Mark Hurd has been co-president of mega-cap technology company Oracle for three years now, having joined shortly after his dramatic exit as CEO of Hewlett-Packard. Hurd's dramas these days are the more typical business type: Oracle's coming to grips that it is a giant competing against nimbler upstarts. Oracle faces competition from the likes of Salesforce.com (CRM) and Workday (WDAY) as well as more size-appropriate challengers IBM and SAP. (Oracle makes database software and many applications that run with it; Salesforce and Workday are online-only application makers, with the former focused on sales tasks and the latter on human resources. IBM and SAP have software offerings similar to Oracle's.)

    The incumbents have been scrambling to match the newbies in "cloud" computing, or software that is offered as an Internet service rather than as programs installed on a customer's computers. Oracle (ORCL) is generally perceived to have been slow to respond to the cloud, having instead pursued an aggressive acquisition strategy that included the unlikely 2010 purchase of server maker Sun Microsystems for more than $7 billion. Moving to the cloud is tough all around for big companies -- like turning the proverbial aircraft carrier -- in part because Internet software generally costs less. That's great for customers, but tough for salespeople, and Hurd, a sales executive earlier in his career, oversees Oracle's sales force.

    In an interview in his Redwood Shores, Calif., office, Hurd spoke with Fortune on the eve of Oracle's massive Oracle Open World developers' conference, which promises to snarl traffic throughout San Francisco from Sept. 22 to 26. Hurd addressed the state of Oracle, including the shift to the cloud, why Oracle will beat its competitors, and whether or not Oracle will abandon hardware. Edited for concision and annotated for explanation, a transcript of the conversation follows.

What will be the highlights at Oracle Open World?

    First of all, it'll be a highlight that we have 60,000 people. It'll be the biggest IT event that I'm aware of ever in the industry to have that many people in one location. We'll have a very large partner ecosystem there. So you'll see Dell (DELL), Deloitte, and EMC (EMC) there. Fujitsu will be there. You'll see a very large cast of partners that will participate. We have a huge list of customer speakers. I'll be doing a speech, and I'll have the CEO of Airbus with me. And the CEO of the New York Stock Exchange too.

And products will you be focusing on?

    We're going to go talk about the fact that we're going to now have in-memory options for our database. [This refers to the ability to store database information on memory chips rather than on disc drives, analogous to the why Apple's iPods shifted years ago from bigger mechanical drives to smaller flash-memory.] We have released a new database called 12C. It is the 12th generation of the Oracle database. The C is for cloud. In-memory will be a big theme.

    马克•赫德戏剧性地辞去惠普公司（Hewlett-Packard）CEO的职务之后不久，便加入了大型科技公司甲骨文（Oracle），如今他担任公司联席总裁已有三年时间。赫德现在扮演的角色是实施更为典型的营运模式：甲骨文已经意识到，它就像是一个巨人，要与身段灵活的科技界新贵进行竞争。甲骨文不仅要面临企业云计算公司Salesforce.com和人力资源软件开发商Workday等公司的竞争，还要与IBM和SAP等规模相当的公司展开激烈角逐。（甲骨文生产数据库软件，以及其他配合数据库软件使用的应用；Salesforce和Workday仅开发在线应用，其中前者专注于销售任务，后者专注于人力资源。IBM和SAP的软件产品与甲骨文类似。）

    传统软件企业竞相与云计算领域的新兴公司结盟，或者争夺作为互联网服务提供的软件，而不是安装在客户电脑上的程序。业界普遍认为，甲骨文对云技术的反应迟缓。实际上，这家公司执行了积极的收购策略，包括2010年，以超过70亿美元出人意料地收购服务器制造商太阳计算机系统公司（Sun Microsystems）。转移到云对大企业而言有些难度，就好像让航空母舰调头一样，其中很大程度上是由于互联网软件通常成本更低。这对客户是好事，但却给销售人员带来了麻烦。甲骨文的销售团队便由赫德亲自负责，他在职业早期还曾担任过销售主管。

    在甲骨文开放世界（Oracle Open World）开发者大会召开的前夜，在位于加州红木海岸的办公室里，马克•赫德接受了《财富》杂志（Fortune）的专访。预计在9月22日至26日期间，旧金山市的交通会因未这个规模庞大的会议陷入混乱。赫德提到了甲骨文的现状，包括向云技术的转变，以及为什么甲骨文能够击败竞争对手，甲骨文是否会放弃硬件等等。以下为对话摘要，为了简洁，我们对谈话内容进行了编辑，同时添加了注释。

