IN THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF PENNSYLVANIA
July 20, 2011
GLASSHOUSE SYSTEMS, INC., PLAINTIFF
INTERNATIONAL BUSINESS MACHINES CORPORATION, DEFENDANT
The opinion of the court was delivered by: Anita B. Brody, J.
July _20__, 2011
FINDINGS OF FACT AND CONCLUSIONS OF LAW
On June 17, 2008, GlassHouse Systems, Inc. ("GlassHouse") filed this action against International Business Machines Corp. ("IBM") and asserted six claims: promissory estoppel, breach of fiduciary duty, negligent misrepresentation, equitable estoppel, intentional interference with prospective business advantage, and unjust enrichment. Compl., ECF No. 1. IBM moved to dismiss all of the claims pursuant to Federal Rule of Civil Procedure 12(b)(6). Mot. Dismiss, ECF No. 4. On March 16, 2009, I granted this motion in part, dismissing all counts except the promissory and equitable estoppel claims. Mem., ECF No. 15. After discovery closed, IBM moved for summary judgment on both estoppel claims. Mot. Summ. J., ECF No. 40. On October 19, 2010, I granted the motion in part, dismissing only the equitable estoppel claim. Mem. & Order, ECF Nos. 51-52.
From May 25 to June 1, 2011, I held a bench trial on the remaining promissory estoppel claim. Both sides submitted proposed findings of fact and conclusions of law, ECF Nos. 95-96, and presented additional oral argument on June 30, 2011. I exercise diversity jurisdiction pursuant to 28 U.S.C. § 1332.
I.FINDINGS OF FACT
1. GlassHouse is a Vermont corporation with its domestic headquarters located in Illinois. IBM is a New York corporation headquartered in New York.
2. IBM is a major technology company that manufactures computer products and sells services related to those products. For some of its products and services, IBM contracts with independent businesses to market and sell to end-user customers ("end-users"). IBM refers to these contractors as Business Partners and the contracts as Business Partner Agreements.
3. The Business Partner Agreements state that IBM approves the Business Partner to actively market IBM products and services at prices and terms that will be set by IBM in future communications.
4. One of IBM's product lines is System z.*fn1 System z products are mainframe computers, i.e., robust computing systems that are able to process many tasks concurrently. They are purchased by large organizations that need systems with significant data capacity and processing speed. The price for a System z mainframe ranges from several hundred thousand dollars to millions of dollars.
5. IBM does not sell or market certain System z products and services directly to end-users. IBM contracts with approximately thirteen Business Partners to do so on its behalf ("zBPs"). z-BPs do not have exclusive geographical territories in which to operate, such that each z-BP is free to market to any end-user.
6. At all relevant times to this action, GlassHouse has been a z-BP. The parties periodically renewed their Business Partner Agreement. Def.'s Ex. 28.*fn2
A. IBM's Pricing Process
7. z-BPs receive prices from IBM in two ways*fn3 by referencing an IBM publication called the Pricing Guide or through a request form called a special bid. The Pricing Guide is over six hundred pages and IBM alters it throughout the year. z-BPs use a special bid form to request a price when a price is not listed in the Pricing Guide, when a price is listed but states that a z-BP needs special approval to use the price, or when a z-BP wants to request a lower price than what is listed.
8. In May 2002, IBM introduced a discount program for its z-BPs called Opportunity Development Pricing ("ODP").
9. The ODP program was ostensibly aimed to reward z-BPs that "performed significant sustained sales and marketing activities to successfully" persuade an end-user to adopt System z. Def.'s Ex. 46 at 12. It was in IBM's interest to develop this program to motivate z-BPs to spend significant unpaid time marketing products.
10. z-BPs were told that only a z-BP that had expended sufficient effort and resources into developing a sales opportunity could receive an ODP discount.
11. The introduction of ODP did not mean that z-BPs could now gain exclusive ownership of a client's account. More than one z-BP could earn ODP for the same sales opportunity. Since ODP was introduced, there have been many times in which IBM has given ODP to more than one z-BP for the same opportunity.
12. In 2006 and 2007, the ODP discount was twenty-one percent.
13. The program allowed z-BPs to assume that they would be entitled to ODP when they were developing a sales opportunity. T.T. 4, 26:7-16, 28:20-21; Pl.'s Ex. 133 at 4.
