- Most AI investments are failing to deliver ROI, with 95% of organisations reporting no measurable returns despite billions spent on generative AI.
- Efficiency is not enough as cutting costs and increasing output does not automatically create business growth or competitive advantage.
- AI leaders redesign workflows, using technology to transform how work is done rather than simply automating existing processes.
- More content does not equal better results. Effective media strategies focus on smarter decisions, not just higher volumes of ads and assets.
- Human judgement remains essential, with the strongest results coming from combining AI, data and automation with human creativity and strategic thinking.
There’s a seductive story being told across boardrooms right now: adopt AI, move faster, do more with less. The pitch is compelling. The dashboards are impressive. And yet, for most businesses, the returns are nowhere to be found.
The most visible consequence has been headcount reduction. But organisations that treated AI as a staffing equation are discovering that removing people without rethinking how work gets done does not create a smarter business. It creates a thinner one.
A 2025 report from MIT’s Media Lab found that despite $30 to $40 billion dollars in global enterprise spending on generative AI, 95% of organisations are seeing no measurable return on investment. McKinsey’s 2025 State of AI survey tells a similar story.
While 88% of organisations now use AI in at least one function, only around a third are scaling it meaningfully, and just 6% qualify as high performers delivering real bottom line impact. Meanwhile, Gartner found that only 5% of marketing leaders using generative AI purely as a productivity tool reported significant business gains.
The pattern is clear. The industry has been fixated on efficiency. Speed, volume, output. But from our experience leading AI implementation and media strategy inside the same agency group, being fast is not a competitive advantage when everyone has access to the same tools.
The real question is whether you are using technology to do the right things, not just to do things faster.
The efficiency trap
It is easy to understand why efficiency became the default objective. Budgets are under pressure. Expectations around quality, scale, and speed have not dropped. In many cases, they have increased. AI and automation offer a credible way to do more within those constraints. But efficiency on its own does not change what gets done. It only changes how quickly it happens.
McKinsey’s research draws a sharp line here. High performing organisations, the ones actually generating measurable value from AI, are nearly three times more likely than others to have fundamentally redesigned their workflows. They do not simply bolt AI onto existing processes.
They rethink the processes themselves. Importantly, while 80 percent of companies set efficiency as an AI objective, high performers also set growth and innovation as goals. That broader ambition is what sets them apart.
A landmark study from Harvard Business School and Boston Consulting Group adds another layer. Consultants using AI completed tasks 25 percent more quickly and produced 40 percent higher quality results, but only when AI was applied to the right tasks.
For tasks outside AI’s competence boundary, what researchers called the jagged technological frontier, AI actually made people perform worse. The first job is not to automate everything. It is to understand where AI genuinely adds value, and where human judgement and creativity are still critical.
More ads is not a media strategy
From a media strategy perspective, Scott Reinders sees this playing out every day.
“The temptation is to use AI to produce more. More ads, more placements, more variations. But without asking whether that volume is actually driving outcomes,” he says.
“Every scrolling feed, every open web banner gets packed with deals and promotions. Scale without signal is just noise.”
Gartner’s data supports this concern. A 2024 survey of marketing leaders found that 87% of CMOs experienced campaign performance issues in the previous year, with nearly half terminating campaigns early. More output has not solved the challenge of cutting through in a cluttered market. If anything, it may be making it worse.
Reinders argues that the real competitive edge in media comes from making sharper decisions about where, when, and how to invest.
“Platforms that deliver strong active attention, TikTok when creative aligns with the format, YouTube non skippable for storytelling, work because someone made a deliberate choice about the environment. That is an effectiveness decision, not an efficiency one.
”For South African brands operating with tighter budgets than their global counterparts, this distinction matters even more. When you cannot outspend the competition, you need to outthink it. That means using technology to improve the quality of strategic decisions, not just the speed of execution.
What effectiveness actually requires
Moving from efficiency to effectiveness is not a matter of slowing down. It is a matter of redirecting where the speed is applied. From our experience, three shifts separate the organisations generating real value from those still chasing output.
From generic to contextual
When every company has access to the same AI models, plugging them in without adaptation produces the same generic outputs as everyone else.
The organisations pulling ahead are the ones embedding their own context into how AI is deployed: how their teams actually work, the signals they respond to, the judgement calls that repeat across real situations. The tool is commoditised. The application of it is not.
From executional to strategic
Too often, AI enters the workflow at the point of execution: generating assets, populating templates, speeding up production.
But if the brief was wrong, or the channel strategy was misaligned, faster execution just delivers the wrong thing more efficiently. The shift is to move AI upstream, into the strategic decisions that determine what gets made and why.
From linear to iterative
Traditional campaign cycles run on a launch-and-review cadence that measures results in quarters. By the time the data comes back, the budget is spent.
Redesigning workflows around continuous optimisation, with shorter cycles, live feedback, and rapid adjustment, reduces the risk of full campaign failures and turns every deployment into a learning opportunity.
How The Up&Up Group is approaching this
These aren’t abstract principles. They’re the foundation of what The Up&Up Group has built.At The Up&Up Group, the team developed the Intelligent Practice framework because the gap between AI adoption and AI impact was widening, both internally and across the industry.
“Intelligent Practice is our model for integrating human creativity, data, AI, and automation across six specialist agencies,” says Izak van der Walt, project director for AI and Automation at The Up&Up Group.
“We documented this approach in our white paper because the experimentation phase is over. Clients and agencies need results, not endless pilots. The cost of unfocused implementation, in wasted investment and team fatigue, is real.
”The organisations that thrive in this next chapter won’t be the ones that moved fastest. They’ll be the ones that knew where they were going.”
Izak van der Walt, project director: AI & Automation, The Up&Up Group, and Scott Reinders, chief operating officer, Connect (part of the Up&Up Group).
