域外学道：我们从ODI-MS 朋友圈中学到的数据联合创新体的构建 What We learned from ODI-MS PLN
最后更新于
最后更新于
我们在此次的朋友圈交流中提出的是我们当前在上海和金融机构合作过程中所提出的金融数据联合创新体的实验。这当然更多偏向于一个实验性的概念而非一个已真正落地的项目，但我们期望的是能够通过我们的不断迭代设计，能够做到最后的落地。
The Fintech data collaboration as we are currently working on is a at its early conceptual stage. Of course, this is more inclined to an experiment than a project that has already implemented.
从需求而言，为什么我们考虑要做这个金融数据联合创新体？一方面从内需而言，各家金融机构有着强烈的突破各家机构间数据壁垒的冲动，从而实现互相间的数据共享，对客户（无论是个人还是企业）能够真正做到全方位的评估，进而更好的提供金融服务。另一方面，从对外角度而言，金融机构的数据对于初创企业来说自然是香饽饽，那么如果能够通过这样的数据供给为金融机构自身带回创新的服务产品，值得投资的标的企业等就自然是再划算不过的一桩买卖了。
From the perspective of demand, why do we consider doing this data collaboration? On the one hand, from the perspective of internal needs, various financial institutions have a strong impulse to break through the data barriers among them to realize joint model buidling. On the other hand, it benefits all data owners if they can get tangible return, either money or innvoative solutions, from SMEs.
上图给出了一个我们的基础架构，绿色部分是我们的数据源自然是各类型的金融机构，既包括国有的也有私有的；黄色的是基础底座，这个我们认为要包含三样要素既许可协议为载体的法律规则，技术设施比如云存储、隐私计算技术等，以及很关键的政府支持（用于打通国有企业）；蓝色的运营部分，我们思考的是通过让产业园区出种子资金，孵化器负责日常运营，去链接这个创新联合体和使用者即创业企业。
The above image provides a simple framework of the collaboration: green pars are data owners who are state-owner of private-owned banks or insurance companies. Yellow part are the infrastructures including technical infrastructure but also license framework serving as legal foundation. The blue part is about how the collaboration itself works or is operated. We will invite one industry park to provide seed funding to make collaboration work.
在尝试构建起这样一个数据创新联合体时，所需要解决的问题众多，我们在这里将主要讨论两个最为关键的问题，其他方面我们的经验可期待之后我们为 PLN 编写的报告。
On the journey of building up the collaboration, there are many tangible issues we face but we will only share two important ones here. For more information, you can expect the final report we deliver as part of the PLN.
在一个强商业化驱动的数据创新联合体中，最为直接去赢得上层支持的是去确保这个联合体可以带回利益，而这就取决于联合体是否有明确的商业模式或者说不同的收入来源。
To a strong commercial data collaboration, it is important show the C-level decision makers that the collaboration can bring back real tangible benefits. This will depend on whether we have a viable business model or revenue stream in place.
在我们考虑这一问题时，把自己过度地局限在了对数据收费的这一焦点问题上。我们试图考虑过将通过对数据做不同的整合、抽象，我们能够将一些数据免费的提供，而对于高价值的、实时的或者高频率更新的数据，我们则采用 API 等方式基于用量来收费。而那些高敏感高价值的数据则可以通过可信计算环境如沙箱等予以封闭式访问使用，而这一类可采用数据作为一种投资去换取初创企业的股份、数据换服务做等价交换等去构建商业模式。
We originally narrow our focus on how to charge for data use. If we produce aggregate abstract data then such data can be free of charge. If data users require real-time data or data frequently updated then we can provide such data in API format and charge the usage by the API rate. Sand-box is suitable for highly sensitive data development and we can ask for a equity share by considering data access as a form of investment.
有意思的是，ODI 所提供的 Sustainable Data Access Workbook 对我们有很大的启发。通过去问自己是否有除了对数据本身收费之外还能创造收益本身就是一个很有挑战的问题。而这个工具的启发性是引导你重新去考虑整体联合体的优劣势，从而基于优劣势或者说机遇和挑战去重新思考可能获得收益的来源。
Interestingly, we expand our vision of revenue stream after using the ODI's Sustainable Data Access Workbook tool. To be honest, it might be already a great challenge to even ask whether there is another revenue model in addition to charging data access. This tool guides you by asking you to unpack the strength and weakness of your collaboration as well as opportunities and threats.
