One of the hardest things to explain to a client (particularly a sophisticated business client) is that some contracts, no matter how carefully drafted and heavily negotiated, may simply be ignored when push comes to shove before a court. For technology lawyers, that conversation often takes place as the client tries to figure out how to retain its key employees or protect its trade secrets and intellectual property through a non-compete clause.

Non-compete agreements present myriad challenges for lawyers and their clients, because their interpretation and enforcement is dependent on so many complex factors external to the language of the contract itself. Whereas the general rule of contract interpretation is that a non-ambiguous contract should be enforced on its terms, the rules for non-compete agreements vary enormously from state to state.

Some states view non-compete agreements as simple contracts subject to all the ordinary rules of interpretation and enforcement; others (most notably California) view them as constitutionally impermissible restraints on trade. New York falls somewhere in between, which makes the New York lawyer’s job that much more difficult.

In New York, “a restrictive covenant will only be subject to specific enforcement to the extent that it is reasonable in time and area, necessary to protect the employer’s legitimate interests, not harmful to the general public and not unreasonably burdensome to the employee.”1

As a practical matter, this gives New York courts wide latitude in determining whether to enforce a particular non-compete provision and, although some post-employment obligation structures are more commonly upheld than others, results can be very difficult to predict.

The recent decision from the Southern District of New York in International Business Machines Corp. v. Visentin,2 denying IBM a preliminary injunction to enforce a non-compete very similar to one it had successfully enforced just a few years earlier, offers a good look at why.

The standard for evaluating post-employment restrictive covenants under New York law is intensely fact based: Rather than simply looking at the language of the contract, courts must look at the course of bargaining that led to it, the burden on the employee (which includes the length of the restriction, its geographic scope and its effect on the employee’s livelihood), the employer’s intent (whether the restriction is designed to protect a “legitimate” interest or simply to reduce competition), whether the restriction is potentially harmful to the public and a host of other factors external to the contract.

These determinations are complex and situational, and to make matters worse they are often addressed without the benefit of a complete record, in the context of an emergency motion for injunctive relief to enforce (or strike) the covenant.

Faced with this complexity, lawyers practicing in the area often rely on “rules of thumb” to give some clarity to their clients: If a non-compete is longer than one year or has no geographic limitations, for example, a lawyer might suggest to the client that it is too onerous. But these rules cannot really be relied on. To take the same example, New York courts routinely uphold unfettered restrictions much longer than one year in the context of the sale of a business or in other situations where the facts require it.3

Notwithstanding these issues, there are a few situations in which courts will typically enforce restrictions on future employment. Naturally, restrictions relating to a former employee’s use of confidential information or the solicitation of clients or employees post-employment are routinely upheld, as are requirements that employees give substantial notice before joining a competitor to aid in the smooth transition of their business matters.

In addition, New York courts will typically uphold restrictions that directly relate to the protection of an employer’s trade secrets, confidential information or other valuable intellectual property; this is usually viewed as a “legitimate interest” under the BDO Seidman test, and it is the most common reason that technology firms employ non-compete agreements.

Unfortunately, the line between protectable and unprotectable business information is not always as clear as it might seem, as IBM recently discovered.

Before BDO Seidman, and for some time after, many technology companies simply expected all of their employees—regardless of position, responsibilities or department—to be bound by some level of non-compete agreement. Once it became clear that the courts would not uphold such across-the-board agreements, the practice in New York changed and lawyers now typically advise technology clients to limit their use of non-compete agreements to key employees with access to sensitive information.


IBM, a sophisticated employer, provides a good example of the right way to use limited non-compete agreements, but even IBM cannot be sure those agreements will survive judicial scrutiny.

In a widely reported decision in 2008, IBM went to court to enforce a non-compete agreement signed by one of its senior tech managers, Mark Papermaster.4

Papermaster had signed a non-compete agreement in connection with his appointment to IBM’s integration and values team (I&VT) in 2006 which required him not to do business with any “significant competitor or major competitor” of IBM, anywhere in the world, for one year after the termination of his employment. In 2008, Papermaster quit IBM and went to work at Apple.

At IBM, Papermaster had been one of the top managers responsible for microprocessor design, largely in the server business. At Apple he was slated to head up the iPhone/iPod division —though Apple had, at that time, recently acquired a microprocessor company.

Despite the seemingly very different jobs, the court found that Papermaster’s work at Apple would likely cause irreparable harm to IBM by the “inevitable disclosure” of the confidential information and trade secrets Papermaster learned at IBM while serving on I&VT and IBM’s technical leadership team.5 The Court thus granted IBM a preliminary injunction enforcing the non-compete.

In 2011, IBM once again found itself in court seeking to enforce a non-compete agreement. IBM must have liked its chances: the employee was another high-level manager, another key member of the “elite” I&VT, and signatory to essentially the same non-compete as Papermaster.

This time the employee, Giovanni Visentin, was going to Hewlett Packard, in many ways a more direct IBM competitor than Apple. But here, the factors lined up differently and IBM lost its preliminary injunction. As a result, Visentin was permitted to begin work for HP without limitation.6

Given the facial similarity of the underlying agreements, and the fact that Visentin and Papermaster were both key managers in the same company, the differences between the two cases are highly instructive.

