadequate to supply in-building coverage and that the enhancers were supplied to mask this reality by creating a deceptive impression of good in-building coverage. However, UCAN's theory is unproven, since UCAN did not examine any Cingular witness about these devices at hearing.¹³

In his rebuttal testimony, Cingular's Bennett addresses the utility's response to its network problems as follows: "As soon as Cingular discovered the problems associated with the increased MOU, it went to great lengths to correct them." (Ex. 400, p. 30.) One issue upon which CPSD's Zicker and Cingular's Jacot and Cruz all agree, however, is that system infrastructure improvements take time to implement and, in some instances, may not be wholly within a carrier's control. Jacot's rebuttal discusses local regulatory obstacles to infrastructure development, such as municipal cell siting moratoria, which may occur while local telecommunications siting ordinances are developed, during

the study of environmental impacts or because of protest by residents. Cingular has experienced such problems in both Northern and Southern California, and Jacot's rebuttal testimony includes a list of problem jurisdictions. Cruz testified: "It can take six months or longer to make the improvements, once a problem is identified." (Tr. p. 847.) Because of this lag time, Cruz suggested that network performance should be reviewed over a broad window of perhaps 18 months.

Data on customer turnover, or "churn," provides another perspective on the misfit between Cingular's customer growth and network development during this period, as well as the impact of the ETF. A market research study entitled "Profiling the Killer - Churn," which is attached to Caceres' reply testimony, shows a significant spike in customer contract cancellations (termed "deactivations") in months 13-16, after contract initiation, once the ETF on one-year contracts ceased to apply. (Ex. 3, Attach. 5.) The research points to network problems as a leading factor in customer-initiated service cancellations. Other market research in the record underscores the importance to consumers of network quality.

5.2 Marketing and Sales Practices

5.2.1 Store Design, Marketing Materials and Sales Disclosures

As noted previously, Cingular markets and sells wireless services directly, through its company-owned stores, and indirectly, via a statewide network of agents and dealers. Company-owned stores account for approximately 15% of the retail sales locations.

The rebuttal testimony of David B. Garver (Garver), Cingular's National Director of Marketing¹⁴, describes this distribution network at some length.¹⁵ Among other things, Garver's rebuttal testimony explains that Cingular promotes a consistent image for its exclusive agents so that all such agents' stores or kiosks have the same "look and feel," which "is accomplished through the fixtures that go into the store, as well as the store layout." (Ex. 406 at p. 3.)

Indeed, street view photographs by CPSD attached to Caceres' reply testimony show prominent display of the Cingular Wireless name and its "Cingular Jack" logo at both a company-owned store and agent stores in San Francisco. The agency disclosure, typically affixed to a glass window above the doorway, is not immediately apparent, by comparison. In the declaration that accompanies the photographs, CPSD's witness Richard C. Maniscalco, who visited 14 San Francisco bay area stores, states that company and agent stores were not readily distinguishable. Based on the common "look and feel," customers who did not notice the agency disclosure above the doorway (or wherever posted) would likely assume they were doing business directly with Cingular. Several customer witnesses describe purchasing service at what they thought was a Cingular store, only to discover later that it was not.¹⁶

Cingular communicates with its sales personnel, both employees and agents/dealers, through dated "Newsflashes" containing various categories of marketing information, such as promotions, rate plan information, sales scripts, and equipment pricing. Sometimes sales training or sales instructions are disseminated in this way. Caceres' Supplemental Report states that within the 3,066 pages of Newsflashes produced she found no:

... instructional information about the limits of Cingular's coverage and system capacity, or any information about how the sales force should address these issues or disclose information pertaining to cellular coverage, network capacity or dead zones. (Ex. 2, p. 7.)

There is no evidence that Cingular's sales representatives and agents were instructed to advise customers about known, major network problems, such as limited in-building coverage or areas lacking coverage altogether. In some instances, however, customers state that sales personnel represented that given cities, towns, or even specific streets had coverage, when they did not. This was Caceres experience when she made eight different site visits, four each in San Francisco and Los Angeles.¹⁷

The record reveals that Cingular's in-store coverage maps, as well as customer brochures depicting coverage, portray continuous coverage over most of the San Diego, Los Angeles and San Francisco metropolitan areas and through much of the Central Valley and Sacramento. They make no distinction among outdoor, in-vehicle or in-building coverage. Ex. 41, a photograph of a large wall map displayed in stores, shows a large-type, highly visible "Coverage Legend" with three different color codes: "Current Coverage"; "Coverage Over Water"; and "Planned Coverage." By comparison, the small type service disclaimer is illegible in the photograph. Ex. 214, a customer brochure (not a photocopy) entitled "Never Pay Long Distance Again," includes a similar map of California and one of Washington. The legend for this map, however, includes a four-color code for four different existing or planned calling areas and below that the following disclaimer: "The coverage depicted on this map reflects rate plan coverage for rate plans available after January 1, 2002." (Ex. 214, emphasis added.) The brochure does not explain to the customer how a rate area and a coverage area differ. Shown Ex. 219 (which appears to be a photocopy of Ex. 214 but marked with Bates stamp numbers), Garver testified that it represents the kind of brochure sales personnel give to customers who ask to see coverage maps, particularly if the store does not have a wall map. Jacot also referred to Ex. 214 as showing coverage.

