Real Data from Six Months of Daily Use
From October 2025 to March 2026 I used Claude Code as my primary coding tool. Every day, on production projects — not toy demos, but systems that real users depend on.
This article isn't Anthropic marketing. It's the honest observations of a developer who went all-in on Claude Code and can tell you what works, what doesn't, and when it's worth (or not worth) reaching for it. The numbers I share are my own subjective estimates from personal time tracking — not an audited benchmark.
My Environment: Projects and Tech
Over these six months I worked with Claude Code on, among others:
- Stadomat — SaaS for cattle breeders (Next.js, PostgreSQL, multi-tenant architecture)
- bartoszgaca.pl — portfolio and blog (Next.js)
- Client MVPs — marketplace, AI tool, admin panel
- MCP Servers — integrations with LinkedIn, GSC, GA4, invoicing
- Automations — n8n workflows, custom scripts, monitoring
Tech: mostly TypeScript/Next.js, with some Python and PHP/Laravel.
Productivity: What Actually Changed
What I Measure
I track work time in 15-minute blocks and compare it against how long similar tasks used to take me. That comparison comes from memory and notes, so treat the figures below as rough, not as a precise measurement.
Results
New project scaffolding:
- Before Claude Code: 4-6 hours (setup, config, starter code)
- With Claude Code: 30-60 minutes
- Rough speedup: 5-8x
CRUD and API endpoints:
- Before: 2-3 hours per entity (model, API, validation, tests)
- With Claude Code: 15-30 minutes per entity
- Rough speedup: 4-6x
Refactoring existing code:
- Before: depends on scale, but typically 4-8 hours for a large refactor
- With Claude Code: 1-2 hours (Claude "sees" the full file context)
- Rough speedup: 3-4x
Writing tests:
- Before: 1-2 hours for a module's test suite
- With Claude Code: 10-20 minutes (it generates, I review and fix)
- Rough speedup: 4-6x
Debugging complex bugs:
- Before: 1-4 hours
- With Claude Code: 20-60 minutes (Claude analyzes the stack trace and code context)
- Rough speedup: 2-3x
Average speedup across all tasks: roughly 3x — with the caveat that this is my subjective estimate, not the result of a controlled experiment.
What Works Very Well
1. Boilerplate and Scaffolding
Claude Code is great at repetitive code. New project setup, ESLint/Prettier/Tailwind config, building UI components from a design system — minutes instead of hours. It usually needs only minor tweaks.
2. Format Conversion
Migrating from Pages Router to App Router in Next.js. Converting JavaScript to TypeScript. Switching CSS Modules to Tailwind. Claude Code handles this smoothly and, in most cases, without errors.
3. Tests
Generating unit and integration tests is an area where AI genuinely helps. Give Claude a component or function and you get a complete test suite covering the happy path, edge cases, and error scenarios. The quality is often better than what many developers write by hand.
4. Documentation and Comments
JSDoc, README, inline comments — Claude Code generates documentation that is actually useful, not generic. It understands the code context and explains "why," not just "what."
5. Multi-file Refactoring
This is one of Claude Code's biggest strengths versus the competition (Copilot, Cursor). It operates at the level of the whole repository — it understands dependencies between files and knows that a change in one file requires changes in another. Refactoring twenty files at once is no problem.
What Does NOT Work (Or Requires Caution)
1. Architectural Decisions
Claude Code is weak at strategic architecture decisions. "Microservices or monolith?", "How do I design the permissions system?" — this still takes human experience. Claude will suggest a solution, but not necessarily the one that's optimal for your business context.
2. Complex Business Logic
Simple CRUD? Perfect. A complex workflow with a dozen-plus conditions, multiple actors, and edge cases? Claude generates code that looks good but can carry subtle bugs. For complex logic I always review the code by hand.
3. Performance Optimization
Claude Code will write working code, but not always optimal code. N+1 queries, unnecessary React re-renders, missing memoization — these still need a human eye. Use profilers and analyze the output.
4. Security-Critical Code
Auth flows, encryption, access control — Claude Code often generates correct code, but not always. For code handling login, payments, and personal data I always have a human do a security review.
