Big data and its developer fallout

大数据及其对开发者的影响

随着互联网社会地盘之争继续成熟,”圈地‘’变得更好理解。在少数几家公司控制95%的社交数据情况下,互联网的封闭和受更多的控制的程度比以往任何时候都强。

术语(及涵盖的概念)大数据在过去15个月一直被广泛提及。这里我指的是用户数据, 它们主要来源于社会企业,如果企业做一个公司范围内的API（应用程序接口），则其可用于构建其他应用程序和业务。

            举一些基本的例子。例如Facebook:开发人员、产品架构师,企业家等可能需要分析姓名、照片或股票。再如Snapchat:分析股票或发送物品的数量。或者Instagram:分析用户,热点或评论。特斯拉:分析汽车位置、能源消费、电量持续情况。此类例子不胜枚举。现代网络是建立在开放的数据交换基础之上。

我经常被问到我认为什么因素可以成就一个好的应用。答案很简单:数据。更具体地说:用户的和他们各自的数据集合。用户也是数据。没有他们,无论你的应用程序多么华丽,都不会有实用性。

进一步的问题总是：“好吧，我怎么得到用户？”

这是价值数十亿美元的问题。现有的社交网站要你相信，通过连接到他们用户就会被吸引过来。只是如果这是真的才奏效。只是为了以正视听而说，你不能买到用户。而且你无法连接到现有的网站去留住用户或做到在两者之间的任何事。用户已经厌倦了新的软件，但更加讨厌一致的软件。合理分配（即找到并留住用户）是创建一个成功的应用程序中最难的部分。

作为开发人员,我们能够深入到Facebook上的社交图谱。开发人员能够通过一个复杂的方式在各种各样的网络和应用产品中注入有意义的数据,最重要的是,我们能够不受带宽限制地请求使用大数据集合。能做到这些只是由于一些大公司说可以使用他们的数据，但并不意味着这些数据是正确的。

在过去的几年里，形势已经有了大规模的转变。可悲的是,说明这种变化的最好方法是通过Zynga的兴衰演变。由于Facebook开放API和允许用户做深层渗透到Facebook图表,相比其他任何公司, Zynga直接通过Facebook图形的API建立了一个令人难以置信的游戏产业。随着时间的推移,Facebook开始更改开发人员如何与特定的数据交互的方式。像Zynga迅速的成长一样,他们迅速衰败了,而且衰败得更厉害。有很多公司,或大或小,遭受类似的灭亡。

现代社会的大数据的玩家就像在一家糖果店里引诱一个没有硬币花的孩子一样引诱你，有的只是无尽的诱惑与好的解决方案的承诺。高品质、快速的数据交换或深层图标渗透的承诺是所有现代大企业API文档的内容。然而，就像生活中的很多东西，魔鬼藏在细节之中，伙计，当心。当然，你可以在期望的速度下获得数据，但一旦你打API门槛，进料将像尼亚加拉大瀑布从水槽龙头流出。如果您的企业依靠其快速检索数据，该怎么办呢？

同样的，你是可以访问组曲线图，并做深入的分析，但稍加深入，你会发现，他们只会给你1%-3%的关于一个具体位置，用户，热点，股票等任何有意义的信息。唯一的好处只有整体用户及其在特定应用中的他或她的活动一小片信息吗?

最后，Zynga和其他许多鲜为人知的企业表示，如果你在其他业务之上建立一个业务，那么你会受其他业务的心血来潮的支配。他们能够而且将会改变你与他们的数据进行交互的方式，这对您的企业的寿命和盈利能力产生重大影响。

创业牧场乐土,无限且快速的数据请求和用户数据的深层渗透，这些承诺都没了。最好的情况下，实际的牧场是一种戏弄。随着互联网和社会企业的成熟,价值是“围墙花园”。通过隔离用户数据在核心业务范围内,市场和销售广告的能力赢得每一次胜利。没错,这些企业做广告确实比别人好,但是一个你可以支付的健壮版本的访问的模型从不能生存。市场太小,不足以支持这种模式。

所以我们要何去何从? 市场仍有巨大的潜力，我们只需要用更进化的方式思考。你不能简单地想：“我要想出一个主意,利用现有的社区,让它茁壮成长。“API就是不允许这种类型的开发了。像Uber最近做的,Pinterest Snapchat也一样,企业家的心态需要成为新社区之一,垂直,专业的方法——一个“围绕一个利益的共同体”。

不要让社交成为您企业的单体支柱，而是让它成为一个功能。没有一个总体的前提下，思考纯粹的社交是具有讽刺意味的孤立，也肯定会迅速在应用商店销声匿迹。

 

Big data and its developer fallout

As the internet social turf wars continue to mature, the land grab is becoming much better understood. With a few companies controlling 95 percent of the social data, the internet is more closed and much more controlled than ever before.