Ex. 214 contains a second disclaimer, placed lower on the page in even smaller type, which repeats that the map does not show coverage:

Map depicts rate areas only, not coverage areas. Actual coverage areas different substantially from map graphics and coverage may be affected by such things as terrain, weather, foliage, buildings and other construction, signal strength, customer equipment and other factors. Cingular does not guarantee coverage. Roaming charges and other charges will be billed based on the location of the site receiving and transmitting the call, not on the location of the subscriber. Extended Home Area [a rate plan] is serviced by a non-Cingular GSM Carrier. Any representations of wireless coverage are based on information prepare by a non-Cingular GSM Carrier and Cingular is not responsible for such representations. (Ibid. emphasis added.)

Cingular and its exclusive agents and dealers use these materials and other brochures which are contained in the record. Apparently others in the distribution network use them as well, since customers who purchased service from non-exclusive agents supplied some of the brochures to CPSD.

In response to questioning about what network performance information Cingular could supply to prospective customers, the utility's witness Cruz testified, "...I think what would be most appropriate is an estimation of where coverage may likely - where it's more likely to be expected." (Tr. p. 842.) He admitted that a rate map does not necessarily provide an accurate depiction of coverage, but stated that a correlation between rate areas and coverage areas exists.

The ALJ queried Cingular's Jacot about a hypothetical prospective customer who wanted to ensure, to the extent possible, the ability to use a handset within a vehicle along a given commute path every workday. She asked specifically what technical or engineering information that prospective customer would need to know in order to make an informed choice among wireless providers. Jacot responded:

Well, first of all, they would obviously want to know, does the carrier provide coverage at all, you know, so they'd know if a signal was available.

They would also want some information about the different handsets they could get and what kind of a - how well those handsets perform, particularly in an automobile.

They might want information about car kits that are available so they know what kind of benefits they can get from having an external antenna mounted on the car.

And then they would like to know what kind of performance -- network -- measured network performance, both in terms of power of signal and in terms of signal-to-interference ratio does the carrier exhibit on that road over the part of that road that they would like to use their phone.

And then they would like to know what plans does the carrier have to provide technical improvements to the quality of the signal on that road in the future. (Tr. p. 951-2.)

Asked by the ALJ whether that information is available to a customer through Cingular, Jacot replied:

Not all that information is available to point -- at the point of sale. They can get information through Customer Care if they would call in and ask about plans to -- you know, to build-out the road in the future. Specific signal-level information is not generally made available to customers. (Id. at p. 952.)

Jacot testified that specific-signal information is available to Cingular's salespeople and the salespeople employed by its agents, but he later admitted that detailed information actually must be obtained from radio frequency engineers within the company. Likewise, though customer service representatives have more information about network performance than that made available in stores, they must turn to radio frequency engineers for detailed information about the likelihood of coverage. Customer service representatives do not have access to the actual propagation maps that these engineers create to estimate (or predict) the probability of outdoor, in-vehicle and in-building coverage at a specific location or area.¹⁸

Asked by the ALJ what additional technical and engineering information Cingular could make available to customers at low cost in order to provide better information, Jacot testified:

The difficulty is that the information is only relevant for a specific amount of time. It's a difficult process for the salespeople to keep abreast for all specific areas in the network over a long period of time about where coverage exists and where it doesn't; and so while the information is generally available for them to find out, having that information in mind or at hand at the time a specific customer comes in, asks a specific question about coverage in a specific area, generally is not something that they have -- have on top of their mind or knowledge or specific knowledge of. (Id. at p. 954.)

This line of examination continued as follows:

Q. [ALJ] If a customer were to explain to a salesperson, for example, that he or she had a particular reason to attempt to ascertain the likelihood of specific coverage, would the salesperson have means to gain further information?

A. [Jacot] Not at the immediate -- I don't think they could at the immediate point of time, but they certainly could take the question from the customer, offer a promise to get back to the customer at a future date -- 24 hours, 48 hours, whatever -- and go back and either through the use of somebody like Kathleen Lee, who is there specifically to support the sales organization, or through -- or through contact back at the RF [radio frequency] engineers, gain the information and come back and provide a response. (Id. at pp. 954-5.)