甲骨文本届开放世界大会有哪些亮点？

    首先，大会将有60,000人参加，这是亮点之一。据我所知，这次大会能将这么多人汇集一堂，堪称业内规模最大的一次IT业盛会。借助此次盛会，我们将拥有一个非常庞大的合作伙伴生态系统。所以，大家会看到戴尔（Dell）、德勤（Deloitte）和EMC。富士通公司（Fujitsu）也会出现在大会上。大家会看到我们庞大的合作伙伴阵容。我们还有重量级的客户发言人。我会发表一个演讲，发言的还有空中客车公司（Airbus）CEO，以及纽约证交所（New York Stock Exchange）CEO等等。

你会重点介绍哪些产品？

    我们会谈到甲骨文将推出内存数据库这个事实。（它是指将数据库信息存储在内存芯片上而非磁盘驱动器上的能力，类似于数年前，苹果的iPod促成了机械驱动向更小的闪存转变。）我们发布了名为12C的新数据库产品。这是第12代甲骨文数据库。C代表云。内存数据库将是一个重要的主题。

    There will be another series of cloud announcements around things like human capital management [think: human resources] and customer experience software [like sales applications]. You'll also hear us talk about a thing called M6, which is a piece of silicon which is in the Spark [server] product line. And there will be a series of other hardware announcements.

Going back to the cloud, what percentage of overall business, expressed either in revenues or profits, is cloud vs. traditional software?

    We don't really give out a cloud number. The reason it's hard to do, is that while we have software as a service offered as a subscription -- and we give out that number: It's in excess of a rate of $1 billion a year -- that would not include all the things in the Oracle portfolio that are used in the cloud. [Investors expect Oracle to record revenues this year in the neighborhood of $38 billion.] So that doesn't include everything we sell to Salesforce.com, which is used for them to run their cloud, for example.

But in your example, what you're selling to Salesforce.com is traditional software. That's how they're buying it. What they do with it is their business. I mean, you don't include auto sales either in your revenue either. No one's asking you, "What's your percentage of revenue from selling automobiles?"

    Yeah, we don't sell too many automobiles. But we do sell software to people that run clouds. Remember, there's multiple elements to the cloud. There's tools. There's infrastructure. And there's applications. In addition, there's not just what you think of as the public cloud. There's a thing called the private cloud, behind the firewall. And the only reason I'm making these distinctions is that the word "cloud" is a relatively complicated term. It's an umbrella term. Cloud means a whole set of things. And by that definition of cloud, we sell tons of things that are used as a private cloud. And then we have our own [software-as-a-service] applications that people can use. And we've been public about that number.

Which is that billion-dollar-plus run rate.

    Yes.

Growing at what kind of rate?

    We don't publish that. It's a good number.

How satisfied is Oracle with its cloud offerings? Did you get into it early enough? Are you where you need to be now? Are you concerned with the younger, smaller competitors that grow much more quickly in cloud offerings? Three questions, I guess.

    Yeah. It sounded like five or six in one thought? [For the record, it was four questions.]

I know you're up to it.

    I think Oracle did a couple things. When there was no such word as cloud, and there was just software over the Internet, Larry started this company called NetSuite. NetSuite pre-dated Salesforce.com. So Larry had this idea that software over the Internet seemed like a pretty good idea way back at the time. [NetSuite, which focuses on online sales software, is worth about $8 billion today, compared with Salesforce.com's valuation of more than $30 billion. Larry Ellison, CEO of Oracle, continues to control nearly 44% of NetSuite's shares.]

    And so we went about the process of building applications organically through R&D, and simultaneously acquiring pieces that we weren't building through R&D. For example, we have a sales automation solution that was built organically by Oracle. We have a service automation solution that we bought, a company called RightNow. They are the leader. And by the way, every company we bought was the leader in their space. In human-capital management [usually known as human-resources software] we built a core product. And we bought Taleo, a recruiting app. We did not have a marketing automation solution. So we bought a company called Eloqua that is the leading marketing automation solution in the world. So this is now a combination of a build-and-buy strategy to build a suite of capabilities that's unmatched in the industry. Nobody has it.

