Learn how to avoid common API misconceptions and the resulting disillusionment.

APIs sometimes are perceived as having almost magical properties. Since APIs are so essential to digital transformation, powering all of our digital experiences, the conclusion is that once you have and use APIs, all the wonders of digital transformation are simply going to happen. Of course, that’s not necessarily the case, and the purpose of this article is to explain some of the useful things where APIs play an essential role, but where APIs are necessary but not sufficient.

Working with many global organizations in the API space, most recently as part of the Axway Catalyst Team, our role very simply is helping customers make good decisions. This, of course, also involves helping customers avoid bad decisions. One such bad decision is to have an incomplete understanding of the complexity of digital transformation and APIs. It’s easy to expect too much from just “doing APIs” without making sure that all necessary factors for success are in place.

I’ll explain five common misconceptions about the powers of APIs. We expect these good things to happen when using APIs, but we also have to understand that they are not quite as simple, and that success will depend on additional complexities also being addressed. Better understanding these complexities from the very beginning will help to avoid disillusionment. It will also help to make sure you consider all necessary ingredients for a successful API upon starting your journey.

Misconception #1: Untangling Monoliths!


APIs are essential for improving modularity, allowing entangled IT systems to be decomposed into loosely coupled ecosystems and thus being easier to maintain and extend. However, modularization is a hard task where finding the right boundaries is the main challenge.


One common motivation to use APIs is to get rid of monolithic solutions that often have a variety of problems associated with them, most importantly being hard to change, hard to scale, and hard to integrate with other solutions. By definition, APIs (in their modern sense of “networked interfaces”) are essential in this picture, because untangling means that the resulting modular structure needs communications to function, and these will be using APIs.

The hard work in this picture, though, is to come up with a proper modularization where the resulting structure follows structured design principles. There are many theories and practices on how to come up with a “good” structured design that represents a complex system. The hard work of doing it will lead to components that then communicate through APIs, but these APIs are an effect and not the cause of the untangling.

Sam Newman’sMonolith to Microservices” provides a useful guide on how to systematically decompose a monolithic system into one that is composed of many communicating components. It is telling that the book has a lot to say about decomposition patterns and strategies, but focuses little on how to design communications between the resulting components. The success of a good decomposition is defined by identifying the right components and then carving out parts in ways that allow the overall system to remain functional throughout the transition. APIs will be created as a side-effect of doing this, but success is a function of decomposition patterns and strategies.

The worst-case scenario is that APIs will make things worse because decomposition was not done correctly. In this case, the newly created API-based architecture suffers from chatty APIs, and the APIs themselves are not all that useful for unlocking new value. This result often is referred to as a “distributed monolith,” but this result cannot be attributed to “APIs negatively impacting performance;” it is just a lousy system design. Also, bad API design choices may further negatively impact the resulting system, but the root cause, in this case, most often is a design where components are not loosely coupled.

Misconception #2: Scaling Up!


APIs allow you to more easily scale IT resources by only scaling up those parts that need to be scaled. However, scaling most critically depends on modularization and shared state issues, which are are hard design tasks for decentralized architectures.


Improving scalability is one of the reasons why architects consider APIs. When you have decoupled components, the reasoning goes, it becomes easier to scale up those parts that need scaling, while other parts of the system can be left alone. In many cases, when people consider microservices or serverless, one of the main values they might be interested in is scaling. For microservices, you then might have to invest in how to implement scaling with something like Kubernetes. For serverless, scaling works easier, were given a sufficiently capable platform, scaling can be done transparently and is simply a question of how often serverless functions get invoked.

Since APIs are essential for both microservices and serverless (they are part of the definition of what these things even are), once again, one might conclude that with APIs in place, scaling becomes much easier. Once again, this is not necessarily the case. When it comes to scaling, and in particular, the selective scaling that API-based components can deliver (only scaling those components that need to be scaled), then, once again, architecture is much more critical than APIs.

A little while ago, when Representational State Transfer (REST) was a much-discussed topic, one often overlooked aspect of it was that one of the REST constraints is for services to be stateless. That way, each invocation of a service strictly is just a function of the invocation (and not of some shared conversation state). This design allows scaling to be simpler because services can be scaled without having to keep track of the conversation state. This is a useful design, but it is one that depends on service designers specifically creating stateless designs. For stateful services, scaling gets considerably harder because requests cannot as easily be used for scaling up as required.