Type of Manager

On a close reading of the Visentin opinion, the most important difference between Papermaster and Visentin appears to have been the respective courts’ views of their positions at IBM.

Both men were key, high-level technical managers, members of IBM’s most important senior management committee with responsibility for some of the company’s most important business initiatives.

The difference is that Papermaster was a “top expert” in chip design and IBM’s “power” processor architecture, while Visentin was a “business manager, not a technical expert.”7 Of course, that somewhat understates Visentin’s role: He was general manger of a group responsible for up to 9,000 deals and $2.5 billion in revenue a year.

It is safe to assume he had some technical knowledge important to IBM. However the Court plainly did not see Visentin as the kind of employee who would have access to detailed confidential information subject to protection by a non-compete.

IBM, to its credit, tried its best to establish that Visentin possessed sensitive and confidential information vital to its business—information of the kind that can be validly protected by a non-compete under BDO Seidman.

It identified seven areas of sensitive business information to which Visentin had access, including troubled client relationships, deal pricing, client pipelines, IBM’s plans to compete with HP (Visentin’s new employer) and IBM’s “cloud computing” initiative and offerings.

Detailed knowledge in any one of these areas would typically be sufficient to justify enforcement of a limited non-compete. Here, however, the court took extensive testimony, developed a record and found, in its very lengthy opinion, that Visentin’s knowledge of all of these areas was sufficiently “high level” that it could not cause competitive harm to IBM for him to carry it with him (in his head) to HP.

For example, the court found that, although Visentin might know about IBM’s plans for marketing, pricing and rolling out its cloud computing offerings, he would not know the technical details of the server architecture or software packages required to implement them. In other words, because Visentin was too senior to be enmeshed in the technical or operational details, the court found that his knowledge was largely not entitled to protection as a trade secret.

As some commentators have noted, this is a somewhat anomalous result: Essentially, the court found that Visentin was too senior—his knowledge too high-level—to be subject to a non-compete. That is, of course, the opposite of the usual view which holds that more senior, irreplaceable employees are more apt to be subject to non-compete obligations.

Prediction Is a Fool’s Game

Under BDO Seidman and its progeny, the outcome of an emergency motion to enforce a non-compete can be all but impossible to predict. The inquiry is highly fact-specific, and the result can turn on any number of idiosyncratic factors. As in the stock market, past performance is no guarantee of future results.

Consider the difference in tone between the Papermaster and Visentin opinions in describing the integration and values team on which both defendants served.

The Papermaster court describes the team as “an elite group that develops IBM’s corporate strategy” consisting of “300 top IBM executives who are considered to be its key leaders.” It notes that “members of this elite group are appointed by the Chairperson of IBM and include the CEO and all twenty of the company’s senior executives” and that members “had access to highly confidential information, including strategic plans, marketing plans, product development, and long-term business opportunities for IBM.”8

The Visentin court, by contrast, describes the team only as “a leadership group that develops IBM’s corporate strategy…charged with addressing some of the strategic and other important issues facing IBM.”9

As described in Papermaster, it is almost a foregone conclusion that a member of I&VT would have access to information sufficient to make a non-compete enforceable under the BDO Seidman standard; as described in Visentin, the group sounds like little more than a glorified management committee.

These two courts, looking at two different men with two different jobs, had entirely different views of the same key committee—and of the associated non-compete. The chip-designer was bound by his non-compete, the key business leader was not. No one could possibly have predicted the result.

Or could they? In fact, the outcomes of cases at this level are rarely in doubt; high level executives do not sit on the beach for long. Two months after IBM got its injunction prohibiting Papermaster from working at Apple, the parties reached a settlement and off he went to head up Apple’s iPod division. Eighteen months later, well before Visentin ever thought of leaving IBM, Papermaster was already out of Apple and on to Cisco. A year is a long time in the tech world.

Stephen M. Kramarsky, a member of Dewey Pegno & Kramarsky, focuses on complex intellectual property litigation.


1. BDO Seidman v. Hirshberg, 93 N.Y.2d 382, 389, 690 N.Y.S.2d 854, 857 (N.Y. 1999), quoting Reed, Roberts Assocs. v. Strauman, 40 N.Y.2d 303, 307, 386 N.Y.S.2d 677 (N.Y. 1976).

2. 11 Civ. 399 (LAP), 2011 WL 672025 (S.D.N.Y. Feb. 16, 2011).

3. E.g. Sovereign Business Forms Inc. v. Stenrite Industries Inc., No. 00 CIV. 3867 (BDP), 2000 WL 1772599, at *14 (S.D.N.Y. Nov. 28, 2000).

4. IBM Corp. v. Papermaster, 08 Civ. 9078 (KMK), 2008 WL 4974508, at *7 (S.D.N.Y. Nov. 21, 2008).

5. Id. at *10.

6. IBM v. Visentin, 11 Civ. 399 (LAP), 2011 WL 672025 (S.D.N.Y. Feb. 16, 2011). The opinion is “only” a denial of a preliminary injunction, not a trial on the merits, but the court held four days of hearings including live testimony and developed a full record prior to rendering its opinion-as a practical matter, the decision on injunctive relief is likely to be outcome determinative.

7. Id. at *2.

8. Papermaster, 2008 WL 4974508, at * 3.

9. Visentin, 2011 WL 672025, at *1.