    此外，还会有一系列与人力资本管理（比如人力资源）和客户体验软件（如销售应用）等领域有关的云技术产品。此外，我们还会谈到M6，这是Spark（服务器）产品中的一个硅片。此外，我们还会在大会上宣布其他硬件产品。

我们再回到云服务。从收益或利润的角度来看，云服务与传统软件在贵公司整体业务中占多大的比例？

    我们无法给出云服务的数字。原因是，虽然我们有以订阅方式提供的“软件即服务”产品，而且我们也给出了这部分的数字：每年超过10亿美元，但它并不包括所有在云中使用的甲骨文产品。（投资者期望甲骨文今年的收入在380亿美元左右。）所以，它没有涵盖我们出售给Salesforce.com的所有产品，而他们可能会使用我们的产品运行他们的云服务。

但在你举的例子中，甲骨文出售给Salesforce.com的是传统软件。这也是他们购买的原因。他们怎么使用这些软件是他们的自由。我的意思是，你也没有在收益中包括汽车的销售。没有人问：“汽车销售占公司收益的比例是多少？”

    是的，我们没有出售太多的汽车。但我们确实将软件出售给经营云的人。记住，云技术包含许多元素。有工具，有基础设施，还有应用。而且，不仅有你所想的共用云。在防火墙后面还有私有云。我之所以进行这样的对比，是因为“云”是一个相对复杂的术语。它只是一个涵盖性的术语。云意味着一整套东西。按照云的定义，我们出售的大量产品被用作私有云。还有我们自己的“软件即服务”应用，可供人们使用。这些数字一直都是公开的。

也就是超过十亿美元的销售额。

    是的。

增长速度如何？

    我们尚未公布这个数字，不过非常乐观。

甲骨文对旗下的云产品满意吗？甲骨文进入云领域的时机够早吗？目前实现的目标是你们需要达到的吗？更年轻、规模更小的竞争对手在云技术领域发展更快，对此你会感到担忧吗？我想这是三个问题。

    是的。乍一听像是五六个问题。（准确的说，一共是四个问题。）

我知道你能接得住。

    我想甲骨文做过两件事。云这个概念还没有出现之前，当时还只有网络软件，拉里成立了一家叫NteSuite的公司。NetSuite的出现要早于Salesforce.com。所以，当时拉里认为，网络软件听起来是个不错的主意。（NetSuite公司致力于在线销售软件，目前市值80亿美元；相比之下，Salesforce.com的市值已经超过300亿美元。甲骨文CEO拉里•埃里克森仍控制着NetSuite约44%的股份。）

    于是，我们开始了通过研发系统开发应用，同时收购并未通过研发制造的应用。例如，我们有销售自动化解决方案，它由甲骨文系统自己开发。而我们收购了一家名为RightNow的公司，获得了服务自动化解决方案。他们都是业内领导者。而且，我们收购的每家公司都是如此。在人力资本管理（通常称为人力资源软件）领域，我们开发了核心产品。但我们收购了一款招聘应用Taleo。我们没有营销自动化解决方法，所以我们收购了一家世界领先的营销自动化服务提供商Eloqua公司。因此可以说，通过自主开发与收购的模式，我们具备了在业内无可匹敌的能力。没有任何其他公司具备这样的能力。

    How do you communicate this strategy to customers?

    When you hear "cloud" and you gravitate to a perspective that everything should be cloud. We've tried to do something a little bit different architecturally. When the customer says, "I really want this," we've built our solution so they can either use it as an online application from the Oracle cloud, or if they like the IP and don't want it in the cloud, they can do it behind their firewall. Our message: Instead of worrying about the delivery architecture, worry about the application. Worry about the IP.

How happy are you with how well your sales force is selling this strategy?

    You're never going to ever say you're thrilled with everything. But I am happy with the progress of our sales organization. I think we changed them quite a bit a year and a half ago, and I think they've made good progress.

One analyst I spoke to pointed out that sales force growth is growing faster than license revenue growth. Why?

    We have invested a reasonable percentage of our sales force growth into this thing called cloud. And when you sell cloud, you sell a subscription. You do not sell a license. [Subscribers to software services typically pay more slowly than license customers -- and perhaps less as well.] That doesn't mean long run that the business models aren't fine, but you do have this outage of productivity as you build up the subscription revenue base.

What was Oracle's strategy in becoming a hardware maker, and are you 100% stay-the-course satisfied with that strategy today?

    Well, we're staying the course, just to take that off the table. Our strategy has been to vertically integrate hardware and software, in that we think hardware and software together, engineered together, provides a better outcome. And we think it provides a better outcome on several planes, and I'll try to describe what they are.

