`
`
`
`Corey Worcester (pro hac vice)
`coreyworcester@quinnemanuel.com
`Renita Sharma (pro hac vice)
`renitasharma@quinnemanuel.com
`QUINN EMANUEL URQUHART AND SULLIVAN LLP
`51 Madison Avenue, 22nd Floor
`New York, NY 10010
`Telephone:
`(212) 849-7000
`
`Terry L. Wit (SBN 233473)
`terrywit@quinnemanuel.com
`QUINN EMANUEL URQUHART AND SULLIVAN LLP
`50 California Street, 22nd Floor
`San Francisco, CA 94111
`Telephone:
`(415) 875-6600
`
`Adam B. Wolfson (SBN 262125)
`adamwolfson@quinnemanuel.com
`QUINN EMANUEL URQUHART AND SULLIVAN LLP
`865 Figueroa St., 10th Floor
`Los Angeles, CA 90017
`Telephone:
`(213) 443-3000
`
`Attorneys for Plaintiff hiQ Labs, Inc.
`
`
`
`UNITED STATES DISTRICT COURT
`
`NORTHERN DISTRICT OF CALIFORNIA
`
`hiQ Labs, Inc.,
`
`Case No. 3:17-cv-03301-EMC
`
`AMENDED COMPLAINT FOR
`VIOLATIONS OF THE SHERMAN ACT
`(15 U.S.C. §§ 1 AND 2) AND THE
`CLAYTON ACT (15 U.S.C. §§ 15 AND 16)
`
`DECLARATORY JUDGMENT UNDER 22
`U.S.C. § 2201 THAT PLAINTIFF HAS NOT
`VIOLATED: (1) THE COMPUTER FRAUD
`AND ABUSE ACT (18 U.S.C. § 1030); (2)
`THE DIGITAL MILLENNIUM
`COPYRIGHT ACT (17 U.S.C. §1201);(3)
`COMMON LAW TRESPASS TO
`CHATTELS; OR (4) CAL. PENAL CODE
`§ 502(c);
`
`INJUNCTIVE RELIEF TO ENJOIN: (1)
`VIOLATIONS OF THE SHERMAN ACT
`(15 U.S.C. §§ 1 AND 2); (2) INTENTIONAL
`INTERFERENCE WITH CONTRACT
`AND PROSPECTIVE ECONOMIC
`ADVANTAGE; (3) UNFAIR
`COMPETITION (CAL. BUS. & PROF.
`CODE § 17200); AND RELATED
`MONETARY RELIEF
`
`
`
`Case No. 3:17-cv-03301-EMC
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`
`
`
`Plaintiff,
`
`vs.
`
`LinkedIn Corp.,
`
`
`
`
`
`Defendant.
`
`
`
`
`AMENDED COMPLAINT
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 2 of 36
`
`
`
`
`
`
`Plaintiff hiQ Labs, Inc. (“hiQ”), by its undersigned counsel, hereby brings this action
`
`against Defendant LinkedIn Corporation (“Defendant” or “LinkedIn”) and alleges as follows:
`
`INTRODUCTION
`
`1.
`
`hiQ brings this action under Sections 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1
`
`and 2, and under the Declaratory Judgment Act, 28 U.S.C. §§ 2201 and 2202. hiQ seeks to
`
`recover the damages it suffered as a result of LinkedIn’s anticompetitive conduct, as well as obtain
`
`a declaration from the Court that hiQ has not violated and will not violate federal or state law by
`
`accessing and copying wholly public information from LinkedIn’s website. hiQ further seeks
`
`injunctive relief preventing LinkedIn from misusing the law to complete its attempted destruction
`
`of hiQ’s business, and prevent LinkedIn from continuing to strangle competition by giving itself
`
`an unfair competitive advantage through unlawful and unfair business practices.
`
`2.