5.2.2 Agent Contracts

Prior to May 1, 2002, the effective date of Cingular's new refund/return policy, agents and dealers sometimes charged customers an additional ETF. Sample contracts in the record incorporate agent/dealer ETFs of as much as $400. Garver's rebuttal testimony states, "[T]he contractual relationship between Cingular and its agents/dealers leaves them free to sell wireless equipment on their own account, at a price which they choose and on terms and condition which they establish." (Ex. 406, p. 4.) If the customer cancels service before the vesting of the activation commission, which Cingular pays agents/dealers for each new customer,

... in many instances, the agent/dealer would lose its investment in the phone because they often resell the phones to customers at a significant discount from the price paid by the agent/dealer. The fee for cancellation for the equipment contract, was, in all likelihood, a means for the agent/dealer to recoup part of those losses. (Id. at p. 7.)

Thus, Garver's rebuttal testimony purports to focus, among other things "on the limited degree of business control" which Cingular has in its relationships with its agents and dealers. (Id. at p. 1.) However, review of Cingular's "Pro Forma Authorized Agency Agreement," attached to Garver's rebuttal testimony, indicates that Cingular holds a great deal of control over them. The agreement, prepared by Cingular and marked proprietary, expressly provides that the signatory owes Cingular "the fiduciary and other obligations of an agent to its principal" with respect to the selling of the "Authorized Cingular Services." (Id., Attach. 1.) Cingular clearly chooses to exercise control in certain areas. For example, the agreement requires the signatory to provide Cingular with copies of any proposed marketing and advertising materials and to obtain advance written approval to use them. Other indicia of control include various compensation terms, such as the activation commission and advertising reimbursement, described in Section 5.3 of today's decision. Cingular also sets activation quotas for agents and dealers.

Most tellingly, once Cingular determined to implement its new ETF policy, effective May 1, 2002, it required agents and dealers to execute an "Amendment to Agency Agreement Re Phone Return Policy," which requires such entities to honor the new policy as of that date.

5.2.3 Wireless Handsets and Other Equipment

Garver's rebuttal testimony attempts to put some distance between Cingular and its agents/dealers with respect to sales of equipment such as wireless handsets. His rebuttal states: "Agent/dealers are not required to inform Cingular as to the terms and conditions pursuant to which they sell wireless phones." (Id. at p. 6.) The rebuttal testimony explains that Cingular was aware that agents/dealers had various return policies which differed from Cingular's express no return/no refund policy--some allowing returns within three days, or seven days, or as much as 30 days, for example. As we recount above, Cingular also permitted its agents and dealers to charge an additional ETF to recoup both the commission forfeited by a customer's early contract termination and the cost of discounting phones.

The record provides limited information on exactly what types of wireless equipment Cingular's agents and dealers sold to customers over the timeframe at issue, which manufacturers' wireless phones (or what models) they carried, what the performance parameters of those phones were, what service and phone packages they offered or how those packages were priced.¹⁹ Garver testified that Cingular encourages agents/dealers to buy wireless phones for resale from Cingular but that they may buy them from other suppliers. Caceres' reply testimony includes, in an attachment, several photocopied pages (two not entirely legible), of the box she received with her wireless handset, apparently a "Nokia 3390." Garver testified that these photocopies appeared to represent the typical box for equipment sold by exclusive agents. Both Cruz and Jacot testified that the handset models Cingular offers are manufactured to appropriate GSM standards and checked for compliance.

5.3 Advertising

Cingular advertises in various media, including newspapers, radio and television. It also exercises substantial control over the advertising efforts of its agents/dealers. Cingular provides all exclusive agents with its written "Cooperative Advertising Program Guidelines," which set out content requirements. The rebuttal testimony of Cingular's witness Garver's includes a copy of these guidelines and states that similar guidelines are provided to exclusive and nonexclusive dealers. In addition, for every activation, Cingular pays $25.00 to the Exclusive Agent's co-op fund, which subsidizes the cost of agent advertising. Cingular does not monitor the ads directly but has contracted with Cooptium, Inc., an outside entity, to do so.

Many of the sample ads introduced into evidence in this proceeding list both company-owned and agent sales locations, but these ads typically do not distinguish them as such. While sample ads reference Cingular's $150 ETF, they do not advise potential customers that an additional ETF will apply. Asked to review the agent newspaper ad referred to in this proceeding as the "When-ever, Where-ever" ad (though its title is "Are you Wireless?"), Garver testified that the nine-line, small print disclosure at the bottom of the full page appeared to be the typical, legal wording Cingular used during 2001-2002. The $150 ETF reference appears in the second line. There is no mention that the ETF may be higher, only the statement, in the eighth line, that "[o]ther restrictions may apply." (Ex. 405, Attach. 3 [San Francisco Chronicle, April 2, 2002].)

