• by Joe Mandese  @mp_joemandese, September 12, 2023

It's been eight years since MediaPost published "Hyperbabble," a regular feature pulling anonymized, verbatim ad tech rhetoric directly from press releases and email pitches as a way of calling out the deluge of meaningless hyperbole inundating the ad world. Based on some recent pitches, I'm thinking of bringing it back as part of our new Media3.0 newsletter. What do you think?

While you mull on that, I figured I'd weigh in this morning on a not-so-anonymized d.scription of something Dentsu is pitching: A new-and-improved way of executing search vis-a-vis generative AI.

I find that interesting for several reasons, including the fact that many already consider generative AI "the new search," or at the very least, will be as soon as Google, Microsoft, OpenAI figure out more intuitive ways for searchers to use it.

Apologies to Dentsu for d.constructing the announcement of its new tool, because I'm sure it truly is innovative, but in the spirit of Hyperbabble, it doesn't really explain what it does, how it works and why it's materially better than the tools your search team has been using before you branded it.

The truth is that search -- like any tech and data powered advertising experience -- has been undergoing constant innovation, much of it taking place within agencies like Dentsu, and especially within its Merkle unit, but we rarely get to see under the hood, just the sizzle-based d.scriptions of it.

I say this as someone who near this time each year, issues a request for submissions to MediaPost's annual agency of the year awards (stay tuned for that), including a category focused on search and performance marketing.

We usually get a lot of submissions, but most of them are heavy on the platitudes, and light on the details about the vision and innovation that make them award-winning. What we pick and what you see published, is the best of the best.

I'm hoping Dentsu once again is among those making a case for those awards, and if it does, I'd recommend it d.scribe what something like d.scriptor can do:

• "Significantly Accelerates Ad Creation Process: Leverages Generative AI to accelerate ideation process, while abiding by character limits and other ad copy best practices. Initial estimates indicate up to 6x time savings compared to manual ad copy optimization approach while reducing overall task complexity.

• "Highly-Relevant, On Brand Ad Copy: Incorporates pertinent information such as current ads, performance insights, and website content to inform ad creation that mirrors brand voice and aligns with existing messaging strategies.

• "Improves Output, Ad Strength & Engagement: Boosts performance by accelerating experimentation through the total available messaging inputs and by generating highly-relevant messaging to improve ad strength and maximize engagement over time.

• "Supports Human-Led Review & Refinement: Engineered to enhance human-led ad creation and approval workflows, increasing confidence & minimizing risk. End-to-end integration with leading platforms provides users with the opportunity to QA, refine, and gain necessary approvals prior to upload.

• "Data Security: Built within dentsu’s enterprise Generative AI models, wherein all data is secured within private Dentsu network. Alternative vendors vary, but none support the full ad copy optimization workflow in a cohesive, secure manner like this tool does."

And focus more on how it does it and why it is different and more innovative than the way it's been done before.