The Franchisor’s Algorithm Ordered the Stock. Who Pays When It Is Wrong?
Head-office AI now makes daily trading decisions for independent store owners, and the losses still land on the franchisee.
Sign up for my Substack daily AI newsletter here.
Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/
Picture a franchisee on a Monday morning, staring at a cold room full of stock she did not decide to order. The forecast came from head office, the order was placed automatically, and the perishables will be written off against her overdraft, not the franchisor’s. South Africa’s franchise sector counts more than sixty-eight thousand franchisees across seven hundred and twenty-seven systems and contributes about one trillion rand to the economy. Absa’s franchising specialist names predictive demand forecasting among the AI tools owners must now adopt to stay competitive. The technology may well improve trading. The question nobody has settled is who carries the loss when it gets the trade wrong.
CONTEXT AND BACKGROUND
Franchising has always been a bargain. The franchisor supplies the brand and the system, and the franchisee supplies capital, local knowledge and personal risk. AI changes the terms of that bargain without changing the paperwork. Greenberg Traurig reports that franchisors are adopting AI for inventory and staffing forecasts and for monitoring individual units. It also reports that regulators and litigants in the United States are challenging technology fees and mandatory vendors as hidden margins and unfair practices.
The risk is no longer theoretical. The first major test case has arrived: Chaac Pizza Northeast, which operates about one hundred and eleven Pizza Hut outlets in the United States, is suing the franchisor for more than one hundred million dollars. It alleges that a mandated AI kitchen and delivery system caused delays and stripped its managers of operational control, and that Pizza Hut would not let it step back from the system. The allegations are unproven.
INSIGHT AND ANALYSIS
The pattern matters more than the outcome of any single case. When head office decides what a store orders, prices or schedules, the decision moves to the franchisor while the consequences stay with the franchisee. Most franchise agreements were written when the store owner placed the order and therefore owned the mistake. Few contemplate a central model trading on behalf of an independent business owner.
Responsibility also becomes harder to trace. Risk Management Magazine notes that when automated systems act across organisations, liability disperses among the model provider, the system provider and the business deploying it. No single party sees the full decision chain, and conventional insurance may not respond. In a franchise, that chain runs from the software vendor, through the franchisor, to the store.
I have previously written about this in the context of AI ordering agents. I argued that autonomous systems can create order volatility of their own even when consumer demand is flat, and that businesses should cap automated order changes and require human approval above set thresholds. In a franchise, those controls belong in the agreement, not only in the software settings.
IMPLICATIONS
South African franchisees have more leverage than they may realise. The Consumer Protection Act requires a disclosure document at least fourteen days before signing and allows a ten-business-day cooling-off period. It also restricts unreasonable fees and conduct that go beyond the franchisor’s legitimate business interests. Before signing or renewing, franchisees should negotiate three terms: transparency about how the system decides, the right to override it with local knowledge, and loss-sharing when a mandated system demonstrably causes harm.
There is a customer dimension as well. When Consumer Reports tested Instacart’s AI pricing, nearly three-quarters of items showed different prices for different shoppers at the same moment. Instacart ended the programme, conceding that its testing had missed the mark for some customers. If head office’s algorithm sets prices by store or by hour, it is the franchisee who faces the customer at the counter.
CLOSING TAKEAWAY
AI can make franchise systems sharper, and franchisors are right to invest in it. The bargain must keep pace with the technology. At its centre lies an asymmetry: head office sees how a system performs across hundreds of stores, while the franchisee sees one cold room, one overdraft and one set of regular customers. A model can look efficient in aggregate while failing badly in a single store. Whoever makes the trading decision should share the cost when it is wrong, and whoever carries the cost should be able to see and challenge the decision. Accountability cannot be franchised away. The time to insist on that is before the next signature, not after the next write-off.
Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net
Comments