One version of another newspaper ad, titled "Never pay long distance or roaming charges again," contains a somewhat different disclosure, again in small type at the bottom of the page. This disclosure uses the convention of bold type subject titles throughout the disclosure text; in the second line, after the subject title "Phone Return Policy/Early Termination Fee," is the following, underlined text: "no early termination fee if service cancelled within 15 days of purchase; a $150 early termination fee applies thereafter." The next to last line includes the phrase, "Additional conditions and restrictions apply." (Id., Attach. 13 [San Francisco Chronicle, June 20, 2002]. Cingular's witness, Dr. Michael A. Kamins (Kamins), Associate Professor of Marketing, Marshall School of Business Administration at the University of Southern California,²⁰ was asked, hypothetically, whether such text was sufficient to advise consumers that agents might charge a higher ETF. He testified: "I don't know if there is confusion, but there is a potential for confusion, granted. Again, assuming that no other questions are asked [by the consumer]." (Tr. p. 1113.)

Cingular offered Kamins to counter CPSD's witness, Professor Anthony R. Pratkanis (Pratkanis), Professor of Psychology in the Department of Psychology, University of California, Santa Cruz,²¹ whose prepared testimony concludes that Cingular's advertising and marketing materials promote a theme of unlimited coverage and capacity that is deceptive, given the kinds of network problems described above.

Pratkanis' review focuses heavily upon the "When-ever, Where-ever" ad, which apparently ran in various newspapers in San Francisco, Los Angeles and San Diego over a two month period in the spring of 2002. He concludes that the ad is misleading because both the quoted language (which appears in a byline near the top of the page) and the four photographs above that byline (a beach volleyball game on an urban water front, a backpacker atop a summit, a yacht under sail, a rock climber approaching a ledge) communicate an impression of extensive coverage and capacity. He testified that this impression is reinforced by language in the three other ads attached to his prepared opening testimony and the selection of some thirty ads attached to his reply testimony - words and phrases such as "anytime minutes," "unlimited nationwide," "talk, talk, talk," in other words:

...a whole series of words, phrases, and so forth, throughout the ads .... that convey that you can use your phone whatever [sic], wherever, any time, unlimited, static free, and so on ... And whatever [sic], whenever is a - think of it as a label for all those words. (Tr. p. 1075.)

Pratkanis conceded that he did not select these ads from the thousands of ads that Cingular offered to make available; rather, CPSD's counsel chose these ads for him. He testified that a broader review was unnecessary because any other ads would either reinforce the deception or would present other themes, but not remove the prior deception.

CPSD's witness Pratkanis also testified that he was unaware that other wireless carriers run ads highlighting words such as "anytime," though when shown three ads by other wireless carriers, he admitted that such words appear in them (e.g., Verizon Wireless-"anytime minutes" and "unlimited night & weekend airtime minutes"; AT&T Wireless-"unlimited night & weekend minutes," "anytime minutes" and "nationwide long distance"; Sprint PCS-"anytime minutes," "unlimited PCS to PCS calling," "unlimited night and weekend minutes," and "nationwide long distance"). (Ex. 509, 510, 511.)

Questioned about the message that advertisers attempt to communicate, Cingular's witness Kamins testified that advertisers sometimes are unaware that their advertisements create consumer misperceptions. He explained:

For instance, Alaska Airlines always advertised that they were a quality airline with a lot of leg room and good food. And what happened was that they ended up with a perception by consumers that they were overpriced. (Tr. p. 1104.)

Pressed to state whether he would deem the "When-ever, Where-ever" ad deceptive, Kamins testified that he would have to know whether the consumer's service expectation was unmet and, then, whether there was an intent to deceive. Upon further questioning, he agreed that the ad might be misleading: "If there's an intent to lead consumers down the wrong path and there's an inconsistency between expectations and performance, yes." (Tr. 1120-1121.) Intent was important to his assessment as a marketing consultant, Kamins testified, though he recognized that legal requirements might be different.

Kamins' prepared rebuttal testimony concludes that price and value are the primary themes running through Cingular's advertising and promotional materials, if one examines the totality of the ads. Kamins testified that his opinion relies upon content analysis of a larger group of ads than Pratkanis examined. Like CPSD's Pratkanis, he also reviewed market research data, including the responses of focus group participants and market research studies generated for Cingular by several market research entities. Kamins interprets the market research data to show that while consumers rate good coverage as the most important attribute of wireless service, price is the attribute most important to them in distinguishing between brands. He testified:

To make an analogy, safety may be the most important attribute that you have in airlines, but people don't decide which airline to fly on the basis of safety-most people don't, even though it's the most important attribute. What's critical is the distinguishing attribute between brands, and I've said before that's price. (Tr. p. 1134.)

