Embrace AI now to future-proof your franchise operations

Image supplied

That was the message from Loren Phillips, principal vizier of AI Strategy Africa at a recent franchiser mentorship breakfast hosted by Nedbank Retail Services and Franchise Coaches.

In the session, Phillips focused on moving past the hype and uncertainty to understand where and how AI can deliver real, measurable value in franchise operations.

Beyond the buzzwords

Cutting through the jargon to focus on specifics, Phillips reminded the audience that generative AI is only one strand of a wider discipline.

"AI is a much bigger story than ChatGPT. It's about machine learning, predictive analytics, computer vision … Generative AI is just one part of it."

She added that large language models (LLMs) – developed using vast volumes of textual information and powering tools like ChatGPT – come with constraints.

"AI matches patterns rather than understanding them, and when the data has gaps, it fills in the blanks with the closest match, which may or may not be accurate."

However, Phillips says that one of AI's greatest strengths is handling imperfect data.

"AI handles messy data really well, and we can help it by building an architecture around our data to help the system better understand it."

While AI won't necessarily clean data, structuring the data makes it more useful and accessible, allowing franchisers to spot compliance issues, improve forecasting, or detect customer behaviour patterns earlier.

Readiness versus reality

"In South Africa, while 70% of businesses feel ready for AI, only 14% have an actual strategy. We're ready to use it, but we just don't know where to start," said Phillips.

Her recommendation: define organisational values before AI implementation.

"Decide what your business values are now, because that's likely to face further evaluation later."

She also flagged data sovereignty – knowing who owns your data, where it's stored, and the laws that apply – and reminded franchisers that 'whatever the technology is doing, the hackers are doing too.'

Phillips warned that every transfer of data between systems increases the surface area for risk. This is particularly critical when handling sensitive information such as ID, details of minors, or health records, where a breach would not only damage trust but also contravene the Protection of Personal Information Act.

Consent and control over data are, therefore, essential. She advised businesses using ChatGPT to switch off the setting that allows information to feed the model.

She suggested enterprise tools like Microsoft Copilot or Google Gemini for Business for added protection, although users must still verify settings themselves.

Phillips added that some companies are now running AI entirely on their own servers, using models trained on only their internal data, from policies and manuals to brand guidelines and customer information.

This keeps sensitive data in house while allowing the AI to deliver responses and recommendations specific to the business and its values.

In sectors with strict privacy rules, such as financial services or healthcare, keeping data local can be essential. Lower costs and simpler tools now make this possible for many more organisations.

Finding the right use cases

Applications vary by sector, for example:

• in fast-moving consumer goods – demand forecasting, shelf stocking optimisation, and AI driven promotions;
  
• in beauty and health – predicting appointment no shows, generating standard operating procedures, and creating localised product education; and
  
• in home and DIY – predictive maintenance and product recommendations.

The common thread is using AI to turn operational data into insights that protect brand standards, enhance customer experience, and support growth.

Phillips suggested the following five guiding questions to match the right AI capability to the right problem:

1. Where are you making repeatable decisions with historical data?
   
2. Where are you losing time, trust, or money through inconsistencies or manual efforts?
   
3. Where is useful data already being captured, even informally?
   
4. Where would faster feedback improve performance?
   
5. Where do you need to predict outcomes based on previous events?
   

Small operators' advantage

For smaller franchise groups concerned about larger competitors, Phillips offered reassurance.

"Your advantage as a small operator is that you can be more resourceful. You're faster. You have tighter control of your data. There are so many businesses moving exceptionally slowly because of bureaucracy."

But she cautioned against giving AI the final say in customer facing decisions.

"Always keep a human in the loop when it comes to critical decision making. You can't just say, "Sorry, AI said so." It's not going to be good enough."

Karen Keylock, national retail manager for Retail Services at Nedbank Commercial Banking, explained the bank's goal to help franchisers support growth of both their brands and their franchisees.

"Nedbank fully appreciates the importance of AI in today's operating environment and, as always, we strive to bring our clients the best tools, advice, and partnerships to help them stay ahead. Through our collaboration with Franchise Coaches, we create opportunities for franchisers to learn from leaders like Loren and translate those lessons into practical strategies for growth."

Keylock added that Nedbank's support goes beyond finance.

"The beauty of banking with Nedbank is that you have a partner that is strategically aligned to your growth narrative, supports you with your financial needs and business aspirations, and helps your business unlock the next level of growth."

A mindset shift

Phillips closed by urging franchisers to see AI as a tool for amplifying – not replacing – human leadership. On its own AI technology is not enough; its success depends on how people adapt to it.

Phillips highlighted the role of change management in reassuring staff that AI is not intended to replace them while preparing them for the likelihood that their roles will evolve.

Getting people on board and building confidence, she said, are essential to embedding AI in a way that benefits both the business and its people.

Her message was clear: start small, be strategic, and take your people with you. Those who act now will set the tone.