As before, there also is a worst-case scenario. If the system design does not consider shared state, scaling may have unintended side-effects. It still may be possible to scale the component on demand, but scaling up the context that it needs (because it is not stateless) may have an unfortunate ripple effect. This possibly can be mitigated by sophisticated state-sharing approaches, but the underlying problem often is that of an “implicit shared state.”

Misconception #3: Monetization!


APIs can be new sources of revenue, allowing cost-effective experimentation with digital offerings and creating new revenue streams beyond the traditional sources of income. However, only a small percentage of API strategies involve direct monetization.


There is an abundance of presentations that highlight how some companies make a lot of money with their APIs. The most prominent examples may be Stripe and Twilio. Some of these presentations talk about the “API Economy” and then jump to the conclusion that this means money is being made by charging for API access. For organizations considering their API journeys, this looks like a tempting route, opening up new sources of revenue just by exposing APIs and charging for them. Some API vendors also rather aggressively advertise this sort of direct API-based income, hoping that potential customers will factor this into their ROI analyses.

However, the reality is that it is a small minority of cases where API value manifests itself directly in paid API access. In the majority of cases, API value manifests itself in enabling more value chains, and a culture in which these value chains can be built rapidly and with little coordination. For this reason, the popular “API-as-a-Product” moniker is a rather poor naming choice, as it seems to imply that the API itself is a product that can be monetized. Talking about “Product as API,” on the other hand, makes it clear that value is created through products, and that APIs simply are the connective tissue that allows this value to be unlocked and used more easily.

Once again, it is undoubtedly true that APIs are necessary for the value being created in today’s digital world, and that improving API strategy increases your chances of realizing more value. But the value is in what you do and how you allow these things to be composed in digitalized value chains. If what you do has little inherent value, then your efforts of digitalizing your scenario may not create a lot of additional value. And even if what you do has a lot of inherent value waiting to be realized, it is unlikely that exposing just the right API is what keeps you from realizing that value. What we see much more often are companies struggling to do the right business and organizational changes, so that the latent value in their products can be realized by building great and valuable experiences around them.

Of course, we also have a worst-case scenario here. If the perception of API monetization is such that API value manifests itself with direct revenue, this introduces an incomplete value analysis for the API journey. Nobody has yet figured out the perfect formula on how to precisely evaluate the value of an API (so that investments could be targeted towards API-enabling those capabilities first that have the biggest potential). Still, it is certainly not as simple as looking at the revenue generated through API access. Fully understanding value chains and evaluating and predicting their success has never been a hard science for as long as organizations have started planning and building them. This is no different for digital value chains, so make sure that you take a holistic view.

Misconception #4: Communications!


APIs are the way how to communicate in digital environments, and via APIs, all capabilities can easily interact throughout the entire organization and beyond. However, this only works when the communication channels are designed and used to rely on a shared model of describing and understanding things.


APIs are a language: They allow peers to communicate to achieve some goal. As such, APIs supposedly should enable teams to become more decoupled, thereby reducing the communication and coordination overhead. In an ideal scenario, inter-team communications entirely happen through APIs or associated channels (such as documentation and support channels). In reality, things often are a bit trickier, and that’s in part because the language of the API needs to be at a useful level for everybody involved.

For example, one scenario we see often is that API efforts start on the IT side. One initial initiative may be to create “System APIs,” which often are relatively low-level APIs representing capabilities provided by certain systems. This often translates to a poor experience for those teams a bit further away from system details: They are looking for solutions to their problems, and not for access to implementation details. It often takes quite a number of these “System APIs” and their orchestration in specific ways to accomplish tasks that, from a business point of view, should be relatively easy to do. There is a “language mismatch” of the APIs being built from the system point of view, and the API consumers looking for APIs that address their domain problems. (For more information about system APIs, check out this video asking “Are System APIs a Good Idea?”)

