Jordan Smith, founder and CEO of mymodel, is building a marketplace where consumers get paid for their data instead of having it scraped, guessed at, and sold behind their backs.
Here’s a stat that stopped me in my tracks. Cookie data is only 47% accurate at predicting someone’s gender. You would get a better result flipping a coin. And this is the data infrastructure that’s powered digital advertising for the last two decades.
I talked about this with Jordan Smith, founder and CEO of mymodel, on a recent episode of Deconstructing Data (my colleague Adam Fitzgerald, our enterprise sales director, co-hosted this one while Jessie was out). Jordan spent time at Microsoft working with enterprise retailers before moving into adtech, and what she saw there convinced her the entire model of buying and selling consumer data needed to be rebuilt from the ground up. Not tweaked. Rebuilt.
Her pitch is simple: stop guessing what people want based on scroll depth and time-on-page, and instead ask them directly, pay them for the answer, and use data you actually know is true. That’s zero-party data, and it’s becoming one of the most important concepts in marketing right now.
What Is Zero-Party Data and Why Does It Matter Now?
Zero-party data is information a consumer deliberately and directly shares with a brand, as opposed to first-party data (collected through behavior on your own properties) or third-party data (bought from someone else, often with murky origins). The difference matters because zero-party data comes with actual intent behind it.
“How can we turn this concept of sharing data in from something that is scary to most people and feels out of the hands of most people to something that can feel like empowering and like you’re participating in something and has some real incentive for the user.”
— Jordan Smith, Founder and CEO, mymodel
Jordan calls her approach to explaining this to users “digestible transparency.” Her benchmark for it is refreshingly simple.
“I coined this term called digestible transparency. How can we be transparent with consumers in the way that my 93-year-old grandma in Florida could understand? That should really be how we look at getting real consent from users.”
— Jordan Smith, Founder and CEO, mymodel
That’s a good filter for any brand’s privacy policy, honestly. If your grandmother can’t understand what she’s agreeing to, it’s not really consent. It’s a legal formality.
Why Is Cookie Data So Unreliable for Marketers?
This is the part of the conversation that really got me. Jordan pointed out that even when people click “accept all” on cookies, the signal brands get back is weak. Scroll depth and time on page get correlated to purchase intent, but that correlation falls apart under scrutiny.
“A stat just came out showing that cookie data is only 47% accurate in predicting someone’s gender. You would be better off flipping a coin in determining someone’s gender than relying on cookie data.”
— Jordan Smith, Founder and CEO, mymodel
I’ve seen versions of this problem for eleven years running a data business. Early on, we’d go to agencies and show them how bloated their “in-market” audiences were. I remember telling one agency their car-buyer audience was 50 million people, when there are maybe a million car buyers in the market at any given time.
“There’s not 50 million people that are interested in buying a car right now… you’re wasting so much money trying to reach this audience.”
— David Finkelstein, Founder and CEO, BDEX
They didn’t want a smaller, accurate audience. They wanted to spend their budget, because an unspent budget meant a cut budget next year. That’s a broken incentive structure, and it’s a big part of why bad data has survived this long in the industry. Nobody was rewarded for fixing it until margins got tight enough to force the issue.
How Is mymodel Building a Consent-First Data Marketplace?
mymodel’s model flips the exchange. Instead of a brand inferring interest from behavior, a consumer builds a profile of their actual preferences and gets compensated when a brand wants access to it.
“Believe it or not, a study came out earlier this year that showed that over 99% of Americans would share data for incentives.”
— Jordan Smith, Founder and CEO, mymodel
But Jordan was clear that the incentive has to be real, not a token gesture. She’s not impressed by competitors offering a 15% discount in exchange for a consumer’s banking and social security information.
“I’d rather have cash in my hands… asking someone to give up all of this vulnerable personal information for a 20% discount or for three bucks, I don’t think that’s going to be enough to incentivize the average consumer to part with something that is very personal.”
— Jordan Smith, Founder and CEO, mymodel
That distinction matters. A discount code feels like a consolation prize. Cash feels like a transaction between equals.
What Should Marketers Actually Be Asking About Their Data?
We got into this because Jordan raised a good point: a lot of marketers don’t know what questions to ask their data vendors in the first place. From our side, timeliness is the one that trips people up most.
“We’ve tested data sets where people send it to us and it’s tied to a mobile ID from seven years ago. Guess what? It’s not around anymore. Timeliness is so important.”
— David Finkelstein, Founder and CEO, BDEX
Eleven years ago, “real time” data meant 30, 60, or 90 days old. That’s not real time. That’s stale by the time it lands in a campaign. If someone was shopping for a car a month ago, that signal is dead today. Beyond timeliness, Jordan flagged verification (is this a real person, not a bot) and actionability (how fast can raw data become an actual insight) as the other two pillars marketers should be pushing their vendors on.
Where Is Consumer Data Headed Next?
Jordan predicted that data transactions between businesses will start looking a lot more like programmatic advertising: automated, consent-gated, and fast. Right now, most B2B data deals are still manual handshake agreements, which doesn’t scale and doesn’t hold up well legally.
She also brought up the Reddit lawsuit against Perplexity, which is a preview of the fights coming over what counts as “public” data and who actually owns it once an AI model has trained on it.
“This is also a problem in B2B data transactions… this lack of clarity of what is public information, what information can an LLM be trained on, what is the consent infrastructure there.”
— Jordan Smith, Founder and CEO, mymodel
I don’t think this gets resolved cleanly anytime soon. But the direction is clear: the businesses that build consent into their data supply now, instead of bolting it on after a lawsuit, are going to be the ones still standing when the regulation catches up.
The Tools Behind the Work
Jordan runs on Otter AI for call recording and follow-up, Attio for CRM (her pick for early-stage startups), and Apollo for sales outreach and prospecting.
The Bigger Picture
The theme running through this whole conversation is trust. Not as a marketing buzzword, but as the actual mechanism that determines whether a consumer will ever hand over accurate data about themselves. Cookies never had to earn trust: they were just there, quietly collecting weak signals nobody consented to in any meaningful way. That era is ending, whether through regulation, cookie deprecation, or just consumers getting smarter about what they’re worth. Companies like mymodel are betting that the winners in the next phase of marketing will be the ones who ask, pay, and actually listen, not the ones who guess.
Connect with Jordan Smith on LinkedIn or check out mymodel.
This article was adapted from an episode of Deconstructing Data, BDEX’s weekly podcast on data-driven marketing. Tune in live every Thursday at 4:15 PM Eastern on LinkedIn.
About BDEX: For companies that need clean identity data to power their products, BDEX offers unmatched quality and execution. Visit bdex.com and click “Talk to an Expert” to get started.
Video
Watch the full episode on YouTube: From Cookies to Consent: Reclaiming Data Power in the New Marketing Economy