The term (and concept behind) big data has been thrown around a lot over the past 15 months. What I’m referring to here is user data, primarily from social businesses that can be leveraged to build other apps and businesses if done within the confines of a company API.

A few basic examples. Let’s take Facebook: A developer, product architect, entrepreneur, etc. may want to analyze names, pictures or shares. How about Snapchat: shares or number of sent items.  Instagram: users, hearts or comments. Tesla: car location, energy consumption, last charge. The list goes on and on. The modern web has been built on an open data exchange.

I regularly get asked what I think makes a good app. The answer is simple: data. More specifically: users and their respective meta-data. Users are data. Without them, no matter how flashy your app is, it won’t work.

The follow up question is always, “OK, how do I get users?”

That’s the billion-dollar question. The existing social sites want you to believe that by connecting to them and through them that users will come. If only that were true.  Just to set the record straight, you can’t buy users. And you can’t connect to existing sites to leverage users or anything in between. Users are tired of new, yet more of the same, software. Distribution (i.e. finding and retaining users) is the hardest part of creating a successful app.

As developers, we used to be able to go deep into the social graph on Facebook. Developers used to be able to inject meaningful data in a sophisticated way to all sorts of web and app products and, most importantly, we used to be able to request big datasets without getting throttled by bandwidth limitations. Just because a few big companies say you can use their data doesn’t mean it’s accurate.

Over the past few years there’s been a massive shift. Sadly, the best way to illustrate such change is to look at the  rise and fall of Zynga. As Facebook opened their API and enabled users to do deep penetration into the Facebook Graph, Zynga, more than any other company, took advantage and built an incredible gaming business directly through the Facebook Graph API. Over time, Facebook began making changes to how developers could interact with specific data and just as quickly as Zynga grew, they fell — and fell far. There are many companies, big and small, that have suffered a similar demise.

The modern social big data players lure you in like a kid in a candy store with no coin to spend, just endless temptation and promises of sweet solutions. The declaration of high-quality and fast data exchange or deep penetration into graphs are the prose of all modernbig businesses’ API documentation. However, like most things in life, the devils are in the details, and boy, watch out.  Sure, you can have the data at desired speed, but once you hit the API threshold, the feed will go from Niagara Falls to a leaky sink faucet. If your business relies on its ability to quickly retrieve data, now what?

Similarly, yes, you can access set graphs and do deep analysis, but dig a little deeper and you’ll find, they’ll give you only 1-3 percent of anything meaningful about a specific location, person, hearts, shares and so on. What good is only a small sliver of an overall user and his or her activity across the app?

Finally, Zynga and many other lesser-known businesses illustrated that if you build a business on top of another business, you are at the whim of their decisions. They can and will change the way you interact with their data, which has major implications on the longevity and profitability of your business.

The promised land of green pastures, endless and fast data requests and deep penetration of user data is over. The actual pasture is a tease, at best. As the internet and social businesses have matured, the value is the walled garden. By keeping user data within the confines of the core business, the ability to market and sell advertising wins every time. It’s true that some of these businesses do advertising better than others, but a model where you can pay for robust versions of access is never going to survive. The markets are just too small to support that model.

So where do we go from here? There is still a huge amount of potential, we just need to think in a more evolved way. You can’t simply think, “I’m going to come up with an idea and leverage an existing community to make it thrive.” APIs just don’t allow that type of development anymore.  Like Uber has done most recently, as well as Pinterest and Snapchat, the mindset of the entrepreneur needs to be one of a new community, a more vertical, specialized approach — a “community around an interest.”

Don’t make social be the single pillar of your business; have it be a feature. Thinking purely social without an overarching premise is ironically solitary, and surely the fastest way to the back of the app store.

 

 

 

From：https://techcrunch.com/2016/07/16/big-data-and-its-developer-fallout/