    First, we do the work that you, the customer, used to do. So historically in the industry you could buy a server from [one vendor]. You could buy an operating system from [another]. You could buy a database from [a third]. You could buy a middleware from [yet another]. And then you give it to your IT organization, and your IT organization glues all this stuff together. Well, our better idea would be, Why don't we just glue it all together for you? We'll test it. We'll warrant it. And we'll do that work, and that cost goes into our R&D as opposed to your IT budget. So point one: We do the work, you don't. Second, this stuff just performs better, meaning that we build hardware to optimize the performance of the software. Third, you get a different support experience. So the fact that we build the stuff together and integrate it together, we now have one version. We have tested all the versions of software together, and we've seen all the problems. And we can create a repository, and any time our customer has a problem, we directly hook up and can auto-patch it in nanoseconds. So our strategy has been to sort of change the game with this -- and then you see a lot of people trying to frankly, you know, imitate that strategy. You've seen it from many of our competitors.

    你们如何将这种策略传达给客户？

    人们一听到“云”这个词就会不由自主产生这样一种观点，即所有东西都应放到云中。我们在结构上做了一些小小的改动。客户说“我非常想要这个”的时候，我们已经开发出解决方案，客户可以将它作为一款来自甲骨文云的在线应用。而如果他们重视知识产权，不希望将内容放到云中，他们也可以将其存放到防火墙后面。我们传达的信息是：不要担心交付架构，多关心关心应用和知识产权吧。

你的销售团队销售这种策略的效果如何？你满意吗？

    并非所有的事情都会让人振奋，不过总体来说，销售部门的进展还是让我感到高兴。我认为，在一年半以前，我们让他们有了一点改变，他们取得了不错的进展。

我交流过的一位分析师指出，销售队伍的发展速度要快于许可收益增长。为什么？

    我们把合理比例的销售力量增长投入到云销售中。而在出售云的时候，其实是在出售一种订阅服务，而不是在出售许可。（软件服务订阅用户支付的速度通常慢于许可客户——而且支付的金额可能也会更少。）但这并不意味着，这种商业模式的长远前景不佳，但在建立订阅收益基础的过程中，确实会出现生产力损耗。

甲骨文在成为硬件制造商方面有什么策略？你始终对今天的策略100%满意吗？

    其实我们一直在坚持，就是为了最终完成这个目标。我们的策略是垂直整合硬件和软件，我们会将硬件和软件放在一起思考，一起进行设计，以提供更好的成果。我们认为，这种策略已经给许多项目带来了更好的结果，我会尽量说明这些项目。

    首先，我们所做的是客户之前的工作。过去，大家会（从供应商）购买一台服务器，从（另外一名供应商处）买操作系统，从（第三名供应商处）购买数据库，再从（其他供应商处）购买中间件，然后把这些东西交给IT部门，IT部门将所有这些东西整合在一起。而我们的想法是，为什么我们不能提前为客户将所有这些东西整合起来？我们会对它进行测试。我们会提供保修。我们来做这项工作，而成本将被计入我们的研发成本，而不是客户的IT预算。这是第一点：工作由我们完成。第二，这种组合的性能更好，意味着我们建造硬件来优化软件的效果。第三，客户会获得不同的支持体验。我们将所有东西整合起来，现在就有一个版本。我们会测试软件的所有版本，而且我们已经发现了所有问题。因此，我们可以建立一个资源库，只要客户遇到问题，我们便可以直接与客户对接，在极短的时间内自动解决问题。所以，我们的策略能够改变游戏规则——而且说实话，现在有许多人在试图模仿我们的策略。许多竞争对手都在这样做。

Like who?

    [Unlike his boss, Hurd avoids mentioning competitors by name. It is no secret that IBM and SAP, also big companies with massive on- and offline software offerings, have made hardware acquisitions.]

You seem to be focused on large competitors that look like Oracle, while Wall Street is focused on smaller, faster-growing competitors.

    If an investor wants exposure to big-cap tech, an investor's going to have to make a decision. You know, do they want exposure to a very, very strong, high-cash-flow company like Oracle with material recurring revenue, with a position to compete with the companies that are small, that you described? By the way, I think that sounds like a pretty cool profile. Or do you want some other big cap company that's got negative growth and, you know, frankly is in the position that it's in. You as an investor will have to make that decision. Now, I think that's a little different than us operationally in the sense that, yeah, we are competing to a degree with those companies, albeit perhaps very North America-centric and in a single process, the way you described HR or sales automation. And I think that's good for the investor. That's good for the investor that we're going to go out and whip their butt. And at the same time, you've got this high recurring revenue cash flow machine behind it.