5. API Hallucinations
Claude Code sometimes "invents" API methods or parameters that don't exist — especially for less popular libraries. Always verify against the documentation. Tip: include the library version in your prompt — in my experience it noticeably reduces hallucinations.
Claude Code, Copilot, and Cursor — An Honest Comparison
GitHub Copilot
Pros: best inline autocomplete. Fast. Well-integrated with VS Code.
Cons: no context beyond the current file. Doesn't understand project architecture. Generates line by line, not whole solutions.
Best for: code completion while typing.
Cursor
Pros: a good AI-first editor. Composer mode is powerful. Fast.
Cons: smaller context than Claude Code. Weaker at multi-file operations.
Best for: developers who prefer a GUI IDE with built-in AI.
Claude Code
Pros: large context (up to 1M tokens depending on model and plan — check current limits). Best project understanding. Multi-file refactoring. Terminal-native, so it integrates with git, npm, and docker.
Cons: terminal interface (not for everyone). Slower than Copilot for inline completion. More expensive.
Best for: experienced developers working on large projects. Building MVPs. Refactoring.
Cost Analysis
Claude Code (Max plan): roughly $100-200/month depending on usage
GitHub Copilot: roughly $19/month (Business) or $39/month (Enterprise)
Cursor: roughly $20/month (Pro)
Prices are approximate — check each tool's current pricing before deciding.
Is $200/month expensive? It depends on perspective:
- Say you save a few dozen hours a month thanks to faster work
- Apply your own hourly rate
- Even conservative estimates put the savings well above the subscription
- This is an illustrative example, not a measured benchmark — your numbers will depend on the kind of projects and your rate
Even if the real speedup is smaller, the cost of the tool pays for itself quickly. For me, Claude Code is one of the best productivity investments I've made.
My Workflow: How I Use Claude Code Daily
- Planning (me): define the architecture, database schema, and API contract
- Scaffolding (Claude Code): generate project structure, models, basic CRUD
- Business logic (me + Claude Code): I design the flow, Claude implements, I review
- Tests (Claude Code): generate tests, I run them and fix edge cases
- Refactoring (Claude Code): "Refactor this module, extract X into a separate service"
- Deploy (me): CI/CD review, production deployment
Key principle: I am the architect and reviewer, Claude Code is the implementer. Never the other way around.
FAQ — Frequently Asked Questions
Is Claude Code suitable for team work?
Yes — each developer can have their own account. Claude Code respects .gitignore, your branching model, and code style. The key is for the team to agree on shared conventions for prompts and a review process for AI-generated code.
Is Claude Code-generated code good quality?
For boilerplate and standard patterns: very good. For complex logic: it needs review. My rule: I treat Claude Code output like a pull request from a junior developer — always review before merge.
How much does Claude Code cost per month?
The Max plan runs roughly $100-200/month depending on usage (check current pricing). With a noticeable speedup in your work the cost pays off quickly — it's one of the more sensible tools you can invest in.
Will Claude Code replace programmers?
No. It may, however, displace the ones who don't use AI. A programmer working with Claude Code is clearly faster than without such a tool. It's a tool, not a replacement — architecture, code review, and business decisions remain the human domain.
Is it safe to give Claude Code access to a company repo?
Claude Code runs locally — the code doesn't leave your machine beyond the context that's sent to Anthropic's API. Per Anthropic's policy, API data is not used to train models by default, but it's worth verifying the current terms before rolling this out in a company. For sensitive projects, consider extra safeguards or consult on security architecture.
Summary: Is It Worth It?
After six months of heavy use: for me, unequivocally yes. Claude Code genuinely changed how I build software — faster, with better baseline code quality and less repetitive work. I can't imagine going back to coding without AI.
Have a project idea? Let's talk for 20 minutes — I build MVPs and products using Claude Code, which usually means faster and cheaper than traditional development. No obligations.
Need Business Automation?
If you want to implement process automation in your company — you don't have to do it alone. Check out AI automation for business or book a free 20-min call — I'll tell you where to start and what results to expect.