Pratkanis' placed a different emphasis on the marketing research, testifying that he agreed with a statement he had read that "[c]overage is the ante to get into the game, and without that you're not in the game ... And then, after that ... some items that come up ... oftentimes less important on their list of drivers, such as price, the color of the phone, what kind of phone you're going to get become quite important ... for a specific decision." (Tr. p. 1054.)

The difference between these carefully nuanced expert opinions is rather subtle. Notably, Cingular's Kamins does not suggest that consumers opt for low price in disregard of known, poor performance. And CPSD's Pratkanis admits that price is an important motivational factor for consumers. The common ground between their opinions is this: Market research suggests that consumers either rank adequate coverage first or they do not rank it at all, because they take it as a given.

Thus, it is not surprising that focus group reactions to Cingular's advertising actually exhibit diverse consumer reactions and interpretations. For example, Pratkanis' opening prepared testimony and Kamins' rebuttal both include, as an attachment, what appears to be the same version of a newspaper ad from Cingular's Spiderman advertising campaign, which Cingular subjected to focus group tests. The newspaper ad, which apparently ran to coincide with the May 2002 release of the Spiderman film, is titled "Never pay long distance or roaming again" and depicts Spiderman swinging from a map of the United States. A spider web, with its center in the center of the country, overlays the map and the Cingular jack logo overlays the center of the web. A subtitle below the map reads "Covering the entire country is now simple. Superhuman abilities not required." (Ex. 38, Attach. 5; Ex. 405, Attach. 4.) Late-filed Exhibit 51 contains 17 verbatim responses from focus group participants who were shown some kind of advertising from that campaign and asked about its main message. Most, though not all, mentioned the cartoon character. Some focused on the rate/price associations, some on the interconnection associations, some apparently made no associations, and one thought the ad was targeted at children.

There is no evidence that Cingular's advertising was patently false. The question we consider in Section 6.1 of today's decision is whether the inferences of broad coverage and the lack of express disclosure of agents' ETFs constitute violations of law.

5.4 Customers' Complaints

The customer complaint evidence draws from seven different sources. We review each of them below.

5.4.1 Customer Witnesses

The record includes 49 verified customer complaints against Cingular. Most are in the form of declarations executed under penalty of perjury, though two customers testified at hearing (Joanne Coxum, who did not submit a declaration, and Matt Zumstein, who appeared to give live reply testimony). These 49 verified complaints comprise the 14 customer declarations attached to the Staff Report as well as declarations or testimony from another 35 customers, 13 of them produced by CPSD and 22 by UCAN. Cingular deposed approximately 25 of these customers. The deposition transcripts or excerpts from them are also in the record, as attachments to rebuttal testimony by witness Michelle Rodriguez (Rodriguez), a Customer Relations Specialist in Cingular's Office of the President, and to reply testimony by CPSD's witness Patricia Esule (Esule) and by witness Michael Shames (Shames), UCAN's Executive Director.²² The deposition transcripts largely corroborate the declarations, as do the computerized customer service records, referred to as "Telegence notes," which Cingular produced for most of these customers.

These 49 customers all complain about unjust and unreasonable rules and practices or about the resultant unjust and unreasonable service. Appendix 2 to this decision summarizes the pertinent details of their complaints in matrix form. Appendix 3 indicates by month and year when 47 of the customers began to experience service problems (two customers did not provide dates). The timeframe is March 1999 through January 2003, though only two complaints cite service problems in the outlying years (one in 1999 and one in 2003.) Twenty complaints target 2001, 16 target 2000 and 9 target 2002.

Several of these customer witnesses originally signed up for service with PBW between 1996 and 1999, but complain that service seriously degraded at some point in 2001. Most of the rest complain of no service at all in the places where they intended to use their wireless phones or of extremely poor service, including spotty coverage and significant capacity problems ranging from routinely dropped calls to frequent network busy signals. Some customers reviewed and relied upon the maps available in stores and marketing brochures, which, as we have seen, suggest wide, unbroken service areas in much of California, including the major metropolitan areas. Other customers explained to sales agents exactly why and where they intended to use wireless service, and were assured that Cingular could provide the coverage they needed.

CPSD's Zicker testified that he had participated in focus groups where people exhibited little understanding of wireless technology but "expected their wireless phone to work wherever they were, whenever they wanted it to." (Tr. p. 749.) Nevertheless, more than a dozen of these customer witnesses expressly recognize that wireless service is far from perfect. Many base their expectations on their own experiences with other carriers or their observations of other wireless users. Some, who switched to Cingular from a competitor in order to take advantage of a more attractive rate plan offer, returned to their former carriers as soon as they could get out of Cingular's contract.

