Strategic DDD and Bounded Contexts

By Michael Samuel Naeem · August 15, 2025 · 5 min read

Every team called it "the account" but Billing meant ledger identity, CRM meant sales prospect, and Auth meant login credential—one database table, three incompatible mental models. Strategic Domain-Driven Design does not start with entities and repositories; it starts with linguistics and boundaries. Bounded contexts delimit where terms mean one thing; context maps show how contexts relate without pretending one unified enterprise model exists.

Ubiquitous language

Product, engineering, and experts share vocabulary in workshops:

• Policy (Insurance): contract terms, not IAM policy
• Quote (Sales): priced offer, not string literal

Code names match speech: IssuePolicyCommand, not ProcessDataRequest. Glossary wiki maintained by product.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Subdomain classification

Core:        underwriting rules engine
Supporting:  agent commission calculation
Generic:     payment capture (Stripe)

Invest modeling depth proportional to classification. Do not build custom payment kernel.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Drawing bounded contexts

Start messy on whiteboard— refine to modules:

[Quoting] --ACL--> [Legacy Rating Engine]
[Policy Admin] --events--> [Billing]
[Claims] customer-supplier [Policy Admin]

Relationships dictate integration style from prior post on ACL.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Team alignment

Conway's law applies: one team per context where possible. Two teams one context breeds model schism. Temporary shared ownership needs explicit governance and merge cadence for glossary.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Evolution from monolith

Modular monolith enforces package-private boundaries between contexts before network boundaries. @Deprecated crossContext lint for illegal imports.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Anti-pattern: unified canonical model

Enterprise data model committees shipping XSD for all divisions move pain to integration without solving semantic mismatch. Prefer explicit translation at edges.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Small product with one clear domain—over-partitioning contexts adds meetings. Revisit when onboarding confusion or cross-team bugs spike.

Context map updated when partnerships or org structure changes. Shared kernel stays minimal—generic IDs and money, not dumping ground for shared business rules.

Conway's law: two teams one context breeds model schism. Event storming before drawing microservice boundaries.

Small single-domain products skip over-partitioning—revisit when onboarding confusion spikes.

Validate this in staging with production-like data volume before declaring done. Capture metrics baseline the week before change and compare for seven days after—subtle regressions hide in aggregates until a large tenant hits the path. Update the on-call runbook with the failure signature and rollback command so responders need not rediscover steps during an incident.

Document the decision, owner, and rollback path in your team wiki the same week you ship. Future you will not remember which environment variable toggled the behavior unless it is written next to the runbook entry and linked from the alert. That habit costs ten minutes per change and saves hours when pagination or auth misbehaves under a single large tenant.

Run the change through your standard PR checklist: tests, observability, and a two-minute rollback drill in staging. Small operational habits accumulate into systems that survive on-call nights without heroics.

Share a short write-up in your engineering channel after rollout: what shipped, what metric you watch, and who owns follow-up. That closes the loop for teammates who were not in the PR and surfaces gaps in docs before the next person repeats the same investigation.

Prefer boring, repeatable process over one heroic migration weekend.

Treat operational readiness as part of definition-of-done: dashboards, alerts, runbook links, and a named owner. Skipping those steps ships code that works in demo and fails quietly in production until a customer or auditor finds the gap.

Resources

• Domain-Driven Design (Eric Evans)
• Bounded Context (Martin Fowler)
• Context Map (DDD community)
• Team Topologies (Skelton & Pais)
• Domain-Driven Design Distilled (Vaughn Vernon)