Most leaders I talk to believe they understand their audiences. They've read the research summaries and they know the headlines. They've sat in enough strategy sessions to have an opinion on what Gen Z wants.

What they've actually though built is a set of untested assumptions.

I wrote about this for Fortune because it's a problem I see at every level of the organisations with which we work. Leaders use generational labels as shortcuts to audience understanding. The shortcuts feel efficient but they are not.

Because the average represents nobody.

Yet, leaders are still leaning into these generalisations and letting them harden into assumptions. Such assumptions consciously and unconsciously shape decisions: who gets hired, which products get built and which campaigns get greenlit.

In hiring, age-based discrimination is causing leaders to overlook talent. Over a quarter of leaders say they wouldn't consider hiring a recent college graduate, citing their perceived lack of soft skills. This is shortsighted, given that Gen Z will make up nearly a third of the workforce by 2030.

In marketing, the commercial risks are just as real. Dating app Bumble's ill-judged 2024 campaign leaned into the stereotype of Zoomers as a near-celibate generation, and it drew significant backlash.

These are not one-off failures. They are what happens when assumptions go unchallenged long enough to join strategy.

The data usually exists. It just doesn't travel.

When I look at where these missteps come from inside large organisations, the problem is rarely a lack of information. It's a lack of circulation.

Insight and marketing teams typically hold layered, segmented audience data. They know that "Gen Z" contains multitudes. But that intelligence rarely makes it to the boardroom in the form it was collected. By the time it reaches the people making big calls, it's been summarised. The nuance has been removed and generalisations have replaced the previous nuance.

The fix is getting existing research to the right people before the decision is already made.

That's what we built Electric Twin to do. Not because it's an interesting technical problem — but because good leaders kept making bad decisions based on information that arrived too late, or not at all. Synthetic audiences allow leaders to interrogate specific micro-audiences before the brief is finished, rather than after the campaign has run — with a speed and precision that simply wasn't possible five years ago.

More so now than ever, leaders have the ability to predict conclusively how to close new gaps in fast-moving markets as they are no longer constrained by the speed and limitations of traditional market research methods.

What this requires from leaders.

First, change how generational shorthand is used inside your organisation. When you use stereotypes conversationally in strategy sessions, they get baked into the thinking before anyone has examined them.

Second, build the circulation routes. Granular audience intelligence needs to be shared beyond the team that collected it. All too often, leaders look at top-level summaries to make big calls. When decisions are made by those a few degrees removed from the data, assumptions creep back in and influence outcomes.

The longer leaders wait to adopt these tools, the riskier the bet becomes — while their competitors target their audiences faster and more precisely.

Generational clichés have always been a shortcut but now they're quickly becoming an expensive one.

Have a read of the full piece in Fortune here.