Home Content Studio Solving The Creative Bottleneck In Data-Driven Advertising

Solving The Creative Bottleneck In Data-Driven Advertising

SHARE:

Will ChatGPT take all of our jobs or just make them easier?

I’m optimistic. Even early use cases for AI and tools like ChatGPT have shown to improve our efficiency and quality of life — and the ad tech world holds several early AI success stories. AI is improving performance, enabling us to finally scale up hyper-segmentation, real-time experimentation and one-to-one feedback loops. But, more importantly, AI is freeing up sharp human minds to be more strategic, curious and explore new ideas.

Easing the creative bottleneck

For the better part of a decade, the ad tech world has told a story of hyper-segmentation and feedback-based experimentation. We’ve split audiences into smaller and smaller groups, moving toward the one-to-one dream. We’ve closed the loop on ad impact, so we can see what’s working, often with surprising results — what wins Clios isn’t always what real people respond to – and adjust strategy and creative accordingly.

Refining messaging means more versions of an ad. If you’re doing A/B (or A/B/C/D) testing, you’re now multiplying all those versions several times over, writing copy for dozens or even hundreds of ads. Then you take real-time feedback and do it all over again. And again. And again.

This creative shift needs to happen fast. Advertisers need to meet tight timelines for ad buys as audience preferences shift and interest fades quickly.

In practice, this change ends up limiting the execution. At best, strategists end up spending too much time writing ad copy or they’re forced to engage additional creative talent (in house or third party) that drives up costs and eats into margins. Ultimately, it makes this kind of campaign cost-prohibitive in too many cases.

Natural language models like ChatGPT or Bard can ease this creative bottleneck. In seconds, AI can generate hundreds of versions of ad copy. It’ll give strategists a shortcut or starting point to accelerate their campaign execution workflows, fully exploring the promise of hyper-segmentation without skyrocketing costs or tanking margins.

Using metadata as a solution

General language models are a remarkable technology. But the same “garbage in, garbage out” that governs the rest of our data-driven ad world applies here, too. We need to give AI the right information and guidance to get quality outputs.

Subscribe

AdExchanger Daily

Get our editors’ roundup delivered to your inbox every weekday.

Naturally, the immediate concern is whether the time previously spent writing ad copy will be spent on “prompt engineering,” which gives the AI all the information and context and frames the prompt to get quality outputs.

Campaign strategists aren’t AI experts, nor do they aspire to spend all their time telling AI what to do. Also, it is a fool’s errand to replace the task of writing content for the task of writing prompts. But this burden should not fall back on advertisers and their agency partners. Ad tech partners need to put in the extra work to integrate AI like ChatGPT in ways that are usable and valuable for clients. And that means spending the time to do prompt engineering on the backend.

Ad tech providers can be better positioned to tackle prompt engineering in a data-driven, automated manner. Some ad tech platforms already have all the information needed to inform AI. If they have the metadata on the meaning and purpose of each ad, from detailed audience profiles to real-time impact metrics, they can accomplish amazing results with AI.

At Fluency, we use this rich metadata to drive a large set of precise automated prompts that pull in the right data fields. All of this happens behind the scenes (though we’re fully transparent about what data is being used) so we can give our end users the simplicity of clicking a button to instantly get a helpful, hyper-relevant starting point in the creative-generation process.

Empowering strategists and creatives

It’s worth saying again: This is not the start of AI taking our jobs. This is AI taking on the most repetitive and often least enjoyable parts of our jobs. It’s also firmly in the realm of “attended AI,” which is fully reliant on human expertise to refine and enhance the raw outputs to ensure quality and protect brand safety.

The value to strategists is clear. They have more time to spend on what they do best and what they enjoy most: being strategic and solving problems. They can focus on delivering technical expertise, going deeper in refining audience segments, testing assumptions and following the “learn in isolation, apply in aggregate” mantra. And they can more confidently encourage clients to invest in these data-driven campaigns, knowing the efficiencies will benefit both client and agency bottom lines.

Copywriters and other creatives are understandably more hesitant to welcome AI into their professional lives. But I have yet to meet a copywriter who dreams of writing 50 versions of a digital ad for a new seasonal menu item or two dozen ad headlines for a used car. Even the most stubbornly proud creatives will admit there’s only so many ways to say, “Visit our website for more details.”

Generative AI like ChatGPT is just a tool that creatives can use to do high-volume, tedious work more efficiently. Much like spell check, the AI provides a handful of suggestions that creatives can use as a starting point, applying their expertise to edit, iterate or ignore entirely.

Ultimately, this is a tool that frees up more time for creatives to spend on the kind of work that truly demands and rewards their talent and expertise: the big ideas, the complex messages and the long-form writing.

Using every tool available

A final concern I hear around AI is whether clients will feel duped paying for AI-generated work.

But let’s be real: Advertisers aren’t paying agencies for their labor; they’re paying for outcomes. And there’s always been an implicit expectation that agencies use the tools available to do the best work possible, whether that’s using typewriters, word-processing software or digital tools for SEO scoring and keyword generation. These are all guided by human expertise that gives those humans more time to apply their expertise in strategic thinking, experimentation and guidance. This is what clients want to pay for.

After all, the best ad strategists of tomorrow, like those of today, will shine in finding patterns that drive better results for clients. We’ve already got the tools to identify those patterns. But ChatGPT and other AI tools will fully empower us to experiment with hyper-segmentation, test all our assumptions, move toward the dream of one-to-one feedback loops – and do it all in real time.

For more articles featuring Eric Mayhew, click here.

Must Read

Forget The FUD, Now DoubleVerify Wants Advertisers To Get Back Into The News

Even brand safety companies think news blocking has gone too far. DV is exploring ways to help advertisers support legitimate news and just hired its first-ever head of news.

To Reduce The Ad Tech Tax, Sovrn Expands Its SaaS Pricing Model

Sovrn is now offering its header bidding managed service, dubbed Ad Management, as self-serve software for a flat CPM fee.

play button with many coins isolated on blue background. The concept of monetization of the video. Making money on video content. minimal style. 3d rendering

Exclusive: Connatix And JW Player Merge To Create A One-Stop Shop For Video Monetization

On Wednesday, video monetization platforms Connatix and JW Player announced plans to merge into a new entity called JWP Connatix. The deal was first rumored in July.

Privacy! Commerce! Connected TV! Read all about it. Subscribe to AdExchanger Newsletters

Buyers Can Now Target High-Attention Inventory In The Trade Desk

By applying Adelaide’s Attention Unit scoring, buyers can target low-, medium- and high-attention inventory via TTD’s self-serve platform.

How Should Advertisers Navigate A TikTok Ban Or Google Breakup? Just Ask Brian Wieser

The online advertising industry is staring down the barrel of not one but two potential shutdowns that could radically change where brands put their ad dollars in 2025, according to Madison and Wall’s Brian Weiser and Olivia Morley.

Intent IQ Has Patents For Ad Tech’s Most Basic Functions – And It’s Not Afraid To Use Them

An unusual dilemma has programmatic vendors and ad tech platforms worried about a flurry of potential patent infringement suits.