微软CEO谈IT转型、亚马逊与Windows 10

    很多公司认为，如果他们把更多的企业应用置于云端，那么他们就不再需要那么多IT人员了。你可能不需要一个SharePoint或Exchange方面的专家。那么，大型企业的IT部门人数现在是否在缩减？

    总会有一些人是网络安全、终端用户设备、协作或数据库方面的专家，这些人是不会失业的。但企业为什么要提供这些职位？就是要帮助企业转型。

    你刚刚谈到了预算缩水的问题。一个有趣的现象是：你可以问问任何一家公司——不管它从事制造、包装、能源、零售还是什么行业——其IT支出的哪一部分正在上涨，哪一部分正在下降。后者就是所谓“过时的IT”——它可能是很多风扇或服务器什么的。即便你是一个以本地应用为主的用户，你在研发或测试新应用时也会用到大量云技术。你可能会在云端上构建移动或网页后端，但你还没有把ERP（企业资源规划）系统放到云端上。

    所以在某种程度上可以说，他们希望用和昨天一样的投资获得更多的回报。然后他们想花更多的投资去做他们正在做的新东西。

    并不是只有IT部门的IT支出很有限，这个现象在营销和人力资源领域也很普遍。一个有意思的现象是企业的分析支出。Power BI（微软的“商业情报”分析服务）是我们以SaaS模式（Saas又称“软件即服务”，是一个用来描述云应用的行业术语）发布的，我起初以为它的用户主要是开发者或营销人员。后来我发现人力资源部门也在使用Power BI，这是因为“人力分析”的重要性。因此，有些领域我原本没有预料到会出现较大的IT支出，但它们却成了IT的开支大户，因为现在IT已经无所不在。

    让我们谈谈数据吧，如何制作非常复杂的数据集，以及如何分何这些复杂数据，令普通人也能看得懂。

    如果你看看我们的数据库业务特别是SQL服务器业务……我觉得非常搞笑的是，很多人认为亚马逊网页服务（Amazon Web Services）是数据领域的领军者，却没有意识到我们通过“古老的”数据库业务早已实现了数据的平民化——这项业务已经发展到80亿美元的市值，无处不在，居于企业的核心地位。

    我们最近刚刚收购了一家名叫Revolution Analytics的公司。凭借我们的数据解决方案，我们将彻底改变所谓的“高级分析”的面貌。我们的机器学习技术将把高级分析和统计学机器学习技术带到全球各地的数据科学部门。我们的Excel绝对拥有这种优势。作为一项SaaS服务，Power BI可以使你通过云端进行丰富的可视化和仪表盘操作，并进行协作与分享。

    我非常关注这些技术的使用。我仔细研究了Office 365的租借和服务使用情况。所有这些数据都是用仪表盘工具（Dashboard）分析的。如果一位利用Office 365的工程师——比如说他正在利用其Exchange Online功能——有一个仪表盘，我就可以发现他的使用情况。所以，它其实已经成为我们理解公司内部的大数据及其协作方式的通用工具。

    而且它不仅仅是报告或仪表盘这样简单，它是实时的数据。这意味着我可以问一个自然语言的问题，并且可以获得回答，做丰富的可视化操作，然后分享回去。这就是我以前提到过的“数据文化”。仅仅拥有大数据，并不意味着你就拥有了正确的见解，或拥有了更好的决判能力，所以你必须要有数据文化。什么意思呢？如果你有了正确的工具，你只需要拥有数据文化就可以了。这正是Power BI的目标。

    Many companies think that if they put more business applications in the cloud, they don’t need as many IT people. Maybe you don’t need a SharePoint expert or an Exchange expert. Is the net number of IT professionals shrinking at large companies?

    There are always going to be people who are experts in security, or end-user devices, or collaboration, or databases. That’s not going to go away. But what’s the reason all of these professions come together? To help the business transform itself.

    You talk about shrinking budgets. So here is a fascinating phenomenon. Ask any company—it could be in manufacturing, packaged goods, energy, retail, what have you—what in their total digital technology spend is going up as a percentage of revenue. Then ask what is coming down. It’s what was considered past IT—maybe a bunch of fans, or servers. Even if you are predominantly an on-premise customer, you’re using the cloud quite a bit for development and testing new applications. You may be building a mobile or web back end in the cloud, but you’re not yet moving your ERP [“enterprise resource planning”] system to the cloud.

    So in a sense they want more out of the same dollar for what they did yesterday. And then they want to spend more dollars for things that they’re doing that are new.

    IT spend is not limited just to the IT department. It’s spread across marketing and human resources. One of the fascinating things I see is the amount of analytics spend. Power BI [a Microsoft “business intelligence” service], which we launched as SaaS [“software-as-a-service,” the industry term for a cloud-based application], I initially thought would be used by developers and maybe marketers. It turns out that HR departments all use Power BI because “people analytics” is a big deal. So in places where I traditionally would not have expected big IT spending, they are becoming big spenders of IT, because IT is everywhere.

    Let’s talk about data, making really complex data sets, and making the ability to parse them available to mere mortals.

    If you look at what has happened with our database business, at the core, with SQL Server…I mean, it’s just pretty hilarious for me to think about people who think of Amazon Web Services as a leader in data and not have an understanding of the democratization of data that we have done with our regular old database business—an $8 billion business that’s everywhere, in the core of the enterprise.

    We just bought a company called Revolution Analytics. We are going to completely change what it means to do advanced analytics with our data solutions. We have machine-learning stuff that is about really bringing advanced analytics and statistical machine learning into data-science departments everywhere. We absolutely have this strength in Excel, but Power BI is a SaaS service that allows you to do these rich visualizations and dashboards and collaborate and share.

    I’ve been very, very focused on usage. I look at what’s happening in Office 365 by tenant, by service. All of that is dashboarded. So if one of the engineers on Office 365, say Exchange Online, has a dashboard I can go discover that usage. So it’s become, in fact, the lingua franca of how we make sense of big data inside the company and how we collaborate on it.

    And it’s not just reports or dashboards. It’s live data. That means I can ask a natural-language question, get a response, do a rich visualization, and then share back. That, I think, is what I referenced before as data culture. Just because you have big data doesn’t mean you have insights or a better ability to pass judgment. So you’ve got to have this data culture. What does that mean? You are only going to have a data culture if you have the right kinds of tools. That’s what Power BI is about.