`
`LinkedIn is the world’s largest professional social network, with over 660 million
`
`members. Particularly relevant here is that LinkedIn rose to dominance in the professional social
`
`networking platform market because it created a social network platform in which users could
`
`publicly post all or portions of their resumes and work history, so that those users could connect
`
`with potential employers or other potential professional contacts. The public portion of LinkedIn
`
`users’ profiles has long been one of the network’s primary selling points, and LinkedIn’s website
`
`and user agreement have long stated that the information users chose to make public remained
`
`theirs to provide to whomever viewed it online. These fundamental representations helped fuel
`
`LinkedIn’s massive growth, which resulted in it obtaining undisputed monopoly power in the
`
`professional social networking platform market.
`
`3.
`
`hiQ, in contrast, is a tech startup nowhere near LinkedIn’s size. hiQ identified an
`
`opportunity for a new kind of “people analytics” service based on the massive trove of public
`
`information individuals chose to share in their professional social networking. By collecting and
`
`analyzing public profile information on LinkedIn, hiQ provided its clients – mostly large
`
`companies – with insights and other data analytics about their employees, such as which
`
`AMENDED COMPLAINT
`
`
`
`
`2
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 3 of 36
`
`
`
`1
`
`employees are most at risk to leave the company or which skills its employees have. This service
`
`2
`
`– the likes of which did not exist before hiQ and constitutes its own unique relevant market today
`
`3
`
`– provides enormous value to both employers and employees, because it creates a way for
`
`4
`
`employers to, inter alia, approach those employees that are at the highest risk of leaving the
`
`5
`
`company in order to either renegotiate their pay and benefit packages (i.e., provide better
`
`6
`
`compensation to the employee), or otherwise address concerns those employees might have with
`
`7
`
`their current employment situation, instead of incurring the significant cost and disruption of
`
`8
`
`replacing a departed employee. Similarly, people analytics help employers identify better
`
`9
`
`positions and potential training for their employees within the company, so that their skillsets are
`
`10
`
`further developed and put to their fullest and best use. Enabling these types of proactive activities
`
`11
`
`results in a win-win situation for all involved. As reported by LinkedIn’s 2020 Global Talent
`
`12
`
`Trends report, “55% of talent professionals say they still need help putting basic people analytics
`
`13
`
`into practice.” hiQ’s services did just that.
`
`14
`
`4.
`
`In order to provide people analytics services to its clients, hiQ does not analyze the
`
`15
`
`private sections of LinkedIn, such as profile information that is only visible when you are signed-
`
`16
`
`in as a member, or member private data that is visible only when you are “connected” to a
`
`17
`
`member. Rather, the information hiQ uses is wholly public information visible to anyone with an
`
`18
`
`internet connection. And, as noted, far from harming LinkedIn members, hiQ’s access to this
`
`19
`
`public information promotes precisely the type of professional and employment opportunities that
`
`20
`
`lead LinkedIn members to make all or even just a portion of their profiles public in the first place.
`
`21
`
`5.
`
`For years, LinkedIn knew about and sanctioned hiQ’s services, because LinkedIn
`
`22
`
`profited from doing so. hiQ increased employer engagement on LinkedIn, because those
`
`23
`
`employers paid more attention to their employees’ LinkedIn public profiles to see when they
`
`24
`
`might leave the company, and/or whether they might be better placed elsewhere within the
`
`25
`
`organization. It also incentivized employers to either directly or implicitly encourage their
`
`26
`
`employees to use LinkedIn in the first place. Employer participation in the LinkedIn network
`
`27
`
`distinguishes that network from other social networks, and, indeed, it is one of the company’s
`
`28
`
`raisons d’etre. As LinkedIn advertises on its website, it is a social network meant to help users
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`3
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 4 of 36
`
`
`
`1
`
`“[f]ind the right job or internship for you,” “post your job for millions of people to see,” and act as
`
`2
`
`the social network for “[a]nyone looking to navigate their professional life.” Moreover,
`
`3
`
`employers, employees, and recruiters often choose to pay LinkedIn subscription and other fees for
`
`4
`
`a variety of reasons. Increased employer and employee engagement through the “free advertising”
`
`5
`
`LinkedIn received from hiQ’s services directly led to higher revenues for LinkedIn.
