Anant Jhingran of Apigee recently opined on GigaOm that our Big Data future will require data marketplaces and data syndication models that few enterprises are currently prepared to support. If one follows the trend toward singularity, it should be clear that an enterprise’s ability to quickly acquire new data sources will require a marketplace that is somewhat akin to the NASDAQ. In other words, it would allow data set contracts to be opened or closed in a matter of minutes or seconds.
Jhingran observes that in the new app economy, organizations no longer own or control all the data they need to make accurate business decisions. If an enterprise is to plan ahead, Jhingran argues that there are three ensuing new sources of data that organizations must be able to capture, measure, and analyze to get a complete view of their customers and businesses.
The first data source is generated by outside organizations that access the APIs your company chooses to make available to the outside world.. The second is data generated by an enterprise’s own applications that are calling an API, and the third is contextually-relevant data culled through external APIs — such as Github, Twitter, or Stackexchange -- and countless others.
Decades ago, companies would build a Decision Support System to aid executive decision-making. These systems rarely had a real-time component; rather, on a weekly or monthly basis, IT would gather copies of massive amounts of data from various servers or mainframes to create a data warehouse that could be queried by human business analysts. In the future, more and more companies will need real-time decision support for analysis by artificial intelligence engines, a radically different model than the days of old. Yesterday's methods and models of information management grind to a halt in such a hyper-analysis model.
The key takeaway for enterprises is that we need to think about how to exercise more control and discipline over API usage. In the future, the enterprise will want to better monitor and control how it uses internal and external APIs and how partners and the public use the APIs the enterprise provides to them. All of this will prompt a discussion about the cost and benefits of supporting APIs both inside and outside the enterprise. This does not mean that all enterprises will be in the business of selling data. Rather, because of how much API activity will explode in the coming years, an enterprise will want to control costs associated with its API infrastructure. Similar to the way the Internet is funded today, enterprises will want to provide free access to APIs for its partners and peers, while having a business model to charge for API access to those relationships that are one-sided.