The State of API 2020 Webinar – Exploring Trends, Tools & Takeaways to Drive API Strategy

Editor's Notes

• #5: The insights provided in this survey were gathered from a global audience of API product owners, architects, developers, QA and test engineers, as well as contributions from both the Swagger and SoapUI open source communities. The report includes perspectives on companies of all sizes, from early-stage startups to established enterprise organizations, and touches on a wide range of industries, including financial services, telecommunications, manufacturing, healthcare, and many others.
• #6: The year 2020 will be recorded as a significant inflection point in the digitalization of business. The impact of a global pandemic and the ensuing business and societal changes have forced many organizations to reprioritize their API strategy and accelerate digital transformation initiatives. The 2020 State of API Report builds on the previous year’s survey to cover multiple stages of the API lifecycle – design, development, documentation, testing, and monitoring. We’ve also included questions into the technologies, standards, and tooling that intersect with APIs. We’ve broken the report into two primary sections: API Strategy, which covers key trends, sector growth, and the primary challenges identified from the 3,000 plus survey responses; and API Categories, which is a deep dive into results across six major categories: API Tools and Technologies, API Consumption, API Testing and Quality, API Documentation, and API Design. APIs continue to grow in number and importance. The better you understand the trends, tools, and technologies within the API ecosystem, the more effective you’ll be in using them. APIs can drive automation across your SDLC, and ultimately deliver new digital applications faster and with higher quality. For those of you that took the time to participate in this survey, we appreciate your willingness to share your knowledge and experience to help shape the success of future API initiatives.
• #7: API adoption is still relatively new for most organizations, with more than half of the organizations only starting to create APIs in the last five years. Despite the promise of digital transformation, for some organizations APIs remain a nascent technology that has yet to be exploited for business gain. Of the companies surveyed, less than 20% have spent more than 10 years developing APIs, while the majority of companies (55%) reported less than five years with API development.
• #8: Interaction between internal systems continues to drive API development, as two-thirds of organizations identified this as their main driver. APIs are the glue that connect applications and data, eliminating the need to hand-code each unique integration. Interoperation between internal systems, tools, teams grew slightly year over year, with 64% of the organizations citing this as their number one driver for API development. Second, and close behind as a driver for API development, is Reducing development cost/time (efficiency) at 58%. For the most part, the drivers behind API consumption ranked in a similar order to last year. In that way, they stayed parallel to the drivers behind providing APIs. This makes sense, since API providers and API consumers make up the supply and demand of their unique markets. The three drivers with significant increases year over year were Partnering with external organizations (+8%), Absorb content (data/features) from external products (+7%), and Interoperation between internal systems, tools, and teams(+7%). Interconnected systems, tools, and 3rd party partnerships increasingly drive API consumption. For the most part, the drivers behind API consumption ranked in a similar order to last year. In that way, they stayed parallel to the drivers behind providing APIs. This makes sense, since API providers and API consumers make up the supply and demand of their unique markets. The three drivers with significant increases year over year were Partnering with external organizations (+8%), Absorb content (data/features) from external products (+7%), and Interoperation between internal systems, tools, and teams(+7%).
• #9: Which API technology challenges do you most hope to see solved in the near future. API standardization continues to be the top challenge that organizations want to solve. The adoption and growth of APIs continues to grow exponentially. Standardization continues to rank as the top challenge (58%) for organizations as they attempt to scale API development. With development increasing, so has the challenge of versioning. The biggest shift, year over year, was the decline of Composability/multi-purpose re-use as a challenge. Customers’ preference for OpenAPI continues to be high, but there’s dramatic growth for GraphQL and gRPC standards. API standards continue to emerge to meet new challenges, with some showing increased adoption. Year over year, the number of customers selecting GraphQL and gRPC have shown dramatic gains, almost doubling for gRPC and just over a 50% gain for GraphQL. Yet, despite growth in these areas, the overwhelming choice of developers continues to be the REST OpenAPI Specification. Originally, this API standard was introduced in 2011 as Swagger, then later in 2017, it was formally released as the OpenAPI Specification. Today, 82% of the survey participants cite the OpenAPI/Swagger as their API standard of choice for defining APIs. Organizations realize the value of API style guides, but for some it’s still an aspiration. With standardization being the number one challenge cited by organizations building APIs, you’d think API style guides would be a given. This year’s survey data indicates that 64% currently have an API style guide or plan to add one. However, behind that metric is the reality that for half of those organizations, an API style guide is an aspirational goal. For another 24% of the respondents, they don’t have a style guide and one is not planned.
• #10: Documentation is rising in importance for API consumers, who still value ease-of-use above anything. While the top 3 most important characteristics in an API remained the same year over year, Accurate and detailed documentation beat out Responsiveness/ performance for the 2nd position with a significant 13% increase. For context, when we asked the same question in 2016, only 25% of respondents selected documentation as a top factor. This year, 63% of respondents chose it. Another sizable year-over-year change was responses for Scalability of underlying architecture, which decreased by 29%. This could be due to the changing web services market and the scalable nature of services like AWS and Azure. API documentation represents big opportunity for improvement. Documentation is so important to the successful adoption of an API, and yet 43% of survey participants indicated their documentation was poor and needs improvement. Most organizations (52%) rated their API documentation Good to Average. Only 5% of the survey respondents felt their API documentation was Very good.