`
`6
`
`6.
`
`LinkedIn eventually realized that it might be able to profit by providing the same
`
`7
`
`type of innovative and revolutionary analytics hiQ pioneered, and it developed its own competing
`
`8
`
`version of that analytics service. Then, in May 2017, LinkedIn abruptly, unlawfully and without
`
`9
`
`cause denied hiQ access to the portion of the LinkedIn website containing wholly public member
`
`10
`
`profiles. hiQ relies on that public data, available nowhere but LinkedIn, for its data analytics
`
`11
`
`business, which, prior to LinkedIn’s conduct described herein, served well-known, innovative
`
`12
`
`clients including eBay, Capital One, and GoDaddy.
`
`13
`
`7.
`
`More specifically, on May 23, 2017, LinkedIn sent hiQ a cease-and-desist letter
`
`14
`
`ordering hiQ to stop accessing LinkedIn and asserting that hiQ’s continued access to the website
`
`15
`
`violated the Computer Fraud and Abuse Act, Digital Millennium Copyright Act, and California
`
`16
`
`Penal Code § 502(c) and constituted common law trespass to chattels. This came as a surprise to
`
`17
`
`hiQ, given that LinkedIn was aware of hiQ’s activities for several years and never once objected to
`
`18
`
`hiQ’s use of this public information.
`
`19
`
`8.
`
`In an attempt to justify this about-face, which followed years of profitable dealing
`
`20
`
`with hiQ and other people analytics providers, LinkedIn asserted (pretextually) that it needs to
`
`21
`
`protect LinkedIn member data even though LinkedIn members have expressly made that
`
`22
`
`information public and LinkedIn has identified no harm to itself or its members. LinkedIn
`
`23
`
`publicly acknowledges on its own website that public profile data belongs to LinkedIn members,
`
`24
`
`not to LinkedIn, and that each member is free to choose the level of public disclosure allowed for
`
`25
`
`his or her own information. LinkedIn members can choose to (1) keep their profile information
`
`26
`
`private; (2) share only with their direct connections; (3) share with connections within three
`
`27
`
`degrees of separation; (4) allow access only to other signed-in LinkedIn members, or (5) allow
`
`28
`
`access to everyone, even members of the general public who may have no LinkedIn account and
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`4
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 5 of 36
`
`
`
`1
`
`who can access the information without signing in or using any password. It is only this fifth
`
`2
`
`category of information – wholly public profiles – that is at issue here: hiQ only accesses the
`
`3
`
`profiles that LinkedIn members have made available to the general public.
`
`4
`
`9.
`
`LinkedIn’s entire stated complaint is that hiQ “copies” the data its members have
`
`5
`
`made public, but LinkedIn asserts no copyright or other exclusive propriety interest in the data and
`
`6
`
`it clearly has none. Moreover, hiQ does not collect all (or even a substantial proportion) of the
`
`7
`
`member profiles on LinkedIn, nor does it compete with LinkedIn by creating a substitute social
`
`8
`
`network or job posting forum. Rather, hiQ accesses public data for a limited subset of users –
`
`9
`
`usually its client’s employees – and uses scientific methodologies to analyze the information. hiQ
`
`10
`
`then provides its clients with this new, refined data that it produced in a form that is by necessity
`
`11
`
`very different from the public profile pages on LinkedIn.
`
`12
`
`10.