14. The concept of z-BPs needing to earn ODP predictably led to disputes amongst z-BPs regarding whether a competitor deserved the discount. IBM attempted to reduce the number of these conflicts. When more than one z-BP wanted to claim ODP on the same opportunity, IBM would conduct a review of the z-BPs' marketing documentation to determine whether one, both, or neither had legitimately earned the discount. In its review, IBM only considered marketing activities in the preceding twelve months.
15. A review could be initiated in one of two ways.*fn4
First, if two or more z-BPs submitted a special bid form
claiming ODP for the same sales opportunity, IBM's certification team
would conduct a "certification review." Second, a z-BP could ask its
designated IBM representative for an "ODP review" when the z-BP had
directly been informed by a competitor that it intended to seek ODP on
the same sales opportunity. This information was to be obtained
through a process called "self-policing."
16. One of the rules of the ODP program was that the thirteen z-BPs engage in self-policing amongst themselves. If a z-BP suspected that another z-BP might be discussing the same sales opportunity with an end-user, the first z-BP should contact the second z-BP directly to ask whether the second z-BP intends to request an ODP discount. If the second z-BP indicates that it will also seek ODP, either or both z-BPs would contact IBM to conduct an ODP review.
17. IBM asked its z-BPs to self-police, but IBM did not monitor the z-BPs' interactions and IBM had no threat of punishment for z-BPs who did not respond to competitors' inquiries or, worse, did so disingenuously. There was also no system in place to ensure that z-BPs received notice that a competitor was working with the same client.
18. If a z-BP was denied ODP pursuant to either type of review, that z-BP was still permitted to work with and sell to the end-user; however, a losing z-BP is at a significant disadvantage if it must compete with a z-BP that has been rewarded the substantial twenty-one percent discount.
19. IBM explained these rules and reviews to z-BPs as its way to prevent one z-BP from entering an opportunity late in negotiations to steal a sale away from the z-BP who had invested the substantial time and unpaid marketing activities necessary to cultivate an end-user's interest in System z mainframes. In other words, only a z-BP that has actually invested resources into developing an opportunity could enjoy the fruits of its labor. Self-policing and ODP reviews were also meant to allow z-BPs to discover and resolve conflicts earlier, before multiple z-BPs redundantly expended significant effort on the same sales opportunity. In practice, however, the procedures were too haphazard to always achieve these goals.
20. IBM documented little of its ODP procedures in the Pricing Guide. Rather, IBM orally conveyed its policies to z-BPs one-on-one and as a group at annual advisory council meetings. As a result, IBM employees who did not specialize in working with z-BPs were unfamiliar with the details of these procedures.
21. From its inception in 2002 and until January 2008, material details of the ODP program were left undefined. In particular, IBM never announced a policy regarding the length of time that an unfavorable ODP review decision was effective. T.T. 5, 22:16-24:23, 60:14-61:15; Def.'s Ex. 5 at 2; Pl.'s Ex. 133 at 6. It was also completely unclear whether or for how long a z-BP would be bound by its response that it did not want to seek ODP on an opportunity--in self-policing or in declining to participate in an ODP review. T.T. 1, 84:9-13; T.T. 5, 25; Nassif Dep. 66:12-67:9, Apr. 8, 2011; Def.'s Ex. 18 at 4.
22. These topics were often debated at advisory council meetings.
23. GlassHouse Senior Vice President Joseph Zozzaro ("Zozzaro") testified that he remembers IBM employee Lesli Yenni announcing a new rule at the advisory council meeting held November 9-10, 2006: that the loser of a review would be locked out from trying to claim ODP for ninety days. T.T. 1, 75:4-76:5, 79:11-22, 84:2-6. Every other attendee that credibly testified regarding this meeting directly contradicted Zozzaro's recollection, including Lesli Yenni.*fn5 T.T. 2, 164:1-167:16; T.T. 3, 38:9-39:21, 126:5- 128:7; T.T.5, 42:24-44:1. There is absolutely no record of this statement in the presentations prepared in advance for the meeting, in any of the notes taken by multiple attendees, or in subsequent e-mails summarizing what happened at the meeting. Def.'s Exs. 36-39, 116-119, 124. Noticeably absent from the parties' numerous exhibits is a single document in which anyone mentions a ninety-day rule in any context before August 14, 2007, the date that GlassHouse now insists a ninety-day lock-out began. While I find Zozzaro generally to be a very credible witness, on this matter I do not credit his testimony. I find that IBM did not announce a ninety-day lock-out period in November 2006.
