When your platform runs on APIs, all of those APIs need to run perfectly. Quickly resolving issues in your API isn’t just helpful, it's mandatory. Latency and error monitoring are only the beginning: a healthy server isn’t the same thing as a healthy product. Resolving error cases and API abuse is easiest with full visibility into your API, which is where API analytics come in.
Nowadays, it’s common to use out-of-the-box solutions for authentication, payment processing, and other critical functions. Still, API analytics solutions are often built in-house, eating up months of developer resources. That’s time your team could spend further developing your product itself.
Luckily, turn-key solutions do exist. Read on to discover three reasons why API Analytics platforms like Moesif can make API developers' work easier, faster, and more efficient.
Reason #1: De-Mystify API Error Logs
"As a scientist, I'm all about data analytics. Finding, displaying, and sharing API metrics like 400/500 errors, when you have thousands of errors and millions of API calls, is very different", Rob Dejournett, CTO of pVerify
Searching through API event logs is time-intensive and single-value metrics often won’t explain complex error cases.
API analytics provide faster queries and deeper insights into a buggy API, supercharging your debugging process. For example, Moesif offers real-time insights into your logs that scales with any volume of API calls. Adding this layer to your API stack allows you to…
- Tail and filter HTTP requests in real time
- Inspect request and response HTTP payloads
- Inspect API logs and replay request in Postman or cURL in seconds
- Segment and aggregate volumes of API calls at scale, using a variety of different parameters
Knowing how your APIs are behaving is vital for both stability and a great user experience. Enhanced visibility into API calls makes log-searching a much simpler process.
Reason #2: Stay In-the-Know with Error Alerts
"As soon as the tool went up and we were able to see the types of transactions happening, we were able to identify security vulnerabilities within our infrastructure. We had transactions from sanctioned countries that were coming through where we weren't blocking certain countries. That was a win for the tool from the get go, literally day one."
-Marat Asadurian, Sr. Manager of Software Engineering, Trulioo
Infrastructure monitoring systems aren’t the best tool for uncovering anomalies in API usage. They’re more focused on things like tracking server uptime, latency, and more generic errors. Static and dynamic alerts take some of the guesswork out of error detection. Alerts can be delivered to many different channels, ensuring that technical teams have ample time to fix issues.
Static alerts can notify you when a static threshold has been reached. For example, you can create an alert that lets you know when users are facing authorization issues on your web app.
Alert if more than 20 HTTP ‘401 - Unauthorized’ status codes are returned in a REST API response within 15 minutes
Moesif’s dynamic alerts provide a broader view of novel errors. They automatically compare the error against your historical data, providing the context needed to discover complex errors. You can tailor our dynamic alert filters to the specific issues your product faces. If you are tracking key users, you might set up a dynamic alert like this:
Alert if a user’s API usage is abnormally decreasing over the last week
Dynamic alerts also provide an additional line of defense against users with malicious intent or an unintentional coding error. For instance, a user may be intentionally or unintentionally hammering an endpoint with traffic. Abnormally large amounts of traffic may overload the server and cause a DDoS outage. To prevent this, you could set up an alert for an abnormal increase in the number of queries against the endpoint. If an endpoint receives an abnormally large amount of requests, this filter alerts engineering and security teams immediately so they can be aware of the situation and be quick to take action.
Alert if an allowed user is creating an abnormally large amount of traffic against an endpoint, within the last 24 hours
When it comes to errors and threats, additional lines of defense can only enhance the reliability of your product. Alerts give your team the time it needs to react appropriately to anomalies in your API.
Reason #3: Streamline Collaboration with Non-Developers
"Moreover, a lot of the built in reports that you have are exactly the kind of product questions that an API company is going to want"
-James Messigner, Director of Dev Experience, ShipEngine
While product and developer teams are on the same side, ever changing requirements and timelines make it seem otherwise. Polished API analytics can help you communicate your ideas to non-developers and help to shape priorities within your product roadmap or bug backlog.
Most modern analytics platforms come with powerful data visualization features, and API analytics are no exception. Moesif offers secure, custom dashboards and reports designed to communicate key API insights to non-technical and technical stakeholders alike.
For example, let’s say you’re on a team building a mobile platform with in-app purchases. Your product team deprioritized a buggy feature that stores payment information for future use. Using visualizations like Moesif’s funnel report, you can justify why debugging takes priority. A funnel analysis is a common way for non-developers to model the journey of their ideal customer, from discovery of your product to purchase.
The funnel report breaks down where your app’s users dropped out of a sales funnel, over the last seven days. Your product team knows that your ideal customer makes at least 100 in-app purchases in that time. So, we they break the funnel into three stages, displayed left to right:
- Customers that made an account with your app
- Customers that performed at least one API call to a payment platform
- Customers that prompted at least 100 API calls to a payment platform
We see that customers who make their first purchase did so 6 hours and 5 mins after signing up, on average. We also see that only about half ever make a first purchase after signup. Using this instantly generated visualization, you can assert that fixing the buggy feature could reduce friction between signup and first in-app purchase.
When multiple teams join forces to offer a great product, it’s vital that all team members understand each other's needs. Good communication between developer, designer, product, and customer teams enables easy collaboration, and less interruptions for developers.
In Summary…
APIs are now the backbone of our interconnected world, but analytics best practices have not kept pace. Currently, developers subsidize much of the inefficiency that comes from an opaque API lifecycle. The solution is not as simple as writing analytics infrastructure in-house. To get the most out of your time as a developer, turn-key solutions are your best option. With an API analytics solution like Moesif, you can…
- Quickly and easily debug errors without tedious log-searching and fragile single-metric tests
- Uncover deeper insights about your API usage than you can with simple infrastructure monitoring
- Automate detection and notification of complex errors, malicious attacks, and latency issues
- Enable clear, precise communication about API usage and product health with non-technical collaborators
- Win back precious hours for deep focus while coding
Want to know exactly how Moesif can save engineering resources? Learn more here.!
This article was originally written for the Moesif blog by Savannah Whitman, marketing manager at Moesif.
Top comments (0)