The AI Divide: Navigating the New Research Landscape

The rush to build proprietary AI shows just how seriously the industry takes it. But the explosion of branded tool, combined with about 30% of staff using AI however they like reveals something else: adoption is chaotic.

Client pushback is still limited. But the cracks are obvious - 95% of users report flaws in AI. If clients start spotting cookie-cutter outputs and sloppy thinking, an AI backlash isn’t far off.

Synthetic data, once hyped, is losing momentum. More stakeholders are opting out, and its use in niche audience analysis is stalling, e.g. synthetic boosts. Skepticism maybe creeping in.

However, AI is irresistible. The time and cost savings are massive, and outputs look convincing, especially when polished or visualized. ‘Attractive plausibility’ is doing a lot of the heavy lifting.

AI is here to stay, and the future market will split three ways:

Tech providers selling fast, cheap, self-serve AI (plus some vertical specialization)

Lean agencies using AI for the project & analytical grunt work, with senior talent adding strategic value

“Human-led” research firms selling depth and context, while using AI behind the scenes, but claiming limited use of it!

Disruption is already here: 35% report staff displacement. Quant was already heading toward automation, but AI accelerates it, even if it is not strictly AI that is the automation.

Notably, the Asia Research survey shows that qual researchers are using AI across a wider range of applications, making their work quicker and more scalable. That side of the industry could surge as ‘qual at scale’ becomes the buzzword.

Fewer people will be hired into the industry but those who do enter and stay the course, will outperform their predecessors as they leverage AI more skillfully. The talent pool will be sustained, smaller but enhanced.

The real question: how many agencies will survive? An 80:20 rule looms large – if clients can get 80% of the insight for 20% of the cost, some will take that option for many projects. That puts generalist agencies under pressure, especially as clients bring standard research in-house. Some will limp on via public sector work, where insourcing is less common, but that market is already shark-infested Red Ocean.

But the missing 20% of insight matters. It can be the difference between success and failure, whether launching  a product, entering a market, reshaping a business. AI tends to generalize, miss outliers, and often hallucinates - that’s its Achilles’ heel. Specialist agencies can own that gap – offering unique perspectives, proprietary insight, industry knowledge, holding an opinion, and the ability to spot the significance of outliers.

Meanwhile, large agencies face a different squeeze: heavy infrastructure, rising overheads, burdensome administrative headcounts & procedures, and pricing pressure from clients that cannot support these costs.

As smaller, AI-enabled players scale up and clients consider cheaper options, the old model starts to creak.