我们通过这个工具意识到两点：第一我们自身如果建立其一个联合体，那么我们本身就是这方面的行业领先者，相应去输出经验本身就是一种可能的商业模式；第二，我们合法提供数据的这种行为本身是对使用者乃至最终端用户的一种「信任服务」，我们确保了数据来源的合法以及使用的合法合规，因此我们也能够从这当中对开发者认证收费或者对希望享受更多优质数据服务的终端用户收费。
We identify two possible approaches to generate revenues by using this tool: first as we are kind of pioneer in building up data collaboration then why not monetize our experience by providing consulting services to other organizations who might also want to set up data collaboration; second, why not charge membership or subsrcription fee if people can trust apps developed upon and cerfified by our data collaboration.
在我们早期研究我们的数据创新联合体时，我们并没有太多将注意力投放到个体之上，虽然我们的数据中国有很多是关联个体的，但我们并不认为我们的数据创新联合体需要去和个体间发生什么关系，而只用处理好和数据拥有者和数据使用者的关系即可，也因此当我们回过头去看当时利用Data Ecosystem Mapping工具绘制的图时，个体是不存在于这张图之上的。
In our early planning and discussion , we never take into account individuals as stakeholder. Partially it is because that we consider our job is to balance the demand and requirements between data owners and SMEs. So even we use the ODI Data Ecosystem Mapping tool there is no such stakeholder group on the map.
而事实是，即使通过前一节的商业模式的思考，我们也会发现试图去将个体作为一类利益相关者是多么重要。更何况如今中共正在制定中的个人信息保护法可能对于我们的数据联合体带来新的挑战：如何去处理好数据联合体和个体权益间的关系，或者说我们该做些什么似的个体信任我们的联合体不会带来滥用和个体相关的数据？
But individuals are quite important stakeholders just as we also discussed in the last section that some new revenue streams could be based upon the interaction with individuals by protecting their rights. It is also much more important now as the peronal informational protection law is under screen and review now in China. Thus it is a key issue for our collaboration to address.
在这一方面我们并没有最终的答案，感谢 ODI 分享的一些尚未正式刊发的研究的启发，我们正在逐渐梳理一套框架分层次的去思考个体权益如何体现在我们的数据联合体中，包括如何去响应法律所抛出的数据处理者应当建立具备第三方独立视角的主体来监管数据使用的要求。
We thanks ODI for sharing some internal research to inspire us on further unpack the question: how exactly we would like to engage individuals in making decisions or regulating our operation or what?
我们期望能够在后续的 4-6 个月中进一步深入探讨这一方向的可能性，包括但不限于组织一些带入具备一定个人权益保护经验的律师和个人代表，来参与我们的数据联合体的探讨会议。
We have no answer yet but look forward to have some clues by inviting individuals to join our discussion in the next 4-6 months.
PLN是一个有其资金支持时长的项目，因此我们即将迎来最后一场工作坊。然而我们希望整个 PLN 的联系能够在之后继续维持，不单单是这一届的各家联合体，更希望借由 PLN 能够促进全球不同联合体间的经验交流。虽然在这一批中，我们应该是相对最为商业化的一家，但是也能够找到一些问题上的共鸣，比如大家都要找到合理的生存下去的方式，一个商业模式。这也就是 ODI-MS 合作能够给大家带来的好处，是他们投入的研究和最尖端的科技的支持。衷心希望之后的联合体能够更好地享受整个 PLN 的过程，也希望我们的工作能够尽快落地，为更多的联合体梳理榜样。
Our journey of PLN finally comes to an end. We hope that we can maintain interaction with other peers after the program and also can join the next cohort to build a international community of data collaborations. Although we probably have the most commercial data collaboration among current cohort but we do share common challenges such as how to sustain each data collaboration. We hope that we can eventually make our data collaboration work and become a new example to many others.