• #11: Performance and uptime are top measures for API success. Measuring the success of an API has become increasingly important, as organizations rely more heavily on APIs for both internal interoperability and external engagement with customers and partners. For the second year in a row, Performance is rated by customers as the highest measure of API success at 72%. Second is the ability to ensure API Uptime/availability, cited by 57% of the customers surveyed. While Performance rates number one overall for API success, to the API consumer Ease of Use and Accurate Documentation are the top choices.
• #12: Microservices is far and away the leading technology area in API growth over two years. Microservices is rapidly gaining ground with many organizations for its fine-grained modularity, flexibility, and speed, all of which are well suited for today’s digital application development. 65% of survey respondents chose Microservices as the technology that will most likely drive API growth. Organizations cited DevOps as the second most impactful technology on API growth coming in at 42%. Limited skills and experience pose the biggest challenges to microservices success. The process of adopting microservices is being throttled by the trouble organizations are having finding developers with appropriate skills. More than half of the people surveyed (52%) cited Experience or skills as the primary obstacle to implementing microservices. Complexity of existing code adds to the challenge of limited skills, with more than one third (38%) of respondents indicating that what’s already in place is too difficult to easily deal with. To sum it up, 67% of the respondents cited the Lack of people and time as the biggest obstacles to getting microservices implemented.
• #13: As more companies move to cloud, it isn’t a surprise that native tools like Amazon AWS API Gateway would get a large mention. Amazon’s API Gateway leads the list with 33% of the participants marking it as their choice. Microsoft Azure comes in third with 22%. Just edging out Microsoft, and somewhat a surprise, is 24% for We aren’t using an API management tool. In fact, 40% of the participants either don’t use an API management tool or use an in-house developed tool.
• #14: The days of static websites with limited functionality are gone. The need to meet customer expectations for engaging online experiences no doubt contributes to why survey participants continue to cite Web as the number one use case for API support (24%). Use cases for APIs that ranked close behind were Mobile (17%), and Back-end system integration (16%), which, just like web, are both intrinsically tied to enabling digital commerce. Also worth noting is the increase in API use cases associated with Events, which could indicate more organizations are implementing event-driven architectures that can dynamically react to customer behavior.
• #15: Collaboration between Dev and Test teams are of utmost importance for organizations. Modern software development has definitely become a team sport. Dev teams are now located throughout the world, and progress on software takes place around the clock. Collaboration has become so important that 76% of organizations surveyed indicated it merited their highest level of importance. In fact, only 8% of the survey respondents indicated collaboration was only slightly or of no importance. Dev teams move away from email in favor of collaboration through Jira and Slack. Team communication is shifting away from email, as technology professionals find it easier to collaborate via project management tools like Jira, and collaboration tools like Slack. Email dropped in mentions by almost a third, while Jira and Slack increased incrementally. The leader at 27% of respondents was Jira, with Slack a close second at 24%.
• #16: When consumers run into quality or performance issues with 3rd party APIs, they first report the problem and then look at their options. This general reaction is also what we saw in our 2019 survey. 70% of API consumers expect immediate notice on issues. More than half of respondents expected fixes to APIs the same day or within 24 hours. Compared to previous years, there’s a trend that API consumers are less loyal to the APIs they work with when faced with performance issues. In 2016, only 30% of respondents said that an issue would lead them to look for a permanent alternative API provider. Their instinct instead was to review service level agreements. In the years since, service level agreements have decreased and the willingness of consumers to look elsewhere has increased, to 34% in 2019, and now 37% in 2020. Why are API consumers less loyal in 2020? It’s likely a result of more competition in API marketplaces, plus a higher demand on API reliability as tools and systems become more connected and dependent. Downtime of a 3rd party API can translate directly to lost revenue because of poor experience for consumers and the services they offer.
• #17: More than three quarters of participants feel API quality is of top importance. A vicious cycle – biggest obstacle to API quality is limited time, biggest obstacle to ensuring API quality is time lost troubleshooting and testing APIs. When paired, these two figures suggest that teams can find themselves in a no-win situation. Pressure from deadlines can result in teams not being able to ensure API quality. When rushed development and testing produce poor quality APIs, then teams must go back and expend time fixing them. Businesses that take the time to initially ensure API quality don’t have the same penalty and can reinvest in a quality-first approach for their next application.
• #18: Organizational size determines if a formal API-design process is in place. As APIs continue to grow within an organization, the challenges associated with building and managing those APIs grow as well. Results from the survey show that 86% of organizations with 10,000 or more employees state they have, or plan to have, a formal API-design process. Compare that to small businesses of less than 100 employees, in which 58% don’t have one.
