Future Perspectives on the API Economy

As someone who creates integration products, I spend a lot of time researching industry and technology trends while talking to analysts, engineers, architects, target customers and my peers. This work inevitably shifts my perspective towards a version of “what’s happening now, what’s likely to happen over the next few years, and what is my role in guiding the industry to the best possible future?” This article aims to provide a synthesis of the most impactful ideas of the past year and their influence on my future thinking as a connectivity product manager. Hope you enjoy the read and look forward to your thoughts in the comments.

APIs are part of the Internet fabric

For some students of modern technological history, the “connectivity” part of the internet was very different just a few decades ago. Through “connectivity“I mean APIs, protocols like HTTP, and agreed architectural models that unlock data. As a result, tech professionals talk about “Modernization of the heritage” projects aimed at exposing old technological silos that would otherwise remain hidden from the digital lifeblood of the business. These so-called digital transformation projects often relied on XML-RPCs to enable integrations with mainframes as the new digital age brought standards such as REST, GraphQL, and Web of Things.


Free for commercial use. No attribution required.

As established companies invest in new APIs to support digital transformation projects, early startups are building on the latest technology stacks. This trend is transforming the Internet into a growing web of interconnected technologies unlike any we have seen before. As the number of new technologies peaks, the underlying fabric – otherwise known as API economics – is fueling the market to undergo technology consolidations with the historically high number of acquisitions.

This trend has two interesting consequences. The first is that all of this drives the need for better, faster, and easier to understand APIs. Many Integration-Platform-as-a-Service (iPaaS) vendors understand this fairly well. Established iPaaS solutions, such as those from Microsoft, MuleSoft, and Oracle, are continually enhanced with new tools as new entrants, such as Zapier and Workato, continue to emerge. All of them are investing in simplifying the onboarding experience in addition to APIs, which essentially speeds up the onboarding time (a level of growing importance in business agility). Some call these experiences “connectors” while others call them “models”. But in the end, the great minds of integration are actively involved in this area.

The second consequence is well-defined, protocol-based connectivity. Looking at the world of REST – a well-accepted architectural style defined in Roy Fielding’s thesis – we see that REST APIs dominate the scene with well-established specification standards such as the OpenAPI specification (formerly known as Swagger ). Not only do these protocols allow cutting-edge iPaaS solutions to agree on what the next world of connectivity will look like, they also lay the groundwork for new experiences – often referred to as innovation – to evolve. More and more technologies continue to emerge, offering visualization and transformation products that understand these standards while bringing more users into the world of connectivity.

I am excited about the potential of this space and its ability to define the fundamental building blocks of the future Internet with APIs as the centerpiece of its fabric.

Also: APIs, microservices succeed as long as the organization does not interfere

Break silos with indexed search and browser-style API discovery

Moving from specialized tools and standards to a simple API discovery layer means that any employee who can write queries and logical flows will also be able to build full-fledged applications and customer experiences. Many leading analysts are now seeing this dynamic as more APIs are consumed by less technical departments like marketing, finance, sales, and human resources.

I see this trend evolving further in two main forms. The first of these is finding and discovering universal APIs. Many of us use Google to search for information, and “Google” endpoints (the addressable location of an API) and data shouldn’t be any different. This means that more tools will evolve, but the approach we take will be fundamentally different; Instead of manually documenting new endpoints with API references and portals, we can start indexing new APIs dynamically based on their machine-readable descriptions. By using techniques similar to the tactics of Google’s crawlers that discover publicly accessible web pages, more users will have access to all endpoints and publicly accessible data.

The second form is about how we explore these APIs and the data they contain. Today, many developers start by searching for an API portal, finding a relevant SDK, and sampling the capacity of an API with API consumption tools like Postman. Less technical users, however, are turning to low-code / no-code solutions that bridge the technical divide by demystifying API access (a skill typically reserved for software developers). It is interesting to think about what will change as the underlying foundations of these protocols and standards evolve. I think we’ll see more browser-style discovery tools soon, where web pages are replaced with endpoints and information is replaced with data. In this world, users can search, query, read, and plug in data instead of worrying about technical API details like URIs, endpoint syntax, query parameters, etc.

Looking ahead, what I find most exciting about this development is that we will see new digital capabilities being created that are closer to the end user and much faster to build. These innovations also trigger a need for business professionals to have a big picture of how everything is connected, while product managers and CIOs need to pay close attention to inconsistencies in the customer experience. or potential compliance, privacy and security issues.

Also: It turns out that low-code and no-code are also valuable for professional developers.

Producing connectivity: protocols vs connectivity as a service

More than ever, users demand access to data. However, many existing solutions are too complex, too expensive or too cumbersome. This creates a technological void which will be filled in the following ways. On the one hand, integration professionals like me will continue to advance connectivity standards. Optimizing for ease of consumption, especially by non-developers, will lead to a new layer of API consumption, so less technical experiences can scale on top of it.

On the other hand, new business cases will be developed for building agile API-facade-as-a-service solutions. As more users demand faster time to market while taking scalability, availability and security for granted, more startups will emerge to meet this need. We are already seeing new entrants involving a productivity infrastructure as a service from Nylas and a unified API from Kloudless that connects over 150 SaaS solutions through a single canonical model. All of this makes it easier than ever to create and maintain connections with external systems.

As we move forward on each front, I suspect the industry will first have to agree on common architectural models as we build new solutions around them.

Data is the new endpoint of security

Data breaches tend to increase, with a record 1,767 publicly reported breaches in the first six months of 2021. Our most common attempts to secure data focus on protecting the infrastructure that provides access to it: terminals. While this approach makes sense for some organizations, as we move more infrastructure to the cloud where the infrastructure is much less under their control, securing that infrastructure becomes more of an issue. We’re adding more users into the mix who can now search, query, and share data with their favorite apps, and we’ve got a recipe for disaster.

To stay ahead of these trends, we must first change our mindset. Instead of protecting endpoints in the new digital world, we need to protect data. This space is full of exciting innovations with new encryption and tokenization standards that further propagate the zero trust model. This trend is also recognized by new startups who are building businesses around the idea of ​​protecting data with encrypted data vaults and use cases ranging from securing PIIs to providing data stores. encrypted HIPAA compliant.

No matter how we evolve our new API layers, at the heart of the “secure” approach will be our ability to discover and work with sensitive data.

Also: API security becomes a “top” priority for business stakeholders

The bottom line

We’re still “rounding the first ground” in terms of defining the next generation connectivity layer and understanding the types of businesses that can be built on it. As APIs are already at the center of many digital transformations, we are clearly seeing a trend of simplifying API consumption with low-code / no-code solutions that get more users to build pluggable businesses. It is gratifying to think of a world where everyone can contribute to the betterment of the business.

Anton Kravchenko is Product Manager at MuleSoft, a Salesforce company. If you are considering or building any products or protocols that touch on any of these ideas, he would love to hear from you.

Previous Book review: Rabbits, Sophie Overett
Next Dear Penny: How do you get a credit score after 30 years in prison? | pennyhoarder