Startups and enterprises have different needs; architecture matches.
Startups:
- Fast-moving — prioritise speed over elegance.
- Small team — 1-3 admins / devs typically.
- Limited budget — cost-conscious.
- Evolving requirements — pivot often.
- Flexibility paramount — minimise commitments.
Architectural choices for startups:
- Out-of-the-box first — don't customise unless necessary.
- Lower-tier licenses — Essentials / Professional / Platform.
- Minimal sandbox tier — Developer sandboxes.
- Simple DevOps — Change Sets or basic CI/CD.
- Simple integrations — Zapier, native connectors.
- Avoid managed packages unless clear need.
- Defer architecture decisions that lock in.
Anti-patterns for startups:
- Enterprise-grade governance overhead — slows velocity.
- Multi-org — overkill.
- Heavy custom code — maintenance burden.
- Big-bang implementations — slower than iterative.
Enterprises:
- Slow-moving — prioritise quality, governance.
- Large team — many admins, devs, architects.
- Adequate budget — but justified.
- Stable requirements — within roadmap.
- Compliance / audit mandatory.
Architectural choices for enterprises:
- Comprehensive governance — CoE, ARB, standards.
- Multi-sandbox tiers — Full, Partial, Dev.
- Sophisticated DevOps — DX, CI/CD platforms.
- Multi-cloud strategy — Sales + Service + Marketing.
- Industry Clouds where vertical-aligned.
- Mulesoft / iPaaS integration backbone.
- Shield for compliance.
- Multi-region / multi-org as needed.
Anti-patterns for enterprises:
- Quick-and-dirty — produces unmaintainable orgs.
- No governance — chaos at scale.
- Single-vendor lock-in — risky for critical systems.
Mid-market:
Between startup and enterprise. Smaller-scale enterprise patterns work; modest governance, formalised but not bureaucratic.
Architectural maturity stages:
- Stage 1: out-of-the-box config.
- Stage 2: customisations within best practices.
- Stage 3: governance and standards.
- Stage 4: Center of Excellence, multi-cloud.
- Stage 5: enterprise-grade with compliance.
Most companies progress 1 -> 5 over years. Architecture should match current stage.
Common pitfalls:
- Startup with enterprise architecture — overhead exceeds value.
- Enterprise with startup architecture — chaos.
- Mid-stage with stage-mismatch tooling — friction.
Senior architect insight: architecture should match company stage, not aspiration. A startup pretending to be enterprise wastes time. An enterprise pretending to be startup creates incidents.
The senior framing: right-size architecture for current need plus 1-2 years of growth. Don't over-engineer for hypothetical scale.
As the company grows, architecture evolves. Plan for evolution, but don't pre-engineer for it.