24. Rather, testimony and contemporaneous documents show that while most System z IBM representatives and z-BPs assumed that declining to participate in an ODP review prevented a z-BP from later claiming ODP on the same sales opportunity, IBM made no announcement or endorsement of a defined lock-out period until after the events leading to this lawsuit unfolded. T.T. 1, 84:9-13; T.T. 2, 162:1-163:5; T.T. 5, 25:3-15; Pl.'s Ex. Nos. 44 at 1, 52 at 2-3, 122; Def.'s Ex. 18 at 4; Nassif Dep. 58:18-61:19.
B. GlassHouse's Relationship with End-User SEI Investments, Inc. 25. In March 2006, GlassHouse sold end-user SEI Investments, Inc. ("SEI") three used System z mainframes.
26. In September 2006, IBM announced that it would no longer provide upgrades for the type of mainframes that SEI had purchased. GlassHouse informed SEI; SEI was naturally upset to learn that it had purchased expensive equipment that would have to be replaced--rather than upgraded--as its data needs grew.
27. Beginning in October 2006, GlassHouse began trying to persuade SEI to replace one or more of the used mainframes with newer System z equipment. SEI would need more data capacity, and GlassHouse advised them of different options that would meet their needs in upcoming years. In addition to sending e-mails and making telephone calls, GlassHouse met with someone from SEI to market IBM technology in October 2006, February 2007, March 2007, and early July 2007. Def.'s Ex. 151 (time line of GlassHouse's marketing activities).
28. Throughout all of these marketing efforts, GlassHouse was unaware of whether other zBPs were or were not marketing to SEI. T.T. 2, 38:19-23, 43:6-25.
29. On or before August 7, 2007, SEI contacted IBM about its frustration with GlassHouse and asked IBM how SEI could purchase equipment without working with GlassHouse. On August 7, IBM provided SEI with a list of IBM's other z-BPs. On August 10, an IBM representative met with SEI about its dissatisfaction with GlassHouse.
30. During the week of August 12, 2007, SEI began discussing mainframes with z-BP Mainline Information Systems, Inc. ("Mainline") because it was displeased with its relationship with GlassHouse. Moscoe Dep. 24-25, Jan. 29, 2010; see Pl.'s Exs. 24, 30.
31. In deposition testimony, SEI representatives explained that SEI was very frustrated with GlassHouse's level of expertise, GlassHouse's failure to retain information about SEI, and GlassHouse's history of changing price quotes. Moscoe Dep. 22-32; Toomey Dep. 14:6-15:1, Jan. 29, 2010. SEI had also contacted a second z-BP in hopes that competition would result in a lower price.
32. On August 13, 2007, after hearing a rumor that Mainline was speaking to SEI, Zozzaro sent a self-policing e-mail to Mainline, asking whether Mainline intended to use ODP pricing for SEI. At the time of this e-mail, GlassHouse had no knowledge of the extent of Mainline's prior involvement with SEI or whether it had engaged in marketing activities worthy of ODP. T.T. 2, 16:18-25, 42:9-43:25; Def.'s Ex. 60 at 1.
33. On August 14, 2007, after more than twenty-four hours passed without a response from Mainline, Zozzaro requested that IBM conduct an ODP review.
34. That same day, IBM contacted Mainline to determine whether it intended to claim ODP. Mainline told IBM that it understood that it did not qualify for ODP pricing. IBM forwarded this information to GlassHouse and concluded that an ODP review was not necessary.
35. On August 16, 2007, SEI discussed mainframe prices with Gartner Consulting, which advised SEI that GlassHouse's prices were too high.
36. Thereafter, from August 17 to 31, 2007, SEI continued to negotiate and intensely work with Mainline.
37. Then, on September 7, 2007, GlassHouse was notified by the IBM certification team that it was initiating a certification review because more than one z-BP had submitted a special bid claiming ODP for the SEI opportunity. The other z-BP was Mainline.
38. An upset Zozzaro immediately began e-mailing IBM employees requesting that the review be cancelled due to Mainline's August 14 response to self-policing. The IBM employees that worked with z-BPs agreed that Mainline should not be considered for ODP. Pl.'s Exs. 44, 52, 54; T.T. 5, 70:2-7.
39. The certification team disagreed, and the two groups of IBM employees argued internally about whether the review should continue. Ultimately, the certification team continued with the review.
