If I wasn’t in IT, I’d like to be a chef. Cooking is one of my favorite hobbies for two reasons, starting with the irresistible joy in cooking red sauce over a stove with a glass of wine in hand and the scent of oregano in the air. Second, cooking feels like a metaphor for my job as CIO at a large data company. Data is the food I cook with every day. It gets sliced, salted, and served a million different ways to suit the end consumer. In many ways, APIs are like data recipes – creating internal structure and order to find new ways to use data to ensure that companies can innovate – and produce better outcomes for consumers.
As CIOs are increasingly called to create and launch new business models and expand topline revenue opportunities, APIs enable power. APIs create a more open, adaptive and nimble future, where CIOs can move beyond small tweaks to real innovation.
[Want more advice from CIOs on driving topline revenue? Download our HBR Analytic Services report, Revenue-generating CIOs: Smart strategies to grow the business.]
Data is the lifeblood of tech and global information services companies. Here at Experian, enterprise data was growing at an exponential rate, and our software engineers were looking for innovative ways to process data files more efficiently. As Experian started to promote the idea of creating an API extraction layer that aggregates and transfers specific data requests for easier data-sharing, our clients were also looking for better ways to process information and create consumer value. They wanted the opportunity to dictate which information they get, how they get it, and not be bothered by information they didn’t want.
Change is always risky for established companies, but being nimble is now a requirement. Consider these lessons I learned as our team geared into the speed of API.
Focus on talent
When you’re the executive chef of a restaurant, the most critical part of the job is assembling a talented team that delivers dinner accurately and on time. You need the sous chef, line cook, pastry chef, and dishwasher with a vision all working independently and together to create a seamless dinner experience. Being a CIO is no different – you need to find talented developers and architects with experience to get the job done.
Precursors to this development approach include things like DCOM, Publish and Subscribe, and Corba. The big difference is the simplicity by which the recent wave of distributed compute capabilities can be implemented. Software developers can pick up this approach more quickly and many universities are teaching these concepts to computer science majors. Architecture is slightly more challenging since understanding how to connect various components and leverage external third party services can be simple or extremely complex, depending on what is being built. Having said that, most strong architects understand the need to embrace the model for scalability and portability and as a requirement to meet customer needs. So the supply of expertise will grow as more companies depend on these capabilities to build applications.
Get team buy-in
For Experian, the biggest challenge is ensuring buy-in from team members around the world. As Experian began implementing API capabilities with the right people in place, it was critical to provide a road map of the direction that the business is going to both to employees and clients. The best results were achieved when developers had exposure to the vision for how their work was making an impact on our business.
My advice: The greatest way forward is one of inclusion. Bring the IT organization, including operations, into the conversation as early as possible. That way firewall issues, security concerns, network bandwidth issues, etc. can all be addressed during deployment as opposed to the typical last minute approach. This applies to both providers and consumers of microservices and API methods.
Most importantly, collaboration is key – and not only between people, but also between people and technology. Research shows how teams of humans and robots can far outperform teams with just robots or teams with just people. Silicon Valley start-ups are exploring ways scientists can work together with AI to help spur advances in molecular biology. Data needs a similar approach. Companies need to reorient how they assess data. Breakthroughs in medicine, healthcare, AI, robotics, even keeping Moore’s law alive, all depend on data analytics.
Good news: Microservices and the API economy, by default, help foster collaboration. One of the basic principles of this technology is re-usability. When an existing service exists, it is more practical to share than to rewrite a capability.
When you’re using APIs, governance and communication become critical to ensure alignment and to maximize efficiency. But governance is by far one of the greatest challenges. If people go rogue and do not follow a robust governance model, you can quickly have a disaster on your hands.
Especially when multiple lines of business within the same company are publishing entry points that may be consumed by the same customer, customers have an expectation of consistency. Consistency around data models and data management are critical. Large financial services customers expect a “customer” record to look the same across multiple systems. Any company with multiple definitions of a “customer” runs the risk of alienating developers who expect a common definition. It is important that governance be considered before any kind of broad adoption is contemplated.
Can APIs help your company be smarter?
At a time when explosive developments are emerging rapidly across nanotechnology, machine learning and AI, the pace of progress can’t be stopped. CIOs like myself must constantly take a hard look at our business and industries to ask ourselves whether technology like APIs can help our companies be more precise with data, deliver more efficient service, and innovate.