Integrations are nonnegotiable for SaaS companies. The average business’s SaaS portfolio encompasses 342 apps. Without integrations, these apps become data silos, and we all know the challenges with those.
Customers expect seamless connectivity. According to G2, B2B software buyers consider integration capabilities a top factor in their decisions. Another survey found more than half of respondents cited poor integration as a reason to explore new software vendors.
Every SaaS team knows they need integrations to connect their products to the other products their customers use, but that doesn’t make building them easy. Early on, many dev teams estimate the engineering time based on the complexity of the integration’s code. They may look at a request and think, “Writing this code should be straightforward; I can knock it out quickly.” Several missed deadlines later, they find themselves spending more time on integrations than on product innovation, and frustration mounts for both dev teams and customers.
It’s typically not coding an integration that’s hard — it’s all the “other stuff.” Teams often forget to consider the complex infrastructure and tooling required to run integrations, let alone deploy them to numerous customers and maintain them at scale. These supporting components consume the majority of integration engineering time. Overlooking them leads to unexpected delays as developers encounter more work and obstacles than anticipated.
Common stumbling blocks
Integrations require a robust, scalable, secure environment to ensure smooth operation in production. The first consideration is provisioning compute resources to execute the code. Then, you must decide how to test, deploy and update your code. Any integration should include detailed logging, performance monitoring, error detection and alerting.
Security is a primary concern for integrations. To safeguard customers’ sensitive data, you need to implement robust protection measures, including secure OAuth 2.0 flows, data encryption and access controls. These systems must comply with security and privacy regulations, like SOC2, GDPR and HIPAA, and require vulnerability monitoring and updates.
All of these steps and decisions are required just to get an integration into production with a single customer. Complexity scales quickly when you consider that most integrations need to serve many customers who need to connect to their own accounts and often have different configuration needs.
Because of that, integrations need to be built in a configurable way so that devs aren’t called in to write custom code every time a new customer needs an integration to behave differently.
Most teams ultimately need to build a user-friendly interface so that customers or customer success teams can easily set up the connections and configurations.
Scaling infrastructure to support hundreds or thousands of active integrations can be a significant time suck. Another challenge is that you usually extend an integration’s capabilities over time or make changes to keep up with third-party APIs. Rolling out changes across your customer base requires more complex versioning and deployment capabilities than many teams anticipate.
Managing technical debt is a critical consideration. Rapid development to meet customer needs can lead to solutions that are difficult to maintain and costly to rebuild.
Figuring out how to handle more integrations without proportionally increasing the size of your team presents many challenges. You must balance the need to support integration with maintaining focus on the core app.
Establishing a realistic timeline
A realistic development timeline demands a thorough discovery phase to understand the specific integration requirements. Knowing what an integration is supposed to accomplish is just the tip of the iceberg. Not looking deeper sets you up for many surprises.
Once you understand an integration’s purpose, you can narrow the scope to determine details like trigger events, system authorization methods and APIs, and data transfer specifications. Teams should talk to end users, seek out corner cases and engage with tech partners.
After establishing specific integration requirements, evaluate necessary compute resources, scalability needs, deployment strategies, monitoring and logging requirements, configuration management and UI elements.
Break the project into manageable milestones, then determine the resources required for each. This roadmap will allow you to account for any dependencies or potential roadblocks. Remember to factor in the time developers spend managing and maintaining integrations once they’ve been shipped.
Strategies to manage integrations
SaaS companies are more than capable of building and maintaining integrations in-house. The question is: Is that the best use of resources?
For companies with very large engineering teams, the answer may be yes. Most other companies may be better off seeking support to accelerate integration development and relieve the burden on their dev teams. This strategy might involve enlisting a third party to actually build the integrations.
Other tools to consider include unified APIs and embedded iPaaS. These solutions enable you to develop customer-facing integrations while minimizing time spent on the underlying infrastructure.
Without a scalable strategy, integrations can destroy your product roadmap. Even the most well-planned timeline will take many months if teams choose to build everything from scratch. As integrations become more numerous and complex, most teams don’t have the resources necessary to keep up. When considering your integration strategy, don’t forget to account for all the “other stuff,” like infrastructure and tooling. Investing in scalable systems and processes is essential to ensure long-term product success and sustainable business growth.