What Artificial Intelligence Has Taken, and What It Has Not

Jump to section

• Overview & contents
• i. Important Notice and Basis of Preparation
• 1. What Artificial Intelligence Has Taken, and What It Has Not
• 2. Executive Summary
• 3. Service Lines, Clients and Pricing
• 4. Capacity: The Shooting Day and the Editing Day
• 5. SWOT and Competitive Position
• 6. Organisation, Rights and Compliance
• 7. Financial Plan
• 8. Break-Even and Debt Service
• 9. Investment Analysis
• 10. Sensitivity and Scenario Analysis
• 11. Risk Analysis
• 12. Implementation Roadmap
• 13. Key Performance Indicators
• 14. Key Assumptions
• 15. Conclusion and Recommendation
• A. Appendix A: Consolidated Financial Summary
• B. Appendix B: Capacity, Mix and Cost Schedules
• C. Appendix C: Funding, Debt and Working Capital Schedules
• D. Appendix D: Risk Register
• E. Appendix E: Glossary

1.1 The proposition

Frame & Field Studios is a commercial photography and content studio in Gauteng, building to 7.5 photographers and 4.5 editors over five years. It sells corporate and executive portraiture, property and architectural photography, event and documentary coverage, brand and editorial campaigns, and motion content — and it deliberately does not chase catalogue product photography.

At maturity the studio bills 1 195 days a year at an average of R23 407, generating R27 959 254 of revenue and R5 338 585 of EBITDA.

1.2 The question any photography plan must answer in 2026

Generative artificial intelligence has changed the economics of commercial imagery, and it has done so within about two years. Traditional product photography is reported at US$85 to US$250 per stock keeping unit against US$3 to US$12 an image for AI alternatives, falling to a few cents at subscription scale. Around 67 per cent of leading e-commerce operators now budget specifically for AI imaging tools, and 83 per cent of creative professionals report using generative AI in their workflow.

1.3 Where photography is still defensible

The work that survives has a common property: the image must depict something that actually exists and actually happened. A generated portrait is not a photograph of your chief executive. A generated interior is not the flat a buyer is being asked to purchase. A generated conference photograph shows an event that did not occur. In each case the value of the image lies precisely in its being a record.

On this plan’s mix the weighted probability of AI substitution is 21.8 per cent of shooting days. Because the exposed lines are not the expensive ones, the revenue-weighted figure is higher at 25.7 per cent, putting R7 185 528 of mature revenue at risk. On a conventional commercial studio mix weighted toward catalogue and product work, the same revenue carries 45.3 per cent revenue-weighted exposure and R12 665 542 at risk.

1.4 Financial summary

1.5 The honest assessment

Seven findings matter more than anything else in this document.

▪ Utilisation is the entire business, and the plan assumes a demanding number. The model assumes photographers bill 71 per cent of working days at maturity. Break-even is 39.9 per cent — 695 billable days, or three fully utilised photographers of the seven and a half on the payroll. Between those two numbers lies the whole return, and 71 per cent is achievable only with a genuine producer function and a booked forward diary.

▪ Editing, not shooting, caps the output. Every shooting day generates about 0.87 days of post-production, and motion work generates 2.4. Editing capacity binds in every year of the plan, not merely from Year 3. Adding a photographer without adding editing capacity produces no additional revenue at all.

▪ Each capacity lever is capped by the other, which is not obvious. Improving the editing ratio by 15 per cent gains only R515 235 because shooting then becomes the constraint, while allowing it to worsen by 15 per cent costs R2 311 197. Raising utilisation eight points gains nothing at all because editing binds, while losing eight points costs R1 530 984. Moved together the two are worth R2 576 175 — more than the sum of the parts. Both levers protect the downside far more than they create upside, and neither should be pursued alone.

▪ Day rates are the most sensitive variable and they are under pressure. A ten per cent movement in day rates swings mature EBITDA by R5 617 680. The plan already assumes rates on AI-exposed lines erode faster than on protected ones, giving a blended increase of only 2.8 per cent a year against costs escalating at 6.0. If clients begin benchmarking all photography against AI pricing rather than only catalogue work, that erosion accelerates.

▪ Year 1 loses money and the covenant is not met until Year 3. EBITDA is negative R602 241 in Year 1 with debt service cover of negative 0.35 times, reaching 2.21 times in Year 3. A studio must build a portfolio and a client list before it can bill, and that takes a year during which the overhead is already running.

▪ This is a people business, and the assets go home every evening. Payroll is 69.2 per cent of the cash fixed cost base. The capital equipment matters far less than the team, which is why the exit multiple assumed here is 4.5 times rather than the higher multiples an asset-backed business would command.

▪ Working capital is heavier than it looks for a services business. Corporate and agency clients pay in about 58 days, and work in progress adds a further 12. Working capital reaches R4 566 135 at maturity — 16.3 per cent of revenue — and exceeds the R3 600 000 facility from Year 4, with the balance funded from retained earnings.