Many customer witnesses report being told they would have to pay an ETF if they cancelled. A few had ETF payments reimbursed after lodging an informal complaint with this Commission. More than a dozen customer witnesses decided to wait out the contract period rather than pay the more costly ETF. Some accepted Cingular's offer of a monthly service charge credit or some other credit in return for service retention, an offer that often was not made until the customer made an informal complaint. Some customers decided against cancellation after being told that Cingular would soon be installing new cellular towers or other infrastructure to improve service quality in a given area. Some were told that upgrading their phone would remedy their service problems but afterwards found themselves bound to a new one or two year contract (and an ETF) without any real service quality gains. Other customers canceled their service contracts and waged disputes with Cingular or its agents, sometimes for months, over the ETF and other charges such as activation fees and wireless phone costs, or had their disputed accounts sent to collection agencies.

Caceres provides another sworn account. Her Supplemental Report describes her unsuccessful efforts in February 2002 to cancel the contract for service she had signed three weeks earlier. Since her only use of the handset was a single call from the Commission's Los Angles offices, which resulted in a fast busy signal, the minutes of use on her account could not have been high. She reports that Cingular's customer service representative advised her to contact the store where she had purchased the handset and service. An employee there told her that "they had a no return policy and that I would be required to pay the early termination fee and any equipment charges. The employee further explained that it would be best if I found another individual to take over my contract so I would not be assessed any cancellation fees." (Ex. 2, p. 8.) This advice is consistent with the directives governing cancellation in Cingular's on-line "Ask Jack" program, a policies and procedures guide for customer service representatives appended to Caceres' reply testimony as Attachment 1.

The record reflects that the contract policy effective in California prior to May 2002 (no return/no refund/ETF), while standard within the Western Region, was by no means a national standard. In fact, as Attachment 3 to Caceres' reply testimony shows, Cingular's other regions had more customer-friendly policies, with return periods varying from three days to 30. Moreover, some of these other regions permitted returns regardless of the minutes of use on the customer's account, and some waived other fees besides the ETF.

Despite the written policy prohibiting refunds and imposing an ETF for early contract cancellation in the Western Region, Maureen Cook, Manager in Cingular's Office of the President (OOP), testified that Cingular actually followed a less onerous, "de facto" policy. According to Cook, the de facto policy allowed ETF waivers for customers with low usage who cancelled within the first 15 days of the contract period.²³ While the record reflects that such waivers did occur, it also suggests that Cingular defined low usage very narrowly, which would have reduced customer eligibility. As Esule points out, dropped calls and redials increase usage. So do frequent calls to customer service to report ongoing service problems. The reply testimony of CPSD's Esule's reviews the allegedly high usage of some of the customer witnesses and finds that when translated into average minutes per month or per day, the usage does not appear to be significant.²⁴

Even if a de facto policy existed, there is no evidence that it actually benefited a large number of aggrieved customers. Several factors suggest the contrary. First, customers were not advised of the de facto policy at the time of sale. Therefore, only those customers who ignored (or for some reason were unaware of) Cingular's express, written ETF and no return/no refund policy would have called to request contract cancellation, particularly within 15 days. Second, Cingular concedes that not every customer ostensibly eligible for the ETF waiver received a waiver offer, given the number of different customer service representatives (some 1500) it employs. Third, many customers who attempted to cancel did so after 15 days-contacting Cingular in spite of the no return/no refund/ETF policy because of accumulated frustration with the poor quality of the service they had experienced. Fourth, the record contains compelling evidence that customer retention was paramount for Cingular, and large-scale contract cancellation is inconsistent with customer retention.

Cingular's corporate materials illustrate the importance of its customer retention policy. The "Ask Jack" program overview states: "All Customer Care Representatives handle calls from customers who request cancellation of service. It is every representative's responsibility to save customers by aggressively identifying issues and providing solutions, which reduce churn." (Ex. 3, Attach. 1, p. 1.) Among the save strategies that customer witnesses report are phone exchanges and upgrades tied to execution of a new service contract and assurances about forthcoming infrastructure improvements. Either easily could, and in some reported cases did, keep a customer under contract beyond 15 days.

Another document entitled "Cingular Promise: Return Policy," which appears to be an internal document generated just before the new, 15-day return policy took effect in May 2002, explains that the no return/no refund/ETF policy "... was agreed upon by all West GMs [general managers]. It makes the sale final and reduces churn. The idea is to sell to the right customer." (Id., Attach. 3, p. (23)0710.) As we have already seen, churn was a problem for Cingular, and the company commissioned marketing studies to assess the reasons for it. Though one might assume that selling to the "right customer" should be coupled with a policy of providing potential customers with sufficient information about network coverage to make the right choice, the evidence reveals that it was not.