• #19: As mentioned earlier, consumers view Responsiveness/ performance as a critical characteristic of the APIs they utilize. API monitoring is a key practice to ensure high API performance. In this year’s survey, we found that roughly a third of respondents (31%) have a monitoring process in place and view it as a top priority.
• #20: 1. API standardization continues to be a top priority, especially for big companies. The growth of open APIs in the last decade has been nothing short of astonishing, but that number is dwarfed by the volume of internal facing APIs. As organizations grow and mature with API technology they naturally begin to create more APIs, create more versions of APIs and start to encounter challenges with managing their APIs. In order to ensure the API created complies with established rules and standards they might turn to an API style guide to govern how it’s development. From the results of this survey we find that less than one third of the organizations actually have an API style guide. This poses a bit of a problem for organizations to reinforce API standardization. Without standardization APIs development becomes bespoke and developer productivity declines. API costs and time to market increase to accommodate changes. APIs that intersect with sensitive data or ones used in regulated environments might expose the organization risk compliance violations. It is little wonder that in this year’s survey, API standardization was chosen by 58% of the respondents as the number one technology challenger they would most hope to solve in the near future.
• #21: 2. API teams want tools that are easy to use, can be adopted quickly, and fit into their existing toolset. The options for API developers continue to grow as the opportunities for incorporating APIs into different technical use cases expands. Software Architects see the potential for APIs beyond just linking applications – from supporting a variety of architectural types, including IoT, to event-driven and microservices architectures. Still, today the predominant standard for API developers is the OpenAPI Specification (OAS), formerly Swagger. In this year’s survey, OAS was chosen by 82% of the respondents as the standard of choice for defining APIs. Worth noting is the surge in popularity for GraphQL, an API that simplifies data query, which has moved up 60% year over year. And gRPC, an open source remote procedure call (RPC) framework, which has increased year over year by 75%.
• #22: 3. Microservices tops in API growth, with DevOps gaining ground. One of the fastest growing IT architectural approaches is microservices, a set of fine-grained, self-contained services designed to implement a business capability. The speed to develop and deploy microservices makes them a perfect complement to cloud architectures. For the second year in a row, survey respondents chose microservices as their top choice (65%) for technology that will drive the most growth in APIs. DevOps and its influence on API growth is also interesting. As organizations look to improve efficiencies in the SDLC, APIs are becoming ideal enablers for automations within CI/CD pipelines. In this year’s survey, DevOps is the second highest choice of respondents for technologies to spur API growth, coming in at 44%. One last point on this: Microservices and DevOps aren’t really technologies -- they’re mind shifts. This highlights how cultural changes around Microservices and DevOps drive API growth, as opposed to a single technology. It’s worth noting as time goes by.
• #23: 4. API consumers are less loyal than they’ve been in the past. We asked API consumers how they respond when they encounter issues with APIs. Compared to our previous two surveys, this year’s data showed an increase in respondents who start searching for a permanent replacement. As more organizations offer digital services, the market is getting more competitive and API consumers have more options to choose from. Simultaneously, API consumers have higher expectations of API providers as applications become more interconnected. Since a provider’s API failing can directly impact the consumer’s application and the consumer’s audience, there is less tolerance for quality issues.
• #24: 5. Cloud vendors emerge as the customer choice for API management tools. IT professionals will sometimes confuse API gateways with API management platforms, when in fact they’re just a part of the solution. Many of the original API gateway industry leaders are recognized for providing API management capabilities – some more than others. When asked how they prefer to expose public APIs, 31% of the survey participants chose API gateways. Following a trend from a year ago, survey participants once again picked cloud providers Amazon (33%) and Microsoft Azure (22%) by large margins over their competitors. Amazon continues to lead this survey question, but the surprise second place finisher was for Not Using an API Management Tool.
• #25: 6. Aggressive deadlines kick off vicious cycles that impact API quality. Respondents told us that the #1 obstacle they face to ensuring API quality is time restraints. As companies are trying to monetize their capabilities, and new vendors enter emerging digital markets, there’s pressure to rush development. However, when we asked about the risk associated with poor quality APIs, respondents flagged the time it takes to troubleshoot and fix issues as the biggest cost. If companies push their API development too quickly and shorten the time it takes to launch quality APIs, they’ll find themselves spending more time in rework. This could kick off a vicious cycle where dev teams have to carry the burden of rework while already lacking the time to ensure quality on their next project. This year’s data reinforces the military credo that “slow is smooth and smooth is fast.” A focus on a quality-first approach, with an organizational awareness of this reality, can prevent this type of cycle.
• #26: 7. Email out. Jira and Slack in. Collaboration between development and test teams was rated at the highest level of importance by survey participants at 76%. It’s therefore logical to assume the tools best suited for collaborating would be the developers’ communication channel of choice. Let’s face it, the plethora of email, message boards, collaboration channels, chat windows, etc., was eventually going to shake out to reveal a winner and loser. In this survey of technology professionals, email is trending down by almost one third year over year, while Jira, the project management platform is leading at 27%. To no surprise, the close second for team collaboration is Slack, which was chosen by 24% of the survey participants.