`
`Because LinkedIn has no legitimate copyright claim, it has instead resorted to self-
`
`13
`
`help by actively preventing hiQ from obtaining the public data in LinkedIn users’ profiles through
`
`14
`
`systematic means. LinkedIn also threatened to sue hiQ under federal and state laws pertaining to
`
`15
`
`hacking and unauthorized computer and network access in order to intimidate hiQ and force it to
`
`16
`
`stop accessing these public profiles. But LinkedIn cannot use those laws for an improper purpose
`
`17
`
`to obtain exclusive proprietary control over wholly public data in which it otherwise has no
`
`18
`
`exclusive interest and which hiQ, and anyone else, can freely access on the world wide web with
`
`19
`
`no log-in credentials or password. Indeed, LinkedIn would not have that data on its website in the
`
`20
`
`first place but for its promise to LinkedIn members that they can publicly disclose that information
`
`21
`
`on LinkedIn for all the world to see and use.
`
`22
`
`11.
`
`LinkedIn’s about-face and active attempts to prevent hiQ (and, on information and
`
`23
`
`belief, all other people analytics providers) from obtaining employment data LinkedIn users make
`
`24
`
`public has had massive, immediate consequences for hiQ. Suddenly, the public data source on
`
`25
`
`which all of its and its non-LinkedIn competitors’ people analytics services rely was taken away.
`
`26
`
`This was long after LinkedIn told its members their profile information was their own and locked
`
`27
`
`them into its professional social networking platform based on that promise. LinkedIn’s about-
`
`28
`
`face, which followed years of profitable dealing with hiQ, prevented hiQ and other people
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`5
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 6 of 36
`
`
`
`1
`
`analytics providers from providing the people analytics their clients needed, and those clients –
`
`2
`
`including, particularly, hiQ’s marquee Fortune 500 clients – quickly disappeared. There is no
`
`3
`
`point paying for a service one cannot receive, and there was nothing hiQ could do to prevent the
`
`4
`
`attrition. Unfortunately for competition, this has also meant that, today, LinkedIn has not only
`
`5
`
`obtained monopoly power in that market through anticompetitive means, but has also crowded out
`
`6
`
`hiQ, LinkedIn’s best-in-class competitor – a competitor that created the market in the first place.
`
`7
`
`12.
`
`As a consequence, hiQ lost the majority of its revenues almost overnight and today
`
`8
`
`is unfortunately a shadow of what it once promised to be. This has caused hiQ substantial
`
`9
`
`monetary damages in the form of, inter alia, lost profits and diminution in business value. hiQ
`
`10
`
`intends to prove its damages at trial, but the Court should also enjoin LinkedIn from denying hiQ
`
`11
`
`access to its public employment information database because LinkedIn’s real motivation is
`
`12
`
`obviously anticompetitive: to prevent anyone but LinkedIn from being able to use that public
`
`13
`
`information for data analytics.
`
`14
`
`15
`
`13.
`
`Plaintiff hiQ is a Delaware corporation with its principal place of business in San
`
`THE PARTIES
`
`16
`
`Francisco, California.
`
`17
`
`14.
`
`On information and belief, Defendant LinkedIn is a Delaware corporation with its
`
`18
`
`principal place of business in Sunnyvale, California.
`
`19
`
`20
`
`JURISDICTION AND VENUE
`
`15.
`
`The Court has subject matter jurisdiction over this action under 28 U.S.C. § 1331
`
`21
`
`and 1337, because (a) Plaintiff brings this action under Sections 4 and 16 of the Clayton Act, 15
`
`22
`
`U.S.C. §§ 15 and 16, for violations of Sections 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1 and 2,
`
`23
`
`and (b) Plaintiff’s first and second claims for relief seek a declaratory judgment under 28 U.S.C.
`
`24
`
`§ 2201 and 2202 that Plaintiff has not violated the Computer Fraud and Abuse Act, 18 U.S.C.
`
`25
`
`§ 1030, and the Digital Millennium Copyright Act, 17 U.S.C § 1201.
`
`26
`
`16.
`
`Under 28 U.S.C. § 1367, the Court has supplemental jurisdiction over Plaintiff’s
`
`27
`
`third through tenth claims for relief because they arise out of the same common set of facts and
`
`28
`
`conduct as Plaintiff’s federal claims for relief.
