AI Briefing: How Lexicon researches its approach to AI naming strategies

In cluttered and noisy industries, naming companies and products has often been part of the strategy for standing out. That continues to be the case when it comes to AI, but there isn’t yet the same depth of research that exists in other categories.

Lexicon — the naming agency behind numerous famous names like Blackberry, Apple’s Powerbook and Adobe inDesign — is hoping to change that as AI names become increasingly important.

Last week, Lexicon released its initial findings from a survey that explores what consumers and developers think about AI. Surveying 150 consumers in the U.S. and another 150 in Germany, the agency asked a range of questions about which industries people think will be positively impacted by AI. The top industries mentioned were telecommunications and health care — which were named by about half of respondents — followed by education, entertainment and security. The lowest were banking, hospitality and travel, retail, and legal services. Another 21% said none of the above.

The goal with the research is to get ahead of the curve, said Lexicon founder and CEO David Placek, who added he’d like to create an AI technology branding practice within the agency. That’s where the research investments can help. While there has been more interest from AI software companies than from other areas like health care, he thinks that will change in the next year or two.

Placek declined to share any examples of Lexicon’s current AI naming clients, but the agency is also behind tech names like Oculus Go, Microsoft Azure, Verizon Fios, Adobe inDesign and plenty of other iconic names of household products ranging from cars to cleaning products.

“I definitely want to avoid talking the hype or anything like that, but I think it’s going to be more challenging to create truly effective brands in this space,” Placek said. “There’s going to be a lot of noise, a lot of clutter, and so we’re trying to figure out: What’s the secret here? What’s the code?”

When consumers were asked to describe emotions they felt when using AI, the top two were “empowered” (52% of respondents said this) and “smart” (40%). Emotions like “energized,” “happy,” “excited” and “delighted” were each mentioned by between 13% and 16% of respondents. However, only 9% said they felt “joyful,” “relaxed” or “calm.” Survey respondents also expressed some of their concerns with AI. Although half said they perceived AI as “powerful,” “innovative” and “efficient,” only 9% chose “responsible,” and just 8% chose “safe” or “secure.”

Before Lexicon released its research, Digiday spoke with Placek about the findings and what they suggest when it comes to AI naming strategies. Below is a part of that conversation, which has been edited for brevity and clarity.

Digiday: How do these insights change your approach to naming AI companies and products?

According to Placek, for a client that has a “really sophisticated buyer,” it might tone down the name with “a kind of mild name that doesn’t raise suspicions and angst and things like that.” However, it’ll vary based on the type of brand, the industry it’s in and how it uses AI. Placek suggested health care as an example of where a company might apply AI to find what leads to a bigger generation shift in various types of treatment.

“Then I think we would sort of have a more uplifting name that supported that,” Placek said. “But I think we have to have an aperture — I’ll stick with lenses — that opens and closes based on what this thing can really, really do or not do. We’re learning as we’re going here.”

Thinking of that same aperture metaphor reminds me of how that needs to change based on speed and light, especially with the AI space moving so fast while some aspects are in the spotlight and others still in the dark.

All the clutter and noise reminds Placek of the early days of mobile apps for everything from searching for stars to apps for playing music. Ultimately, it became a lot harder for any one to stand out as the universe of apps expanded to thousands.

“I think we’re gonna have the same thing happen here,” he said. “With some apps, some products, some companies, there’s going to be this huge surge of things. And a lot of that will go away — just like how we lost a lot of companies from 1995 to 2005 when they just kind of filtered out. But we’ve got to figure it out. How do we develop strong signals for our clients as opposed to joining the noise in the marketplace?”

How does the naming strategy differ for enterprise companies and products versus consumer brands? For example, various enterprise-grade AI models like Cohere’s Command family of models or Snowflake’s Artic.

One prediction from Placek is that there will be a “gravitational pull” for companies to have a lot of new names in their portfolio. But when there is industry contraction, companies might realize their customers are confused by so many names. Instead, he suggests clients be very focused with naming strategies.

“Some things don’t need to have an AI brand next to it,” he said. “You could simply have a message [like] ‘supported with AI’ and oftentimes you don’t even need to say that. This is like the word ‘digital,’ right? In a few years, we will just expect things to have some level of artificial intelligence in it. So from an architecture standpoint, I would advise clients to choose your battles, be very focused and never over promise. If anything, under promise and over deliver. Keep it clean. Keep it simple.”

