Every Agency Says They’re Using AI Today. So How Can Businesses Tell Who’s Actually Delivering Results?

Artificial intelligence has quickly gone from being a competitive advantage to something almost every marketing agency claims to offer.

Whether you’re a business owner hiring an agency or a marketing leader reviewing your budget, the question is no longer “Does this agency use AI?” Instead, it’s “Will AI actually help my business grow?”

After all, businesses have spent the past two years investing heavily in AI platforms, yet many still struggle to prove whether those investments are translating into more customers, better campaigns or stronger returns.

According to Safder Ali, Co-Founder and Chief Operating Officer (Southeast Asia) of AI-native marketing technology company Brrandom, the technology itself isn’t the biggest challenge. The bigger issue is how businesses and agencies are choosing to use it.

AI Is No Longer The Competitive Advantage

Three years ago, agencies that embraced AI early enjoyed a genuine advantage. Today, that’s no longer enough. “Almost every agency now says they use AI,” says Safder. “The difference isn’t the tools anymore. It’s how your business is built around them.” He explains that many agencies have simply added AI onto existing workflows, while continuing to operate much as they always have.

Brrandom, however, took a different approach. “We started as AI-native. Our cost structure, team design and delivery model were built around AI from day one,” he says. Rather than deciding which AI platform to use after receiving a client brief, the agency first determines which of its specialist AI practice areas is best suited to solve the client’s business problem.

“It’s a structural difference rather than a feature difference,” says Safder. “Features can be copied within months. Business structures take years to rebuild.”

Why Many Businesses Still Can’t Prove Marketing ROI

One of the biggest frustrations among marketing leaders today is proving return on investment (ROI). Despite having access to more customer data than ever before, many organisations still struggle to connect marketing activities with actual business outcomes. Safder believes the issue usually isn’t the AI platform. Instead, it’s the way organisations are structured.

“The technology to prove ROI has existed for years,” he explains. “The problem is that data often sits in silos.” Marketing teams, CRM teams and sales teams frequently work independently, making it difficult to build a complete picture of customer acquisition and revenue.

In fact, Safder says one of the first things Brrandom checks isn’t campaign performance; it’s whether the client’s measurement systems are working properly. “We’ve seen companies running sophisticated campaigns, but their analytics aren’t even tracking conversions correctly. ” Without reliable data, AI has little chance of producing meaningful insights. “AI can’t prove ROI on top of broken plumbing,” he says. “Fix the plumbing first, and suddenly the AI ROI problem becomes much easier to solve.”

The Companies Seeing Results Focus On Business Outcomes

Many businesses proudly announce AI initiatives. Far fewer can point to measurable commercial results. According to Safder, the companies seeing the greatest success have three things in common.

First, they measure AI against commercial outcomes, not productivity metrics. Rather than focusing on hours saved, they track indicators such as customer acquisition costs, qualified leads and conversion rates.

Second, they redesign workflows instead of simply adding AI into existing processes.

Finally, leadership actively uses AI themselves.

“When decision-makers have personally used the technology, conversations move away from hype and towards practical applications,” he says. Companies chasing trends often run isolated pilot projects that never influence revenue. Businesses seeing meaningful results usually take a much simpler approach. “They focus on one part of the customer journey, measure it properly and improve it over time.”

Has AI Actually Made Marketing Cheaper?

Many founders assume AI automatically reduces marketing costs. The answer, according to Safder, is both yes and no. The cost of producing marketing assets has certainly fallen. Tasks that once required weeks of creative work can now often be completed in days.

However, successful businesses rarely keep those savings. Instead, they reinvest them. “The smartest companies use the savings to test more ideas, create more campaign variations and enter more markets,” says Safder.

As a result, marketing budgets may remain relatively unchanged, while output increases significantly. “The investment shifts away from production and towards strategy, measurement and decision-making.” Businesses that simply use AI to cut costs often see limited improvements. Those using AI to increase the speed of experimentation tend to achieve much stronger marketing performance.

Southeast Asia Isn’t One Market

Following Brrandom’s expansion from India into Southeast Asia, Safder says one lesson quickly became apparent. “The biggest misconception was assuming Southeast Asia is a single market.”

Instead, every country behaves differently. Strategies that work well in Malaysia often require substantial changes before succeeding in Indonesia. Language is another challenge. Campaigns written in the way local consumers naturally speak consistently outperform direct English translations or generic AI-generated copy.

“It’s not something off-the-shelf AI can do particularly well without serious local tuning,” he says.

The agency also discovered that while some industries have highly sophisticated digital capabilities, others are still developing the basics of marketing measurement. Perhaps most importantly, relationships remain central to doing business across the region. “Capability gets you the meeting,” says Safder. “Trust gets you the mandate.”

What Clients Are Paying Agencies For Is Changing

As AI makes content production faster and cheaper, businesses are becoming less willing to pay agencies simply for execution. Instead, clients increasingly expect agencies to provide strategic advice and accountability for business outcomes.

“They’re paying for judgement,” says Safder. That includes helping businesses decide which audiences to target, which markets to prioritise, how campaigns should be measured and what the data is actually saying. Execution still matters. But increasingly, it’s the strategic thinking behind the execution that clients value most.

Most AI Investments Don’t Fail Because Of The Technology

Businesses have invested heavily in AI software over the past two years. Yet Safder estimates that only a minority of those investments generate meaningful commercial value. Interestingly, he doesn’t blame the platforms themselves.

Instead, unsuccessful implementations usually have something else in common. The business buys software without assigning ownership, redesigning workflows or defining success metrics. Successful AI adoption, on the other hand, is often surprisingly simple. “One platform. One use case. One accountable owner. One commercial KPI.”

He compares AI adoption to hiring a new employee. “You wouldn’t hire someone without defining their role, manager and performance targets. Businesses should approach AI investments the same way.”

The Future Of Agencies May Look Very Different

While many believe AI will eventually replace marketing agencies, Safder sees a different future. He believes agencies will continue to play an important role, but they’ll look very different from today’s traditional agency model.

Rather than relying on large teams carrying out repetitive execution work, successful agencies will become smaller, more senior and increasingly AI-enabled. “The real value isn’t the tools,” he says.

“It’s judgement, experience and recognising patterns across different industries and markets.” His prediction is that AI-native agencies with relatively lean teams will increasingly outperform much larger traditional agency networks. For businesses, that means choosing an agency based less on whether it uses AI, and more on whether it knows how to turn AI into measurable business outcomes.

As Safder puts it, “Technology is becoming table stakes. The real competitive advantage comes from what you build with it.”