Self-service integration — including data preparation — increases agility and bridges information silos. Integration specialists, architects and IT managers should evaluate three complementary architectures to empower self-service: embedded integration, stand-alone integration and data preparation.
KEY FINDINGS
- The increases in data volume, velocity and variety — introduced by cloud, mobile, the Internet of Things (IoT) and big data — are forcing organizations to pursue self-service integration. Both “citizen integrators” and “ad hoc integrators” are embracing self-service integration. Citizen integrators include data scientists, power users and business users, whereas ad hoc integrators include line of business (LOB) developers.
- Self-service integration is possible through three complementary architectures: embedded integration (in data sources or in data consumers), stand-alone integration and data preparation. Each architecture includes several subtypes, which in turn come with their own strengths and weaknesses.
- Without the proper governance and architecture, self-service integration can cause many issues, such as inconsistent results, security violations, increased data sprawl, and brittle architecture caused by unmanageable point-to-point integrations.
RECOMMENDATIONS
- Advocate bimodal integration. Mode 1 focuses on trust, governance and reusability, whereas Mode 2 focuses on agility, speed and risk reduction. To take a step further, declare upfront which mode you intend to use for each integration project.
- Harvest the work of citizen and ad hoc integrators into enterprise assets for a shared data access layer. Shared data access provides trusted and reusable data services that comply with governance policies, enabling a dynamic digital marketplace.
- Choose integration architectures systematically based on use cases. Integration specialists, architects and IT managers should proactively engage citizen and ad hoc integrators and guide the selection of self-service integration architectures. Adopt assessment criteria, such as transactional context, target audience, integration scope and data quality.
- Use internal and external digital marketplaces to foster open exchanges of digital assets such as data, applications, algorithms and integration modules.
[For more information on self-service integration, please see Gartner’s research “Comparing Three Self-Service Integration Architectures“.]