Monitoring API Integrations: Stories from Fusebit and their Customers in the Field

Updated:
3 minute read

According to Blissfully’s Annual SaaS Report, companies between 250-500 employees on average use 123 SaaS applications to support their business, and will replace a staggering 39% of them over a two year period. The number goes up to 46% for smaller companies. To survive in this landscape, offering the best-in-class solution for a specific problem is becoming table stakes for SaaS vendors and is no longer enough to ensure continued success. Successful SaaS vendors provide standout integration and customization features in their platforms to increase conversion and retention.

Building robust and effective integrations is frequently challenging. At Fusebit, we hear from customers every day, as they struggle with reliability, isolation, security, and versioning, among other concerns. Monitoring, however, is the top pain point in almost every conversation. Development-time observability concerns crop up first, but are quickly replaced with the need to ensure reliability as production load ramps up and APIs evolve and break.

Here are some common pain points and failure modes in the words of our customers:

  • “I have no cohesive traceability of callouts to add-ons/webhooks and existing add-ons” ~ Senior Director of Software Development at German multinational

  • “Our story is bad, and partners are struggling with inspectability. We frequently hear: ‘I am not getting the events, tell me what you are sending to me’ or similar” ~ VP of Technical Product Management at same German company

  • “It’s hard to know when we have been called and developers don’t want to “miss” a webhook call. In [our case] it could make the difference between an order being paid for or not.” ~ Engineering Manager at Bay Area startup

  • “We have to defensively rate limit against customers and big players (Skype, Stripe) when providing an API.” ~ CTO at Bay Area startup

  • “Zendesk will turn off our webhook if it fails… a few minute outage on our side will cause deactivation of the webhook. We had to build a proactive monitoring and notification solution to deal with this” ~ Engineering Manager at Bay Area startup

  • “Because integrations are machine-to-machine it’s hard to know what’s going on. People flood us with traffic and forget to turn it off. Nobody sees the messages… our customers are bad at logging and monitoring.” ~ Co-founder and CEO of Bay Area startup

  • “Our integration got turned off recently because we didn’t respond in time. Sometimes it gets turned off don’t know why. We started monitoring our partner’s webhooks status page, we even built automation around it to get a little bit more visibility.” ~ Engineering Manager, Seattle startup

Against this backdrop of systematic under-investment in API integration monitoring on behalf of even the major SaaS players out there, companies seeking to build robust integrations are frequently left filling in the gaps with their own monitoring infrastructure. That’s where solutions like Moesif can save developers weeks of unnecessary upfront work, not to mention the ongoing peace of mind and reduction of operational pain.

Most monitoring solution providers focus on a customer’s own API surface, however in the integrations space, monitoring the integration partner’s infrastructure is just as important. Here Moesif’s offering stands out with their tailored feature set for monitoring third-party APIs:

Debugging and testing

  • Pinpoint the exact JSON or XML key that caused an API call to fail with live event stream.
  • Compare millions of successful and failing API calls and audit what changed

Analytics

  • understand the request volume going into integration partners, measure how poorly performing third-party APIs impact your own customer experience.

Reliability monitoring

  • Get alerted when partners are failing and ensure they’re supporting the issue
  • See the global health of which partners are having failures and how often
  • Correlate multiple failing transactions together to avoid alert storms

In building the integrations their customers are asking for, SaaS vendors frequently have a choice: whether to build integrations in-house, or leverage an integration platform (iPaaS) partner such as Fusebit. The former approach makes sense if the SaaS vendor will only develop a handful of integrations and plans on becoming an expert at maintaining them. The latter approach is a better fit if the number of integrations is expected to grow, and the vendor chooses to leverage the technical and operational know-how of an expert partner in the space. Moesif’s solution can be used in conjunction with either approach, while still delivering significant value.

So whether you’re building integrations yourself or working with a partner, don’t end up like the Fusebit customers quoted above; invest in first-class monitoring and analytics!

Leave a comment