Our review of the record notes the financial mitigation Cingular extended to a number of customer witnesses, often after the customer lodged a complaint. Cingular's mitigation measures include offering credits toward monthly service charges. We consider the impact of such mitigation in Section 6.2 of today's decision. However, mitigation does not negate the problems attributable to Cingular's no return/no refund/ETF policy and its disclosure failures; to the contrary, in certain cases it validates the complainants' allegations.

5.4.2 Informal Complaints to CAB

Between January 1998 and October 2002, CAB received over 1,000 informal complaints by letter or email about one or more of the issues raised in this OII, according to Caceres and CPSD's other primary witness in this area, Janeen Long (Long).²⁵ These informal complaints number either 1,057 (CPSD's contention) or 1,049 (Cingular's contention).²⁶ The nominal difference results from the use of different categorization measures and other assessments by Long and Cingular's Cook, and by others working with them. Cingular contends that these numbers are still too high by about 300 because they should not include complaints made before January 1, 2000 or after June 6, 2002, the date the OII issued. CPSD responds that the OII specifically charged it to continue investigating complaints. Cingular also contends that the existence of informal complaints like these cannot be deemed proof of utility wrongdoing.

Long presents several matrices in her prepared testimony (opening and reply) that indicate the subject of an informal complaint, such as coverage problems (e.g., no reception), capacity problems (e.g., dropped calls), concerns about the ETF, or misrepresentation. The prepared rebuttal of Cingular's Cook also includes a matrix, which adds information such as the minutes of use accrued prior to cancellation of the contract and whether Cingular ultimately waived the ETF or offered the customers other credits. Cook's testimony stresses that Cingular waived nearly $118,000 in ETF charges and awarded more than $50,000 in other credits to these 1,000 + customers who sent informal complaints to CAB. She calculates that of the approximately 900 of these customers who cancelled their service, ultimately only 117 were charged the full ETF by Cingular and another 167, a partial ETF. Again, Cingular reports no information about the agents' ETF charges, and review of the matrix shows that some customers waited out part of the contract to avoid this additional ETF.

We agree with Cingular that absent corroborating information such as may be gleaned by further investigation to assess the credibility of the complainant and the nature and circumstances of the grievance alleged, an informal complaint to CAB should not be accorded the same weight as a declaration or affidavit, since an informal complaint is not a sworn statement. However, Cingular's own evidence, such as Cook's matrix and the Telegence notes provided for a subset of the informal complaints, verify some of the pertinent facts alleged in them, such as the nature of the customer's concerns, the response provided by Cingular or its agents, and the corresponding dates. Evaluated in this way, these informal complaints tend to corroborate the sworn testimony of the customer witnesses. We need not address the parties' disagreement over the total number of "related" informal complaints, as our penalty calculation in Section 6.2 of today's decision is not assessed on a per customer basis.

5.4.3 UCAN Complaint Database

The 22 customer witnesses UCAN produced were drawn from 74 complaints.²⁷ The other 52 complaints, like the informal CAB complaints, are unverified. The complainants' names and contact information, and the specifics of their grievances (which repeat the coverage and capacity issues described above), were all made available to Cingular. The fact that some of the complaints were lodged with UCAN after Shames appeared in a televised news story about the Commission's issuance of this OII does not render the complaints suspect per se, though Cingular appears to suggest it should. More importantly, while Cingular challenges some of the details asserted by the complainants, it does not seriously undermine their basic allegations. While we do not accord these complaints the same weight as sworn testimony, we find them generally credible and again observe that they tend to corroborate the sworn evidence offered by the customer witnesses.

5.4.4 UCAN's Deadzone Project

UCAN argues that its Deadzone Project provides additional proof of Cingular's coverage and capacity problems. Beebe's opening prepared testimony states that this "interactive" database that UCAN maintains "allows cellular customers in the San Diego area to submit the name of their service provider and where they have experienced dead zones in service coverage." (Ex. 200, p. 5.) UCAN reports that from August 31 to September 29, 2002, Cingular subscribers reported 487 dead zones, more than for any other carrier and about 200 more than the second ranked carrier. Cingular's Rodriguez, challenging the usefulness of the database, testified that when she pulled up UCAN's website on the Internet, she was able to enter her address in the Sierra Nevada foothills into the Deadzone Project database. Though Rodriquez testified that she does not have Cingular coverage at home, neither does she live in San Diego. While no doubt some (and possibly most) of the entries in the database represent San Diego area dead zones, we give little weight to the database in this proceeding.