40. On September 13, 2007, the certification team announced its review decision that both GlassHouse and Mainline had earned ODP on the SEI opportunity.
41. Both z-BPs submitted proposals to SEI. SEI decided to work with Mainline despite the fact that its bid was more than $100,000 higher than the price offered by GlassHouse.
42. In October and November 2007, pursuant to Zozzaro's continuing complaints about the review decision, IBM held meetings to discuss GlassHouse's belief that it had been treated unfairly.
43. The heated dispute between IBM employees about whether a certification review should be done demonstrated how problematic it was that only a select few IBM employees were well-versed in the ODP procedures IBM communicated to its z-BPs. Presumably in recognition of the need to memorialize its oral communications, in October 2007, IBM began attempting to document current ODP procedures.
44. IBM also solicited internal and external feedback to develop better ODP policies to prevent a similar controversy from occurring in the future. IBM apparently realized that its hands-off approach to self-policing obliterated the purpose of having z-BPs self-police at all because, in January 2008, IBM began requiring self-policing to be done in writing with copies sent to an IBM employee. In January 2008, IBM also announced a rule that a z-BP could not challenge an "ODP decision" until ninety days had passed, regardless of whether the decision was pursuant to a review or due to a z-BP stating that it would not seek ODP in response to a self-policing inquiry.
45. After the disputed 2007 sale, SEI continued to work with Mainline and purchase additional products and services through them.
In New York,*fn6 promissory estoppel requires proof of a clear and unambiguous promise, reasonable and foreseeable reliance on the promise by the party to whom it was made, and injury sustained in reliance on the promise. Schwartz v. Miltz, 77 A.D.3d 723, 724 (N.Y. App. Div. 2d Dep't 2010); Thome v. Alexander & Louisa Calder Found., 70 A.D.3d 88, 104-05 (N.Y. App. Div. 1st Dep't 2009); Binkowski v. Hartford Acc. & Indem. Co., 60 A.D.3d 1473, 1475 (N.Y. App. Div. 4th Dep't 2009); Clifford R. Gray, Inc. v. LeChase Constr. Services, 51 A.D.3d 1169, 1170 (N.Y. App. Div. 3d Dep't 2008).
GlassHouse claims that when IBM awarded both z-BPs the ODP discount in September 2007, IBM reneged on its promise to exclude Mainline from ODP on the SEI sale. GlassHouse contends that IBM had promised z-BPs a meaningful period of ODP exclusivity after self-policing or an ODP review decision and, beginning in November 2006, that this period was ninety days. Pl.'s Proposed Findings 42. GlassHouse argues that its marketing activities for SEI were done in reliance on these promises. Id. at 43-44. Its injury was the loss of the SEI account. GlassHouse's argument presents two theories of the relevant promise.
A. GlassHouse's Theory of Exclusivity for a Meaningful Period of Time
GlassHouse's first narrative of the relevant promise is that at some unspecified time before 2006, in an unspecified way, IBM as an entity made its z-BPs believe that IBM would "lock out" a z-BP from claiming ODP for a "meaningful" period of time after either IBM had decided through a review that a z-BP was not entitled to ODP on a specific opportunity or when a z-BP said it would not seek ODP in self-policing. GlassHouse utterly failed in its effort to base a promissory estoppel claim on its contention that IBM promised to give GlassHouse a period of ODP exclusivity for a "meaningful" period of time in the circumstances that unfolded in August 2007.
GlassHouse failed to show factually that IBM made such a promise. GlassHouse did not present evidence of IBM announcing that it would enforce a lock-out for a meaningful period of time, nor could it show circumstantial evidence that IBM enforced lock-out periods in the past so that I could infer such a promise had been made. Even if a promise had been made, legally the promise is too unclear and ambiguous to be enforceable.
The subjective terms of the promise plague the other elements of the claim as well. Any reliance by a business on such a nebulous standard would have been entirely unreasonable. It is also impossible to assess whether the promise was broken--and thus caused GlassHouse's injury--because I do not know whether three weeks is a "meaningful" period of time. Indeed, by its own terms the alleged promise left this determination to IBM's discretion.
B. GlassHouse's Theory of a Ninety-Day Lock-Out Period
GlassHouse's more specific characterization of the relevant promise is that at the advisory council meeting in November 2006, IBM representative Lesli Yenni told z-BPs that IBM would enforce a ninety-day lock-out period on a z-BP after it was found ineligible for ODP by IBM in an ODP review or after a z-BP told another z-BP that it did not want ODP during self-policing. The resulting theory of promissory estoppel, that IBM had promised to exclude Mainline for ninety days once Mainline responded in the negative to self-policing, also fails on all three elements.
