For years, pundits and thought leaders have talked about data economies, churned out metaphors such as “data is the new oil,” and generally broadcast the idea that if enterprises aren’t thinking about ways to extract value from data, they must be missing out.

Even so, many enterprises aren’t extracting as much value as they’d like from their data, with surveys suggesting a majority of “big data” efforts fail. One reason is that the data isn’t being integrated in the right ways and circulated among the people who can make use of it.

Think of currency. If all the money in the world were locked in a vault, never circulating, it would not actually have any value. Interactions and transactions around currency are what make it useful. The same goes with data. If data is just sitting on a server somewhere, not packaged so others can interact with it, or if it’s being exposed to only a small group of would-be innovators, it probably won’t produce much value.

How can businesses make data accessible so innovators can leverage it to create new value streams? Siphoning value from data may require many strategies and technologies, but the place to start is almost always application programming interfaces, or APIs.

APIs are how software talks to other software and how developers leverage digital assets, including data, to create new applications. API also aren’t new. Every company has APIs — possibly thousands of them — in some form or shape. So why aren’t all these APIs already helping companies turn their data into new value? Because not all APIs are created equal. An enterprise won’t make much progress putting rich datasets behind sloppy, undocumented APIs that can’t be easily consumed by developers outside the teams in which they were created.

To create APIs that are easy for developers to consume and that encourage developers to leverage a business’s data, an organization may have to rethink its approach.

At some companies, for example, APIs are constructed on an as-needed, ad hoc basis by individual teams. This can result in many bespoke APIs, many APIs needlessly duplicated, and only a few APIs documented well enough and designed consistently enough to be of ongoing use.

Instead of perpetuating these mistakes, leading digital enterprises increasingly apply product management methodologies to their APIs, including product managers and product teams overseeing a full lifecycle from conception to market introduction and evangelism to maintenance and iteration. Put simply, APIs should not be seen as technological minutiae curated by IT teams but rather as strategic business products that translate data and other valuable intellectual property into a form developers can leverage.

Developers are essential participants in the process of turning data into value because they use APIs to build applications that add additional services atop the data and that produce new connected experiences for end users. Most businesses should expect to rely on internal engineering talent for some of this development — but many of the biggest, most data-driven companies turn to external developer ecosystems to help broaden their scope of innovation.

AccuWeather* makes its weather data available to third-party developers, for example, and as a result, its services are being leveraged for a wider variety of applications, and its intellectual property is being used, experimented with, and innovated upon in more ways. More than 60,000 developers have signed up to use AccuWeather APIs since the company launched a developer portal in May 2017.

Google Maps is one of the most familiar APIs that has attracted an external developer ecosystem. Many people who’ve ordered a car through a ridesharing service may have noticed Google Maps powering the service’s mapping and navigational capabilities, for example. What may be less obvious is that this relationship works both ways; Google’s APIs are not only consumed by the ridesharing service but Google also sometimes consumes APIs from the ridesharing service, which is what allows Google Maps users to order ridesharing directly from within the Google Maps app. Google’s mapping data can be made more useful because it’s been shared with ridesharing services, and the ridesharing services’ data can be made more valuable by being shared too. The sharing — the circulation of data — is what drives the value.

APIs are “how to unleash data in new ways,” said Kate Bae, vice president of product leadership at Nielsen*, a global information and measurement firm that provides market research on consumer habits. Nielsen works with dozens of companies through its Connected Partner Program to aggregate data and package it for different needs, and it offers APIs through its developer portal so that others can build applications that leverage the firm’s wealth of information, from audience segmentation data, to TV ratings data, to retail measurements.

Though developers and APIs are common factors in translating data into value, enterprises still have many different pathways to success. Some companies, for instance, may find that giving partners and developers free access to data is the optimal way to drive adoption and innovation — and ultimately to turn the data into revenue. Others may find that because the data itself is already so valuable, the best path is to monetize it directly by charging developers for API access. Datasets and data sources are diverse, and so too are the strategies and tactics for maximizing those sets and sources’ value. Even so, the first step to untangling this diversity — and to develop ecosystems of innovators who turn data into value — is to treat API as products.

*AccuWeather and Nielsen are clients of my employer, Google Cloud’s Apigee team.

Originally posted on Apigee by Anil Sagar on March 30, 2019.