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`6
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 7 of 36
`
`
`
`1
`
`17.
`
`This Court has personal jurisdiction over Defendant LinkedIn in this action
`
`2
`
`because, on information and belief, LinkedIn’s corporate headquarters and principal place of
`
`3
`
`business is within this judicial district, and LinkedIn has engaged in substantial business within
`
`4
`
`this district.
`
`5
`
`18.
`
`Venue is proper in this District pursuant to 28 U.S.C. § 1391(b)(1) and (2) because
`
`6
`
`Defendant LinkedIn conducts substantial business within this District and a substantial part of the
`
`7
`
`acts or omissions giving rise to Plaintiff’s claims occurred in this District. Venue is further proper
`
`8
`
`in this District pursuant to 28 U.S.C. § 1400(a) because this action relates to copyrights and
`
`9
`
`Defendant LinkedIn resides in this District.
`
`10
`
`11
`
`INTRADISTRICT ASSIGNMENT
`
`19.
`
`Pursuant to N.D. Cal. Civil Local Rule 3-2(c), this case is, inter alia, an intellectual
`
`12
`
`property action appropriate for assignment on a district-wide basis.
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`FACTUAL ALLEGATIONS
`
`LinkedIn, the Professional Social Networking Platform Market, and the Public Member
`Profile Portion of LinkedIn’s Professional Social Network Platform
`
`20.
`
`The core of LinkedIn’s business is a professional social network platform that
`
`aggregates the profile information of well over half a billion professionals, their interrelationships,
`
`their posts, and their cross-endorsements. LinkedIn states that the purpose of the service is to
`
`“promote economic opportunity” and provide a place for professionals “to meet, exchange ideas,
`
`learn, and find opportunities....” See Exhibit 1 (LinkedIn User Agreement). For individuals, it is
`
`a place to network and seek out jobs; for companies and recruiters, it is an essential platform to
`
`post jobs and source talent. For all, the focus is on business relationships.
`
`21.
`
`As opposed to other social networks, such as Facebook, Instagram, Twitter, etc.,
`
`LinkedIn’s focus on professional social networking puts it in a separate category. Users
`
`participate in LinkedIn’s network because its purpose is to connect professionals of all stripes, and
`
`to act as a constantly-updated repository of users’ business information. This is different than
`
`other social networks, which often focus on non-business-related aspects of their users’ lives (e.g.,
`
`sharing family photos, posting thoughts or news stories, uploading videos). Thus, the primary
`
`AMENDED COMPLAINT
`
`
`
`
`7
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 8 of 36
`
`
`
`1
`
`service LinkedIn provides its users – the provision of a business-focused social network – is
`
`2
`
`different in purpose and scope from other social networks, and satisfies a different need for its
`
`3
`
`users. In practice, LinkedIn users often have other social network accounts, but they do not
`
`4
`
`typically use those other social networks for business purposes. As a group, professional social
`
`5
`
`networking platforms therefore constitute a separate relevant market, because consumers would
`
`6
`
`reasonably switch to other professional social network platforms for business purposes (if they
`
`7
`
`could), but would not similarly use more traditional social networks for the same purposes.
`
`8
`
`22.
`
`LinkedIn was launched in May 2003 and, within one month, had over 4,500 users.
`
`9
`
`This was approximately the same time period in which Facebook was founded (after MySpace,
`
`10
`
`arguably the first major social networking platform, demonstrated the popularity of this new form
`
`11
`
`of connecting online), but the two then-fledgling social networking platforms had very different
`
`12
`
`purposes. As the concept of social networking caught on in the broader populace, LinkedIn
`
`13
`
`experienced meteoric growth because of the new professional social networking platform market it
`
`14
`
`created. Over the next 15 years, LinkedIn added hundreds of millions of users to its base, now
`
`15
`
`with over 660 million members. Today, LinkedIn is established in more than 200 countries,
`
`16
`
`receives more than 100 million unique visitors a month, and reports that up to 60% of U.S.
