Conversion is strong and cost per acquisition is up again, yet more spend is not producing more growth. If you run marketing for a South African business, you will have been asked why. The honest answer is that we are only converting people who already have intent, not creating new demand.
The solution requires a shift in investment towards building consideration, earning trust and creating demand. The problem is that these investments take time to deliver, often two or three quarters, while the pressure to demonstrate immediate returns continues to intensify. NIQ’s CMO Outlook for 2026 highlights the contradiction: only 69% of marketing leaders surveyed globally believe their CEO and CFO support long-term brand investment, down eleven points in a year, yet 84% prioritise return on investment when allocating budgets.
The uncomfortable truth is that we have become exceptionally good at harvesting demand, while steadily losing the ability to create it.
Demand is not created at the point of purchase
Creating demand means giving people a reason to want your product or service. Harvesting demand means capturing customers who have already decided they want to buy. Most marketing metrics celebrate the latter, and marketers know it: 52.9% told Marketing Week’s 2025 Language of Effectiveness research that their campaigns are too focused on performance and sales rather than brand building.
Awareness at the top, intent at the bottom, and a hole in the middle
Most brands I work with are not short of awareness, and a competently run paid social and Google account will reliably convert anyone who intends to buy this month. What is missing is the part in between, where the decision is actually made.
In South Africa, preferences build gradually through conversations, recommendations from friends, creators people trust, reviews and WhatsApp groups. By the time someone types your brand into Google, they are validating a preference that was formed somewhere else.
The middle has moved to places you do not control
Rogerwilco’s 2026 South African Customer Experience (CX) Report, drawing on insights from 2,000 consumers and 50 business leaders, shows what happened to that missing middle. Fifty-eight percent of consumers rate customer reviews on a company’s website among their most helpful research sources, and 44% say the same of independent platforms like HelloPeter. Only 16% of business respondents expected customers to value on-site reviews. Meanwhile, AI search use has nearly doubled in a year, from 12% to 23%.
The consideration part of the journey is now influenced by reviews you did not write, AI answers you did not influence, WhatsApp groups you are not part of and creator content you did not brief.
AI is making differentiation harder, not easier
AI is becoming the great equaliser: similar optimisation models, audience signals and automation tools make execution easier and differentiation harder.
There is another way AI is influencing differentiation, and it is easy to underestimate. When someone asks an assistant which medical aid to choose, which bakkie holds its value, or where to order dinner tonight, they are handed an answer.
The engine assembling that answer reads not only your campaign, but what other people have said about you, in places you do not own and cannot buy. The brand nobody talks about is far less likely to make it in. Yet only 27% of business leaders in the CX report have a dedicated strategy for appearing in AI-generated answers.
The middle of the funnel is harder to plan
Top-of-funnel work can be broad: one idea, well made, in a big burst. Bottom-of-funnel work can be systematised, because the person has already decided. The middle has to answer different people with different needs.
One segment is worried about the total cost of ownership. Another was burned by a competitor and needs proof you will not do the same. A third understands the product perfectly and is waiting to hear that someone like them has already chosen it. Addressing that takes more segments, creative versions, channels and patience than either end of the funnel allows, but it is where the demand for your bottom-of-the-funnel results is created.
What to take into the budget conversation
If the middle of your funnel is going to get funded, somebody needs to make the case in the language of the people holding the budget. Start with the data you already have. Bottom-of-funnel customer data shows which groups are worth the most and where you are genuinely competitive but have never said so out loud. Test one, and when it works, the budget naturally follows. Three things help:
- Separate the two jobs in your reporting: Track what share of your conversions arrive already knowing who you are. If it is shrinking while your performance numbers hold steady, what you are reporting is not efficiency. It is an early signal that demand creation needs funding.
- Reframe the request: Cutting mid-funnel investment does not save money. It commits the business to buying every future sale at full price, in an auction, against competitors who did not cut. That is a cost structure rather than a brand preference, and finance teams speak that language fluently.
- Go and read what is written about you in the places you do not control: Reviews, forums, comparison content, AI answers. It is where your customers are forming their view, and a more persuasive exhibit in an exco meeting than another chart about brand affinity.
The hole in the middle of your funnel is rarely a gap in thinking. It is a gap in funding, and it closes once you can show finance what leaving it open costs.
