PMI builds commerce engine to glean customer insights

Counterfeit tobacco sales account for as much as 75% of South Africa’s total market. And while Mary Mahuma, CIO for Southern Africa PMI, admits that the challenge facing the business is significant, she finds solutions by tackling the root cause of the issue: customer insights. According to her, other FMCG brands also struggle to clearly understand consumer behavior, how they engage with brands, and what they actually want. This is especially true in rural and informal markets.

“One might expect a brand like PMI to try to address these challenges by focusing on big fish,” she says. “But there’s so much value in better targeting our strategies toward understanding the hidden market for tobacco products.”

This market, consisting of small, independent convenience or general trade stores, is often hard to reach. “But this doesn’t mean the customers in these communities are any less important,” she says. In fact, they represent an untapped market because purchasing behavior is shaped by affordability and availability, rather than traditional brand loyalty. “By understanding the informal economy, where demand exists but is largely being met by illicit supply, PMI can close commercial gaps and access data that more accurately reflects the realities of consumption in South Africa,” she adds.

Follow the product

PMI has a trade engagement platform for large retailers and wholesalers, which the business uses to enhance loyalty, provide training, share data, and drive higher sales performance. But this tool didn’t effectively reach those general trade stores.

So to better engage this market, PMI had to deploy a solution tailored to their needs. “Feature phones are commonly used by people living in rural settings, and digital literacy levels are low, so we had to make sure that whatever we built wasn’t too complex,” she says. This omnichannel capability leverages unstructured supplementary service data, allowing a user to access a service by typing a short code without an internet connection, and helping address cost and connectivity barriers.

The goal was to streamline ordering and delivery processes, unlock valuable customer insights, and ultimately drive better decision-making. “You often hear people say follow the money, but in this case, PMI needed to follow the product,” says Mahuma. “When we did this, we realized there was a big gap in our understanding of who was buying our products and where our products were ending up.”

The omnichannel platform makes it possible for shop owners to easily place an order for additional stock, for example. Understanding customer habits means ensuring the platform is available in various languages, since many informal traders in rural communities are from other countries.

In addition, this project also looks to encourage consumers to move from combustible cigarettes to new options, adds Mahuma. “But in many of these communities, there are often education and literacy gaps, which is why this commerce platform also needs to serve as an education function,” she says. “We wanted to use these channels to promote smoke-free alternatives and to inform the consumer about what’s out there.”

Looking north for inspiration

When developing the platform, Mahuma and her team looked to other African markets facing similar challenges, and with similar demographics. “Senegal’s approach to addressing this problem was inspired by Uber,” she says. “They created a platform that allowed traders and small retailers to place orders and have their stock delivered the same day.” But this wouldn’t work in rural South Africa, where shop owners sometimes order just 10 cigarettes, which they sell to customers as singles. “We knew that giving retailers the ability to self-order was key, but we also knew it’d be too costly to send a scooter to a remote location to deliver 10 cigarettes,” she adds. So they hired scooter drivers to deliver bundled orders, making it more economically viable to handle such small product volumes.

Armed with this data, PMI now has a better idea of where its customer base is growing and where the brand’s market share is dwindling. “In the past, we’d send out trade representatives to try to get a clearer picture of what was happening on the ground,” she says. “But this strategy wasn’t scalable, and it often meant the data we got was inconsistent and incomplete.” Now the business is starting to use data to forecast when a retailer might run out of a product and ensure this doesn’t happen. When supply gaps are quickly filled by competitors or illicit products, you can’t afford stockouts.

While the solution was built in South Africa for the South African market, it’s also being piloted in other African markets. To achieve this scalability, the commerce engine needed to be amended. This market-agnostic design was essential to support different operating models, business processes, partners, infrastructure constraints, and regulatory requirements. “To get this right, we had to break it down and determine which engine components we’d need to change across markets, and which could be reused,” she says. “Although Africa is a single continent, each country faces distinct challenges, and we need to address them in tailored ways.”

For Mahuma, the project’s goal was to expand PMI’s footprint by ensuring the brand’s products were on shelves when consumers looked for them. “So far, we’ve seen an increase in what our wholesalers typically order, which shows that the flow of products has improved,” she says. “This wasn’t a Big Bang approach where we changed processes overnight. We tested the engine and went back several times to check how it was working, and then made changes and improvements as we went along.”