This problem is best solved by looking at APIs as business-level capabilities: They must have a value proposition why others might use them, and solving that problem in the easiest way possible then becomes the guiding principle for identifying and designing APIs. While this may sound easy, it often is a hard problem to solve because it requires a functioning dialog between business and IT. This may be a natural thing for companies born digital. However, it can be a real challenge for companies where business and IT are still two relatively disconnected parts of the organization.

The worst-case scenario is that technical APIs deepen the rift between business and IT. IT thinks they are delivering value because they make capabilities available. Business thinks “this is just another tech thing” because they don’t see clear business value being exposed in the technical APIs. The remedy for this is for both sides to understand each other better, and to work towards establishing and using a shared language. IT needs to focus its efforts on exposing APIs that have a clear value proposition and accomplish a business purpose. Business needs to start “talking API,” not to the point where they design JSON that they expect as responses, but to the level where they think in terms of designing and delivering products as digitalized value chains, based on existing APIs, and exposing their APIs. That is, in essence, what a true “API Culture” is all about.

Misconception #5: Organizational Issues!


APIs can help with improving the way how teams communicate and how the organization can innovate. However, APIs can only help the organization to communicate better, but they will not cause organizational changes.


Large organizations are slow at adapting to changes, while more modern organizations are demonstrating how to improve agility through reorganization. Such companies are re-organizing the way they build products, and this reflects itself in API-centric organizations and cultures.

There is no way escaping Conway’s Law, meaning that to build something that is non-monolithic and can be changed more efficiently, you need an organization that has the same structural properties.

APIs play an essential role in this picture because they significantly reduce coordination overhead. Instead of doing one-off “integration” projects, capabilities are exposed through APIs, and anybody needing them can simply access them, ideally without ever talking to the team that owns that capability. This approach scales better and creates an abundance of options when it comes to teams designing new experiences and looking for existing skills that they can leverage. In a perfectly functioning organization, teams are decoupled and can work at their own speed, delivering and evolving products as needed.

In reality, changing a large organization to this ideal picture of decoupled and empowered teams is a much harder challenge than designing the APIs that they will produce and need. And of course, even if an organization has been changed successfully, it still takes careful design and execution of an API strategy and an API program. But in our experience, once it comes to these API-specific issues being the remaining problem, the hardest part of transformation has been solved.

The worst-case scenario for this is that organization leaders see digital transformation first and foremost as a technical challenge and focus on technical indicators. Once APIs are created, the assumption is that good things will follow. We see many organizations losing months or years of precious time because they don’t fully appreciate the implications for the business and organizational pillars. In our work with large companies, a substantial share of our time is to make sure that all the pillars necessary for digital transformation (business, organizational, technology) are addressed. From that perspective, APIs simply are a side-effect of digital transformation and not a cause.

Conclusions

This article presents five common misconceptions about what APIs can do, and makes the point that they actually can do these things, but not by themselves. APIs are necessary but not sufficient for many important challenges and opportunities in today’s increasingly digital organizations. The most important takeaway of this article is that it always takes a holistic view: APIs are part of many solutions in the digital space, but they are never the solution themselves.

Like all popular technology trends, APIs are going through the hype cycle. At first, they seem to be the solution to many problems. Then comes disillusionment, when it becomes clear that they are “just plumbing” and that many other factors go into making APIs useful other than just having them. After this, there is the hard work to get onto the plateau of productivity, where APIs are an essential part of many solutions in the digital space, but other factors go into these solutions as well.

Luckily, organizations nowadays can benefit from a lot of experience, they can choose between many available products, and they can rely on trusted advisors for their journey. All these issues discussed above are very real and relevant challenges for many organizations today. Make sure that your API strategy and your API program take a holistic view. Many reports about “digital transformation initiatives” show high failure rates, and by looking beyond just the technical aspects, you drastically improve your chances to make meaningful changes that truly transform your organization.

This article is based on a presentation given at the API Conference in Berlin in October 2019. The slides are available online at http://dret.net/lectures/apicon-2019/api-limits.Watch the presentation below:
Erik Wilde

About Erik Wilde

Erik Wilde focuses on technologies, standards, and design issues of open service landscapes. His areas of interest include Internet and Web of Things, semantic technologies, service description and documentation, and questions of how to design for openness and evolvability. Erik is a frequent speaker at key events; follow him on Twitter.