With the rise of the Microservice business model, more and more companies are using third party plugins within their services. For example, if your new company needs to build a location based App, more than likely you would look to an existing location API, like Uber or Google’s, that you know would work well. Another example, you want to add weather forecasts to your service, would you be more willing to build this from scratch? Or, look to something like Open Weather Map or AccuWeather API to perform that function for you? These days, more often than not the answer would be to use the third party API.
How do you choose the correct third party API? Research would be the best answer, but research needs quality, bias free data to be beneficial, and that type of data is almost impossible to find on APIs. Currently, you are relying upon the API’s SLA on Uptime and Latency, however if you start shopping around you notice that almost everyone says their API will be up 99.9% of the time.
We all wish every API was working as it should 99.9% of the time, but this just isn’t the truth. Now the question is, how do you find the actual Uptime, and average Latency? This is where a tool like APImetrics and the ability to run reoccurring automated API tests can be so very useful.
Let’s go back to the Weather API example, where we were wanting to choose a third party API to run Weather Forecasting within our APP. A quick search yields a Top 3 Weather API list that reads as follows: 1. Open Weather Map 2. AccuWeather 3. The Weather Channel. (Let’s assume that Price, Documentation and Call Limits are all the same for this example.) With a combination of APImetrics Public API Health data, and the APImetrics tool, you would be able clearly see which API has the best Pass Rate, Avg Latency, and Max Latency, then subsequently make the correct decision for your service.
Beyond just choosing which API, Reoccurring Automated API Testing also assists with staying ahead of errors and user complaints. We have all been in a situation where something crashes and the first notification you receive is a flood of angry responses from Users, which sends you scrambling to find and fix the issue. With an automated test set up the moment there is an issue, you would be the first to know.
As the API Economy continues to grow, there is an ever increasing number of APIs to utilize. Make sure you are making educated decisions based off hard data when choosing which API to utilize. Once you have found that correct API, keep an eye on it with APImetrics API Performance Monitoring.
