The Objective
You’d be forgiven for thinking that “ROMO” has something to do with the former Dallas Cowboys quarterback. But you’d be wrong.
ROMO, which stands for “return on marketing objectives,” is the abbreviation for a new metric being pushed by Orange Apron Media, The Home Depot’s retail media business, Digiday reports.
Zach Darkow, Home Depot’s senior director of marketing measurement, argues that ROMO is a better guideline for advertisers than ROAS, which focuses on directly attributable conversions that can be tied to specific ad impressions.
The ROMO process is more arduous, however, since campaigns might be credited by, say, brand lift surveys conducted over months during and after a campaign. ROMO measurements also require vendor help and take a great deal more time to create.
By comparison, a two-day ROAS report that claims every dollar spent magically turned into $4 earned has an enchanting simplicity.
Darkow previously elaborated on the notion of ROMO in a presentation at Programmatic I/O in Las Vegas last year (not to brag). He also noted that his former employer, Target’s Roundel ad business, preferred a metric called WISS (web-influenced store sales).
Although WISS never took off, retailers like it because it aims to capture the often intangible boost they generate for brands.
Breach Of The Peace
Elon Musk’s X is dealing with the fallout of a massive data breach. Billions of current and inactive accounts, including bot accounts, are reportedly now available online.
Passwords are not part of the trove. Instead, the breach is of personal emails tied to accounts and publicizes all sorts of valuable metadata attached to those accounts, reports Dataconomy, a German tech and business news publisher.
Account metadata reveals location information and other important contextual signals about an individual, like devices used over time and posting history. While it’s not as obviously valuable as a password, bad actors can use this type of data to perpetrate newer, more sophisticated forms of online fraud.
People nowadays aren’t so easily duped into clicking buttons they shouldn’t. Instead, fraudsters string them along through conversations over days – and sometimes even weeks or months – and they can do this because they have access to rich histories of online behavior that can be tied to real-world events.
“One clear lesson from any Twitter breach is that phishing campaigns tend to ramp up fast,” Dataconomy writes. “In the case of the X breach, attackers could craft incredibly convincing emails using real metadata.”
Dollars Ex Machina
Did the Trump Administration use an LLM to plan its new global tax policy?
It appears that way to some onlookers.
According to Inc’s James Surowiecki, the White House likely took the US trade deficit for each country and divided it by the total goods imported into the US. (The Office of the United States Trade Representative essentially confirmed this later on, albeit in an intentionally overcomplicated fashion.)
Essentially, it’s like forcing a store to charge you more because it’s not buying enough or anything from you, which is hardly a standard economic policy. So whose idea was it?
Out of curiosity, engineer Rohit Krishnan decided to pose the same prompt to ChatGPT, Claude, Gemini and Grok, and all four produced the same response: trade deficit divided by total imports. (DeepSeek, for what it’s worth, was the only one that dug into the potential consequences of such a simplistic plan.)
Interestingly, although many online now seem eager to start a Butlerian Jihad, Krishnan himself is more forgiving.
“I think the fact they’re using LLMs is good but they def need better prompt engineers,” he posted on X, later adding: “a second question would’ve explained the issues with the approach in exhaustive detail.”
But Wait! There’s More
Since Trump’s tariffs announcement, Apple has lost more than $250 billion in market value, with shares down as much as 8.5%. Tesla, Nvidia and Meta were down 6%, and Amazon shares fell by 7.2%. [TechCrunch]
TikTok owner ByteDance faces a privacy fine of more than $550 million for shipping European user data to China. [Bloomberg]
Meanwhile, the looming TikTok ban is starting to feel like third-party cookie deprecation 2.0. [Adweek]
YouTube quietly removed “gender identity and expression” from its hate speech policies. [User Mag]
Creating intentionally deceptive deepfakes can now land you up to five years in prison in New Jersey. [Fortune]
You’re Hired!
Infillion taps Jeremy Woodlee to lead its enterprise business unit. [release]
LegalShield names Jaymee Johnson as CMO. [release]