Is there a need to file a lot quicker, because there might be a timeline or a waitlist or competition for those names?

A few decades ago, Lexicon would often have three or more months to go through the process with a client. However, that’s now down to six or seven weeks, Placek said, adding that sometimes it’s less than that.

“Everybody is moving really as fast as we can go,” he said. “Hopefully, we don’t get more time pressure than we already have. They always asked, ‘Hey, can you do this in three weeks?’ And we say no. If you want to do it right, if you want your version of a Sonos, or Febreze, or a Powerbook, you need to give us six, seven or eight weeks to do it.”

Other AI-related stories from across Digiday

Prompts and Product: AI and other announcements

  • OpenAI debuted new capabilities for ChatGPT, including ways for the chatbot to view and identify real-world images, read text on a screen, and interact in a more natural way in audio conversations. The features are powered by a new model called GPT-4o, which OpenAI said is faster and more efficient than GPT-4. The company also made GPT-4o available for free to all users and rolled out a desktop version of its app.
  • Google hosted its annual Google I/O developer conference, which was packed with a range of generative AI updates for search, music and video generation, and ways to see the world using new AI agents like Project Astra and Google Lens. (Read Digiday’s recap.)
  • Search experts expressed concern about how Google’s new AI search updates could harm internet traffic for publishers, content creators and advertisers. Meanwhile, privacy experts are worried that Google’s plans to use AI on devices to protect against spam and AI deepfakes might become a privacy concern if it’s misused.
  • At the 2024 upfront event, legacy networks and digital streaming platforms touted new AI capabilities for audience targeting, shopping and other uses.
  • Congressional committees in the House and Senate both held hearings about 2024 election security and the risks of AI. In the Sentate Intel committee, leaders of several U.S. intelligence agencies testified about their concerns and what they’re doing to mitigate the risks of AI deepfakes and other concerns.
  • U.S. Sen. Chuck Schumer unveiled a plan to address AI safety, which comes after his yearlong series of talks in 2023 with various tech companies, AI researchers and other experts.
  • Slack and Squarespace both faced criticism from customers for allowing content on their platforms to be used for training AI models.
  • Sony Music sent a letter to hundreds of AI companies warning them against unauthorized use of content from the music label when training AI models.
  • Reddit announced a new partnership with OpenAI that will let the company train its AI models on Reddit posts. The deal comes just months after Reddit revealed Google is also paying the platform tens of millions of dollars to train its AI models on Reddit content.
  • The Open Markets Institute and others contributed to the U.K. Competition and Markets Authority to review partnerships between tech giants and AI startups.
  • A new class action lawsuit against the AI voice-cloning startup LOVO alleges false advertising and fraud.

Ads from AI: News from the creative side of generative AI

  • KFC faced criticism for its use of AI in a new ad that invited fried chicken fans to generate images with six fingers. (Fast Company)
  • Jewelers have begun using generative AI tools for creating images and videos in new advertising campaigns. (The New York Times.)

Hellos and Goodbyes: Noteworthy hires and departures at AI companies

  • Ilya Sutskever, OpenAI’s cofounder and chief data scientist, announced he was leaving the company. The startup also suffered another key departure with the resignation of Jan Leike, co-lead of OpenAI’s Superalignment trust and safety team. According to Leike, the startup’s safety culture had “taken a backseat to shiny products.”
  • Anthropic hired Instagram co-founder Mike Krieger as its new chief product officer, leaving some to wonder what the creator of the OpenAI competitor — and creator of the Claude chatbot — might have planned for improved user experience and other updates.
https://digiday.com/?p=545291

More in Media

Media Briefing: Publishers’ Q3 earnings show revenue upticks despite election ad pullback

Q3 was a mixed bag for publishers, with some blaming the U.S. presidential election for an ad-spend pullback.

Workplace policies poised for seismic shakeup post-election

Topping the list of expected changes: a rollback of many health insurance reforms provided under the Affordable Care Act, better known as Obamacare.

News publishers didn’t sustain a traffic bump in the 2024 presidential election week like they did in 2020

Unlike the drawn out process of the presidential election in 2020, this year’s election quickly revealed that Donald Trump would be the winner – and that meant less of a sustained traffic bump to publishers.