1. IBM's Promise
GlassHouse did not prove that IBM made a clear and unambiguous promise that it would bind other z-BPs to their responses in self-policing. Specifically, GlassHouse could not prove that, prior to 2008, IBM promised that z-BPs who were denied ODP--through self-policing or through a formal review decision--were "locked-out" for any particular amount of time. Findings of Fact ¶¶ 21-24. GlassHouse succeeded only in showing that many z-BPs and IBM employees had assumptions and expectations about what should happen or what would be fair. Id. It has not shown that IBM ever communicated a promise regarding these matters. Promissory estoppel cannot be used to enforce a plaintiff's expectations about what would have been "fair;" a court can only enforce a clear and unambiguous promise that a defendant actually made.
2. GlassHouse's Reliance
GlassHouse has not proven that it actually relied on a promise, and I also conclude that it would have been unreasonable for it to do so. GlassHouse failed to show a causal relationship between the alleged promise and its marketing activities to SEI. GlassHouse's marketing before and after the date of the alleged promise were of the same character and frequency. Id. ¶ 27. For it to show reliance that was induced by a promise, GlassHouse's behavior must have been changed by the promise in some way. The evidence shows that GlassHouse's marketing activities did not increase in number or in intensity after the November 2006 advisory council meeting at which the promise was allegedly made. An oral promise that does not actually induce reliance by the promisee need not be enforced against the promisor.
Furthermore, given the constant threat that any z-BP could be working with any client at any time, any reliance related to a lock-out promise would not have been reasonable until a specific lock-out of a particular z-BP had begun. z-BPs had no territorial restrictions, could compete against each other for the same clients, and could even both earn ODP on the same sales opportunity. Id. ¶¶ 5, 11. IBM had no system for z-BPs to know whether they were speaking to the same clients. Id. ¶ 17. Therefore at any time in the months of GlassHouse's marketing activities with SEI in 2006 and 2007, any of twelve other z-BPs could also have been working with SEI and also assuming ODP without GlassHouse's knowledge. GlassHouse never asked SEI if it was working with another z-BP. Id. ¶ 28. When GlassHouse sent its August 14, 2007 self-policing e-mail to Mainline, it had no way of knowing how Mainline would respond or whether Mainline had legitimately earned ODP on the SEI account. Id. ¶ 32.
Any marketing activities expended before August 14, 2007 were done in reasonable reliance that GlassHouse would receive ODP, but they were not reasonably done in reliance on an expectation that other z-BPs would be denied ODP. This distinction is crucial because it means that GlassHouse relied on a promise that IBM kept when it gave GlassHouse ODP. GlassHouse could not expect that any other z-BPs would be locked-out of the SEI opportunity until GlassHouse received a self-policing response from Mainline on August 14. Then, and only then, could GlassHouse have expected to have a recess from competition without another z-BP, and even then it would only be safe from competition with Mainline in particular. It follows that GlassHouse could not have done anything in reliance on IBM's alleged promise until August 14. This is a significantly shorter time period in which reliance was possible, and GlassHouse did not demonstrate evidence of reliance in the three weeks following August 14.
3. GlassHouse's Resulting Injury
Finally, GlassHouse was unable to prove that its injury was caused by IBM. There is ample evidence of SEI's pre-existing displeasure with GlassHouse and its impatient desire to end its relationship with GlassHouse. Id. ¶¶ 29-31, 35. Indeed, SEI decided to work with Mainline despite the fact that Mainline's bid was more than $100,000 higher than the GlassHouse bid. Id.
¶ 41. Even if IBM had enforced a ninety-day lock-out period against Mainline, I have no reason to believe that SEI would have purchased the equipment from, or otherwise continued to work with, GlassHouse.
I conclude the following:
* IBM did not make a clear and unambiguous promise to give GlassHouse discounted pricing to the exclusion of others.
* GlassHouse did not rely on a promise of exclusivity when it marketed to SEI, nor would it have been reasonable for GlassHouse to do so.
* GlassHouse's loss of the SEI account was not sustained in reliance on a promise by IBM.
ANITA B. BRODY, J.
Copies VIA ECF on _________ to: Copies MAILED on _______ to: