How much does an integration cost?

One of the most common questions we're asked by utilities and partners is "how much does the integration cost?" As consultants, we would typically answer "it depends." However, as one of the leading providers of an Enterprise iPaaS, we are able to use our experience from a variety of integration projects to offer some indicative numbers.

There are several cost drivers that should be considered when estimating an integration project as well as the Total Cost of Ownership (TCO) of an integration solution:

• Platform costs - mostly recurring (OPEX):
  • Licenses or subscription costs for the integration platform and all environments (including test, stage, production environments), its modules and, when specified and required, third-party components for various environments
  • For on-premise / data center deployments cost for hardware, server infrastructure, virtualization services, operative systems, backup systems need to be considered
• Infrastructure costs - recurring (OPEX):
  • For public or private cloud deployments the cost of utilized cloud services, the compute power to be used, data ingress and egress, data exchange in a hybrid cloud environment are calculated
  • For on-premise / data center deployments the server and infrastructure operations, network operations, and data center operations are taken into account
• Integration Project costs - one-time investment (CAPEX):
  • Implementation including installation, provisioning, and configuration of the platform services. Also, the costs of delivery related to enabling connectivity to applications, systems and services and their following integrations. The, setup of CICD pipelines and DevOps tooling support also need to be considered.
  • Design of the integration architecture, security architecture, integration analysis, interface specification, API design
  • Design, development, and module-testing of defined integrations
  • Test management, test case specification, test data (anonymized) preparation, test execution
  • Project administration, project management, migration & cut over management, documentation, end user training
• Integration operational costs - recurring (OPEX):
  • Platform operational costs including monitoring, management, and continuous improvement of platform components and services, and cloud optimization costs
  • Integration management costs for monitoring, application management, maintenance, bug fixing, and continuous optimization of delivered / deployed integrations
  • Business operational costs that include business activity monitoring, manual handling of business process deviations, manual handling of data quality problems, management of business processes and value chains

When estimating the TCO of your integration solution here are some tips based on our extensive experience with integration projects:

• Look carefully at the holistic integration architecture to identify unnecessary "jumps" or data exchange between various cloud providers (or private cloud and on-premise applications). This will help you to avoid exploding cross-cloud data transfer costs (data egress). It is not uncommon for companies to get a "surprise" when they see how costly their cloud operations become with a badly designed integration architecture.
• Establish a unified logging & tracing module from day one. A crosscutting big data log analytics solution that monitors both platform operations and all the integrations deployed will be a huge advantage during development and testing and also during operations.
• Everybody loves DevOps. However, very few like and are good at the "ops"-part. Think through, plan for, and build a team that will operate, manage, maintain, and optimize the platform, as well as develop and deliver the integrations.
• Set up CICD pipelines and DevOps tooling support from the beginning of your integration project to ensure efficiencies in delivering integration solutions. In a fast-paced, digital world, you must be able to deliver high quality, new integrations, quickly and in an extremely agile way.
• Design and deliver solutions that support domain experts and business users in monitoring and managing the integrations from a business/ data point of view. Provide them with the tools that reduce the need for manual activities to rectify data discrepancies. However, even with the best and most modern integration solution in place, process failures or data quality problems can occur that need manual intervention.
• Place the creation and preparation of (anonymized) test data on the critical path of your project. Many organizations underestimate the effort needed to create and prepare test data for both the desired "happy path" but also all the test data for deviations and exception handling.