How IT Entrepreneur R. Grows Instead of Shrinking
Many IT companies fear that artificial intelligence will accelerate development so much that fewer people will be needed. But the opposite is possible: those who cleverly rethink their offerings can actually grow more profitably and stably with AI.
Using the example of the fictional IT entrepreneur R. (200 developers), we show how this works – step by step.
1. The 2000 Projects Paradox
R.’s concern is understandable: if his teams deliver twice as fast thanks to AI, sales would theoretically have to acquire five times as many projects to keep everyone busy.
But this calculation is based on a fallacy – it applies old offering logic to a new world.
AI does not shorten development times to replace people, but to increase value creation per unit of time.
That means: productize, bundle, repeat – not always chase more.
2. From Project to Platform
• Focus instead of patchwork: choose 2–3 industries where you already have experience and develop standardized solution modules there.
• Deepening instead of broadening: industry knowledge beats feature variety.
• Reuse: each module should be 80% reusable.
• Examples:
- AI document processes for insurers
- Quality inspection with vision AI in industry
- Copilots for customer service in e-commerce
👉 This turns “project work” into a repeatable value stream.
3. Outcome Instead of Effort
The crucial shift: customers buy results, not hours.
Formulate your offerings as outcome sprints with clearly defined deliverables.
Example structure:
• Build Sprint (2 weeks) – working prototype with tests
• Hardening Sprint (2 weeks) – security, eval tests, cost monitoring
• Integration Sprint (2 weeks) – integration with customer systems
Each module becomes a product with a fixed price.
This way, efficiency becomes an advantage – not a risk.
4. Your Own AI Platform as a Growth Base
R. invests internally in a “Delivery OS” – a framework that accelerates all projects:
• Code templates & prompt library
• Eval harness for quality & hallucinations
• Token budget control & cost monitoring
• Security modules & PII masking
This is not “overhead,” but IP building.
Every improvement increases the reuse value of the next project.
5. Pricing Models That Grow With You
An AI-powered IT house needs pricing models that reward acceleration, not punish it:
• Outcome price per sprint
• Run-&-optimize subscriptions (monthly)
• Bonus for reaching KPIs (e.g., -30% process time)
• MSA with call-off catalog: negotiate once, then call off flexibly
• Channels & marketplaces: lower acquisition costs, predictable leads
6. From Sales to Pipeline
R. replaces classic project hunting with a subscription pipeline:
• Goal: 150–250 active accounts with ongoing sprints
• Tools: demo videos, vertical landing pages, benchmarks
• Partners: industry consultants and BPO players as feeders
• Metrics: lead velocity rate, sales cycle length, net revenue retention
👉 No longer find 2000 projects, but maintain 200 relationships.
7. New Ways of Working: Pods Instead of Chaos
Efficient delivery means constant teams:
• 5 developers + 1 PM + 1 QA/AE = 1 pod
• Pods pull work from the subscription backlog, not from random projects
• Eval gates check quality after each sprint
• Cost SLOs prevent uncontrolled token spending
Result: predictable utilization, stable quality, clear accountability.
8. Cultural Change: From Coder to AI Engineer
R. systematically builds a skill matrix:
Domain | Data | App | Ops | Trust.
• Further training: pair programming with AI, evaluation culture, prompting.
• Incentives: bonuses on customer KPIs instead of overtime.
• Innovation time: 10% per week for own tools or accelerators.
This turns fear of AI into new competence through AI.
9. The Numbers Example
For orientation:
One pod delivers 22 outcome sprints per year at 15,000 USD each = 330,000 USD
- Run subscriptions (5 × 3,000 × 12) = 180,000 USD
- Evolve add-ons (2 × 2,000 × 12) = 48,000 USD
= 558,000 USD per pod/year
With 40 pods (200 developers), that’s 22 million USD in revenue –
with predictable subscriptions, reusable IP, and higher margins.
10. 90-Day Plan for R.
Weeks 1–2: industry focus & inventory of existing solutions
Weeks 3–6: define outcome sprints, create platform modules
Weeks 7–10: onboard design partners, build references
Weeks 11–13: start go-to-market, activate channel partners
Result: first recurring revenues from month 4, scalable structure from month 6.
Conclusion: AI Accelerates – You Decide Where
R.’s initial concern is justified, but not inevitable.
AI can displace jobs – or strengthen companies that use it correctly.
The difference is not in the technology, but in the attitude:
Courage to productize. Courage to focus. Courage to grow.
Checklist for your company
☑ Define 2 industries
☑ Formulate 8 outcome sprints
☑ Activate eval harness
☑ Publish 3 demos and 2 references
☑ Onboard 2 design partners
☑ Secure 1 channel partner
☑ Plan pods & backlogs
☑ NRR ≥ 120%, target margin > 45%
Growing with AI does not mean running faster.
It means building smarter.