Beyond Reach: Can AI-generated ads turn attention into action?

In the wonderful world of digital advertising, there are few things quite as reassuring as a dashboard full of green arrows. Impressions are up. Clicks are up. Engagement is up. The campaign has reached several million people, and the marketing team is preparing a presentation explaining how thoroughly everyone has been reached.

Whether any of those people subsequently bought the product is, of course, another matter.

As we’ve previously discovered, everyone is aboard the AI bandwagon now, brands especially. The industry is delighted to discuss how artificial intelligence is transforming creative production and marketing workflows.

Also Read: The great AI content glut: When everyone can make everything

AI can make infinite content. But can it make anyone care?

ChatGPT Ads just hit $1bn run rate. Now comes the price test in India

Start asking whether AI-generated advertising is actually delivering better results, however, and the enthusiasm can give way to a rather impressive barrage of no comments.

Apparently, discussing what AI could do is more comfortable than discussing what it has done.

There is reason to ask. An August 2026 NP Digital study of 100 companies found that fully AI-generated material accounted for 52.1% of new content but attracted just 4.9% of organic traffic. Human-created content, meanwhile, represented 14.5% of output but accounted for 87% of organic traffic. The figures point to a striking disparity between the amount of AI content being produced and the audience it attracts organically.

But organic reach is only part of the equation. Advertisers can pay to put AI-generated creative in front of defined audiences, personalise it and test dozens of variations. That raises deeper questions: Does getting an ad in front of someone mean they will notice it? And does getting their attention mean they will act?

After all, reach can be bought. Attention is harder to secure. And a click, however enthusiastically celebrated, is not quite the same thing as a customer.

When the numbers start looking good

“AI has solved the production problem, not the attention problem,” says Samir Asher, Founder & COO, Tonic Worldwide.

Established platforms, he explains, already have audiences, habitual usage and distribution working in their favour. What AI cannot guarantee is that the content being delivered will interest those audiences. “The ceiling is not AI, the ceiling is sameness. The mistake today is using AI as a content factory instead of an intelligence layer.”

Nevertheless, Indian advertisers are reporting measurable results from campaigns incorporating AI-generated creative.

Appliance manufacturer Usha used Amazon Ads’ AI-powered image generation to develop creatives for its BLDC fan range, alongside audience targeting and a broader advertising strategy.

According to an Amazon Ads case study, the campaign delivered a 2.4-fold increase in branded searches, 36% higher ad-driven page views and a 32% improvement in return on advertising spend (ROAS). Usha also reported quarter-on-quarter sales growth exceeding 50%, while the AI-generated creatives achieved higher click-through rates (CTR) than static creatives.

The figures are platform-reported, and the campaign involved changes beyond creative production, so the improvements cannot be attributed to AI-generated imagery alone. Still, they illustrate an important distinction: an advertisement need not earn an organic audience to produce a commercial result.

Angad Saimbi, Associate Director – Business and Growth, Yellow Seed, believes consumer intent is central to that distinction. “Inside a product, AI is helping someone do something they already came there to do. There’s an existing audience, context and habit.”

Yaron Tomchin, CEO, Mobupps, points to the structural advantage of closed platforms, which already have habitual users, behavioural data and built-in distribution. But he cautions that scale alone is no substitute for value. AI can accelerate production, personalisation and optimisation, he says, while originality, context and trust still depend on human insight.  

On the open web, by contrast, content must earn attention from scratch.

For Tushnik Thakur, Deputy Head, Content & Creator Business, IN, AnyMind Group, the opportunity lies in using AI for creative variations and dynamic personalisation rather than relying on it to generate the underlying creative idea.

He points to AI-generated, user-generated-content-style videos that demonstrate products, answer questions and address purchase objections during the consideration stage, saying, “The technology creates speed and scale; the strategy, human input and consumer truth create relevance.”

An inconvenient conversion question

Successful distribution, however, does not guarantee success further down the marketing funnel.

A June 2026 conference paper by Hong Kong Baptist University researchers Huimin Li and Qing Li examined more than 150 unique video advertisements in a randomised field experiment. AI-generated advertisements achieved higher CTR, stronger viewer retention and fewer negative reactions than human-created advertisements, but performed less well on conversions.

The publicly available abstract does not provide the numerical performance gaps or detailed campaign conditions. Nevertheless, the finding highlights a distinction advertisers cannot afford to ignore: an ad can attract attention without persuading someone to buy.

Apurv Verma, Executive Director, SRV Media, reports a different result from campaigns he has worked on. “In my experience, fully AI-generated videos have seen lower CTR because audiences are now exposed to a high volume of similar AI content. Videos that combine real human characters and AI elements have performed better, resulting in stronger CTR and overall performance metrics.”

Verma’s experience is not a controlled industry-wide comparison, but it suggests that how AI is used may matter more than whether it is used at all.

A September 2026 working paper by researchers from Columbia Business School and other institutions examined more than 16 billion display advertising impressions and 116 million clicks. It found that ads featuring AI-generated images could achieve higher CTR than those using human-generated images, provided the AI imagery did not visibly resemble stereotypical AI content.

The study concerns display advertising rather than video and does not establish whether additional clicks produced additional sales. Taken together, the research and industry experience resist a simple verdict on whether AI-generated advertising “works”. Performance varies by creative execution, format and the outcome being measured.

Relevance is not repetition

There is also the question of how much AI-generated advertising consumers are prepared to encounter.

“Attention hasn’t increased just because content supply has, so unconstrained volume can mean greater creative fatigue, weaker engagement and more opt-outs if communication isn’t relevant,” says Srinivasan Subramani, VP – Growth & AI, CleverTap.

This is particularly important in customer relationship management, where brands can use AI to personalise messages for consumers they already reach. Generating another communication is easy; giving its recipient a reason to welcome it is rather more difficult.

Keyur Dhami, SVP – Customer Success (Key Accounts) & CoE, WebEngage, similarly argues that engagement depends on the relevance, context and originality of the message, not merely the ability to personalise and produce it at scale.

That distinction becomes more consequential when a click is only the beginning of a lengthy purchase journey.

Prashin Jobhalia, CMO, House of Hiranandani, points out that real estate buyers are researching major financial decisions, not simply scrolling for quick information. The company therefore “combines AI-driven efficiency with human, market-specific insight to provide the depth and credibility such decisions require.”

A consumer might click on an attractive property advertisement. Persuading them to purchase the property is an altogether different assignment.

The metric that matters

None of this means AI-generated advertising is inherently less effective than conventionally produced creative. It can also be commercially worthwhile without outperforming human-made work on every metric: comparable results at substantially lower production costs may represent a meaningful gain.

But the ability to create and distribute more advertising makes it increasingly important to distinguish between metrics that demonstrate activity and those that establish business value.

CTR measures clicks. ROAS measures attributed revenue against advertising spend. Neither, on its own, tells the whole story of incremental sales, profitability or lasting customer relationships.

AI can help advertisers test more messages, adapt creative faster and reach consumers more efficiently. It cannot guarantee that the person encountering an advertisement will want the product.

The industry has always talked about reaching the right person with the right message at the right time. AI has made producing and delivering that message considerably easier.

The rather old-fashioned business of persuading someone to act on it remains a separate challenge.

And until that happens, all those lovely green arrows on the dashboard are still just arrows.