`
`17
`
`companies use it to hire staff.
`
`18
`
`23.
`
`One of the main keys to LinkedIn’s initial and then sustained growth was that it
`
`19
`
`acted as a repository and platform for users to share their work history, also known as a curriculum
`
`20
`
`vitae (“CV”) or resume information. LinkedIn member profiles contain resume information such
`
`21
`
`as education, skills, publications, certifications, and employment history. They also contain
`
`22
`
`endorsements from colleagues and others within the member’s professional social network.
`
`23
`
`Members can connect their LinkedIn profiles to those of colleagues around the world.
`
`24
`
`24.
`
`As particularly relevant to this lawsuit, users had – and, as of this filing, still have –
`
`25
`
`the option to publicly share as much or as little of their CV as they want. Most LinkedIn users
`
`26
`
`choose to share at least a portion of their CV, because it allows other professionals to search them
`
`27
`
`out for networking or recruiting purposes, and because it also helps elevate their professional
`
`28
`
`profile in online searches. This “public option” has long been one of LinkedIn’s main value
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`8
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 9 of 36
`
`
`
`1
`
`drivers, because it facilitates the site’s core professional social networking services and is one of
`
`2
`
`the key reasons users first signed up for, and then maintained, their LinkedIn accounts. Without
`
`3
`
`the ability to publicly share CV information, LinkedIn would not be the dominant professional
`
`4
`
`social network platform it is today. Indeed, the entire purpose of a social network is to share
`
`5
`
`information with those in your network, or who wish to join your network. LinkedIn always
`
`6
`
`promoted this core tenet of social networking (although, as discussed below, it has recently
`
`7
`
`reneged on its promises that the information its members choose to share publicly shall remain
`
`8
`
`freely accessible).
`
`9
`
`25.
`
`As LinkedIn grew, so did its power in professional social networking platforms. It
`
`10
`
`was the first mover in the professional social networking platform market, and its explosive
`
`11
`
`growth was self-reinforcing with respect to its market power. A social network derives a major
`
`12
`
`part of its value from the number and scope of users in the network; i.e., the more users in the
`
`13
`
`network and the broader the scope of those users, the more opportunities one has to be social and
`
`14
`
`connect, and the more one is incentivized to be in that network rather than in smaller, less-
`
`15
`
`populated networks. LinkedIn was, for many years, the only real professional social networking
`
`16
`
`platform and became the de facto, dominant standard for such networking. Today, LinkedIn’s
`
`17
`
`collection of profiles is a one-of-a-kind resource. It is the single largest, most up-to-date, and
`
`18
`
`authoritative repository about the world’s professional community. There is no comparable data
`
`19
`
`source anywhere else in the world, and no current likelihood that any other company can craft a
`
`20
`
`similar database of information and users. LinkedIn is the unquestioned, most dominant
`
`21
`
`professional social networking platform in the world, counting, on information and belief, well
`
`22
`
`over 75% of all professional social network users in the United States.1
`
`23
`
`26.
`
`Put simply, LinkedIn has monopoly power in the professional social networking
`
`24
`
`platform market, and it has erected substantial barriers to new entry by any professional social
`
`25
`
`network platform that wishes to rival its power. The first such barrier is simply the number of
`
`26
`
`27
`
`28
`
`
`
`
`1 To the extent professionals are members of other, smaller professional social networks, hiQ
`understands, on information and belief, that they typically also belong to LinkedIn.