5.4.5 Complaints to the Attorney General's Office

Attachment 6 to Long's reply testimony contains copies of letters and emails, some with attachments, complaining about Cingular that 12 different customers sent to the California Attorney General's Office. These complaints bear dates between November 2001 and February 2003. Three of the 12 customers also wrote to the Commission, but only one of them appears in the tallies of CAB complaints discussed above. Since the record contains no evidence that directly corroborates the 11 new complaints, Cingular argues that these unverified allegations have no evidentiary value.

We disagree. We recognize that CPSD did not reveal these complaints until it distributed its reply testimony two weeks before hearing, and we do not know whether the Attorney General disclosed these complaints to Cingular before that time. We do not condone "sandbagging"; the obvious fairness issues aside, it tends to undermine useful record development, which actually disadvantages all parties. CPSD should have ensured that Cingular was aware of these complaints in order to permit timely and useful investigation of them. However, we note that over half of the complaint letters are addressed to Cingular, list Cingular as a recipient of a copy or attach copies of prior correspondence with Cingular-on balance, this suggests that Cingular knew or should have known that many of these customers were extremely dissatisfied. Nonetheless, Cingular provided no vindicating evidence at hearing with respect to these customers.

5.4.6 Cingular's Cross Streets Records

CPSD argues that some 144,000 "trouble tickets" generated by Cingular's Cross Streets software program beginning in early 2000 constitute another source of complaints and evidence Cingular's violation of law. Cingular's customer service representatives have been using this computer program since the first quarter of 2000 to create an electronic record, termed a trouble ticket, when a customer calls about network coverage and capacity problems such as no service, dropped calls or a continuous system busy signal. These calls come in on Cingular's "611" or "800"numbers, which provide a direct line to the customer service department. If a customer service representative is able to resolve a problem by providing information to the customer, the trouble ticket is closed. Otherwise it is referred elsewhere within the utility until resolved (this is termed an escalation).

Attachment 6 to the opening testimony of CPSD's Long, titled "Cross Streets Query Results," contains a sample printout from Cross Streets that includes the following information for each trouble ticket in the sample: trouble ticket status, identification number, customer's phone number, date/time and nature of problem reported, location, phone manufacturer, etc. Long's opening testimony aptly describes Cross Streets as a "bridge between Customer Service and the Network Engineering Department, enabling customers to be informed about dead spots, areas of no coverage, and closest working and planned cell sites..." and also providing network engineers with real time reports about actual network problems. (Ex. 8, p. 11.) The rebuttal testimony of Cingular's witness Kathleen M. Lee (Lee), Cingular's Network Sales and Network Issues Manager for the West Region, explains that Cross Streets permits Cingular:

to disseminate the information concerning each trouble ticket to the RF [radio frequency] engineering teams for explanation and/or resolution. The explanation and/or response is then uploaded by my team into the Cross Streets program so that CSRs [customer care representatives] have access to that information to relay to the customer. (Ex. 402, p. 5.)

Cingular strongly contests CPSD's characterization of these 144,000 plus trouble tickets as individual customer complaints for which the utility should be penalized. CPSD concedes that these trouble tickets have not been reviewed to eliminate duplication (which may occur when a customer makes more than one call about the same problem, for example).²⁸ Cingular also argues that it would be both unfair and counterproductive for the Commission to levy a penalty for use of a tool with such clearly beneficial customer service applications.

Again, because we do not assess a penalty on a per customer basis, we need not address these contentions. As Cingular admits, one purpose of the software is to track network problems, and even a cursory review of the Cross Street printouts in the record indicates that many trouble tickets report network coverage and capacity problems of the kind described in other evidence.

5.4.7 Internet Petition Signatories

The Staff Report claims that 4,953 electronic signatures to a petition on an Internet website provide another credible source of complaints against Cingular. CPSD has included a printout of the petition, entitled "Cingular Wireless: Petition for Better Service" as Attachment G to the Staff Report. The petition purports to express the views of "the paying citizens of not just California, but ALL paying customers of Cingular Wireless" and the electronic signature portion appears to permit four entries: name, comments, a "yes" or "no" response to the question, "Are You Satisfied With Cingular Wireless?" and a text response to the question, "What Do You Want Cingular To Do?" (Ex. 1, Attach. G.)

While many of the comments briefly refer to the kinds of network coverage and capacity problems we have already seen, without more they do not establish a nexus between such problems and Cingular's no return/no refund/ETF and limited disclosure policies. Other comments have nothing at all to do with wireless service but instead include obscenities, sexual innuendo, and political commentary.

Also troubling is the relative anonymity of all of these electronic signatures. Some include only first names and none includes information (email addresses, telephone numbers, etc.) that permits even rudimentary verification that the signatories are or were Cingular customers in California. While some of the signatories may in fact have signed to register genuine concerns,²⁹ we have no choice but to assign very little evidentiary weight to the petition.