`
`AMENDED COMPLAINT
`
`
`
`
`9
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 10 of 36
`
`
`
`1
`
`LinkedIn users that currently exist. As noted above, a social network derives its value (and power)
`
`2
`
`from the number and scope of users in the network; a concept called “network effects.” A new
`
`3
`
`entrant would need to sign up a substantial portion of LinkedIn users for its own platform in order
`
`4
`
`to even begin to take share away from LinkedIn. Second, the information LinkedIn currently
`
`5
`
`houses on its servers is another barrier to entry, because a new rival would need to obtain similar
`
`6
`
`amounts of information about its users in order to take share away from LinkedIn. Third is the
`
`7
`
`cost and logistical expense of creating a worldwide network with the same resources as LinkedIn.
`
`8
`
`This includes not only the front-end website and/or application members use when accessing the
`
`9
`
`social network, but also the massive back-end infrastructure needed to facilitate such large
`
`10
`
`networking operations, and to generate members’ ongoing trust in the network’s reliability and
`
`11
`
`security.
`
`12
`
`27.
`
`LinkedIn grew to what it is today by telling its prospective members that their
`
`13
`
`employment information would always remain their own, even if they include that information in
`
`14
`
`their member profiles. In its User Agreement, LinkedIn admits it does not own the data that its
`
`15
`
`members decide to share publicly on LinkedIn’s website. LinkedIn explains to members that “you
`
`16
`
`own the content and information that you submit.” See Exhibit 1 (LinkedIn User Agreement) at
`
`17
`
`3.1. LinkedIn is unequivocal that members control their profiles: “You control the visibility and
`
`18
`
`reach of your LinkedIn profile.” See Exhibit 2 (Public & Private Profiles | LinkedIn Help).
`
`19
`
`LinkedIn gives members the ability and right to specify which portions of their profiles will be
`
`20
`
`visible to their direct connections, their network (those within three degrees of separation), all
`
`21
`
`LinkedIn members, and the “public.” The “public” setting (which is at issue here) allows
`
`22
`
`“[a]nyone [to] see [members’] public profile[s] in search engines, as well as in apps and other
`
`23
`
`services.” Exhibit 3 (showing public profile settings). Public profiles may be reached via third-
`
`24
`
`party services (e.g., Google and Bing) and directly via a web address (URL) that LinkedIn creates
`
`25
`
`for its members. Thus, LinkedIn acknowledges that a public setting will permit access for anyone
`
`26
`
`in the world with an internet connection. The Privacy Policy advises members: “Your profile is
`
`27
`
`fully visible to all Members and customers of our Services. Subject to your settings, it can also be
`
`28
`
`visible to others on or off of our Services (e.g., Visitors to our Services or users of third-party
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`10
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 11 of 36
`
`
`
`1
`
`search engines.” Exhibit 6 (LinkedIn Privacy Policy) at 3.1.
`
`2
`
`28.
`
`By encouraging the proliferation of public employment information in order to
`
`3
`
`exponentially grow its professional social network (and therefore lock its users into that network),
`
`4
`
`LinkedIn has effectively become a public forum where professionals can meet and exchange ideas,
`
`5
`
`information, and news articles. LinkedIn describes itself as a “community” and users are able to
`
`6
`
`post publicly, share other members’ posts, and comment on other members’ posts. LinkedIn
`
`7
`
`cannot – consistent with the free speech clause of the California Constitution – selectively exclude
`
`8
`
`hiQ from this “public” forum, even if LinkedIn’s servers are considered “private” property. The
`
`9
`
`California Supreme Court has definitively interpreted these constitutionally guaranteed free speech
`
`10
`
`rights as precluding an owner of private property from prohibiting such access when the property
`
`11
`
`has been opened to the public and constitutes a public forum. The United States Supreme Court
`
`12
`
`has in turn upheld this California constitutional right as against a challenge that it amounts to a
`
`13
`
`“taking” of private property under the United States Constitution. LinkedIn cannot have it both
`
`14
`
`ways even on its own web servers: promising a public forum and public access on the one hand,
`
`15
`
`and then selectively excluding members of the public on the other.
`
`16
`
`29.
`
`Since its launch in 2003, LinkedIn has created numerous successful revenue
`
`17
`
`streams, including selling services to corporations that help with their recruiting and sales
`
`18
`
`processes, memberships to corporate and individual users, targeted advertising based on members’
`
`19
`
`usage of the site, and others. All of this follows on LinkedIn’s status as the world’s largest and
`
`20
`
`most recognizable professional social networking platform, and derives from its key decision early
`
`21
`
`on to allow members to make any portion of the member profile public. As of hiQ’s launch in
`
`22
`
`2012, LinkedIn’s annual revenues were on the order of nearly $1 billion, a number that had nearly
`
`23
`
`quadrupled by the end of 2016. In late 2016, LinkedIn was purchased by Microsoft Corporation
`
`24
`
`for $26 billion.
`
`25
`
`26
`
`27
`
`28
`
`
`
`hiQ Labs Creates a New “People Analytics” Market Based on Professional Social
`Networkers’ Publicly-Shared Information
`
`30.
`
`hiQ was formed in July 2012 and has since raised $14.5 million in multiple rounds
`
`of funding. It once had 24 employees, the majority of whom were in its San Francisco office, and
`
`AMENDED COMPLAINT
`
`
`
`
`11
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 12 of 36
`
`
`
`1
`
`11 of whom had advanced degrees, including several PhDs. Today, however, hiQ’s workforce has
`
`2
`
`been reduced to just one employee, because its business unfortunately cratered as a result of
`
`3
`
`LinkedIn’s anticompetitive acts described herein.
`
`4
`
`31. When it was founded, hiQ identified a previously-unmet need among employers
`
`5
`
`(particularly large employers, such as Fortune 500 companies) to identify those employees that
`
`6
`
`were at the highest risk of leaving the company, as well as uncover the full scope of current and
`
`7
`
`potential employees’ skills. Although one would expect this latter task to be the simpler of the
`
`8
`
`two, it has proven an elusive goal over the years, because employees often do not know to share
`
`9
`
`their full skillsets with their employer, or do not know that skills they possess might warrant a
`
`10
`
`different or better job (i.e., better “fit”) within the company.
`
`11
`
`32.
`
`hiQ’s innovation was the realization that the facts employees choose to share about
`
`12
`
`themselves publicly in a professional social network strongly predict their employment
`
`13
`
`satisfaction and also demonstrate what skills they can bring to current and future jobs. By
`
`14
`
`utilizing modern technology to analyze this data, employers can eliminate inefficiencies in their
`
`15
`
`organizations and also reduce transaction costs for themselves and their employees by highlighting
`
`16
`
`which highly-valued employees are most at risk of leaving the company. Employees seeking new
`
`17
`
`jobs must spend time and effort to do so, and employers losing valued employees must spend
`
`18
`
`time, money, and energy looking for new hires (or identifying those within the organization that
`
`19
`
`can either be promoted to replace the lost employee, or have their jobs combined with some of the
`
`20
`
`departing employee’s responsibilities). By predicting employee behavior and identifying
`
`21
`
`hidden/adjacent skills based on the information those same employees or employee prospects
`
`22
`
`share publicly, hiQ reduced these costs and helped their clients create a happier, better-
`
`23
`
`compensated workforce that was simultaneously more suited for their jobs.
`
`24
`
`33.
`
`In response to this realization, hiQ developed “people analytics,” a new type of
`
`25
`
`predictive data analytics aimed at providing employers in-depth, predictive insights into their
`
`26
`
`workforce. People analytics generally work by performing computerized analyses of employees’
`
`27
`
`public professional information and history that then show which employees are at higher risk of
`
`28
`
`looking for a new job, and which may have skills that are not being utilized in their current job.
`
`AMENDED COMPLAINT
`
`
`
`
`
`
`12
`
`Case No. 3:17-cv-03301-EMC
`
`
`
`
`
`Case 3:17-cv-03301-EMC Document 131 Filed 02/14/20 Page 13 of 36
`
`
`
`1
`
`hiQ created two specific analy