How to overcome beige: braver, bolder marketing in the age of AI slop

Hattie Matthews, Co-Founder & CEO at Uncharted
Key points Hattie Matthews is CEO and co-founder of Uncharted, a creative agency built for a changing world. She previously helped grow Karmarama from 15 to 300 people ahead of its acquisition by Accenture in 2016, and started her career as an artist liaison and venue co-runner at Glastonbury. This is what she covered. The […]

Key points

  • Marketing is defaulting to “beige”: safe, forgettable, functional work, driven by tight budgets, risk-averse stakeholders, and a flood of AI-generated content. One industry report found 57% of UK ads make people feel nothing at all
  • Emotional campaigns outperform rational ones significantly (31% profitability uplift versus 15%), and campaigns that try to do both tend to land in the middle
  • Hattie’s five-point framework to de-risk braver work: seek human emotion, avoid algorithmic bias through diverse teams, use rapid prototyping, pre-test before spending on media, and manage stakeholders deliberately rather than dropping a finished idea on them
  • AI and data tools codify the past, they don’t invent the future. Diverse human judgement is what actually produces distinctive work
  • Pre-testing shifts a conversation from gut feel to something you can back with evidence, but it should inform a decision, not replace it entirely
  • Bringing stakeholders along the journey, rather than presenting a finished idea for sign-off, reduces fear and makes brave work feel like the safer choice, not the riskier one

Table of Contents


Hattie Matthews is CEO and co-founder of Uncharted, a creative agency built for a changing world. She previously helped grow Karmarama from 15 to 300 people ahead of its acquisition by Accenture in 2016, and started her career as an artist liaison and venue co-runner at Glastonbury. This is what she covered.

The problem with beige

Hattie opened with the scale of the challenge: digital content volume is forecast to grow 10x in the next few years, with 80-90% of that generated by AI. Against that backdrop, one industry study found 57% of UK ads make people feel nothing at all, and separate research found 29% of marketers describe their organisations as highly risk-averse, not just cautious.

The root cause, in her view, is safety bias, our natural preference for the tried and tested, because our brains treat change as a potential loss rather than a possible gain.

“Nobody ever got fired for buying IBM… but maybe they should. Or maybe they’re just not going to get promoted.”

Seek human emotion

Campaign effectiveness data backs this up directly: purely emotional campaigns in the IPA databank deliver roughly double the profitability uplift of rational, feature-led ones, 31% versus 15%. Campaigns that try to do both tend to land in the middle rather than winning on either front.

Hattie’s own example was a year-long project with TopCashback, shifting the brand from rational messaging about cashback amounts to leaning into the emotional “double hit of joy” people described feeling. The campaign was tested regionally in Manchester and the North West before wider rollout, and delivered a 41% increase in brand awareness alongside measurable business growth.

Her practical tip: shift the creative brief itself from “what do you want the audience to know” to “what do you want them to feel.”

Avoid algorithmic bias

Because algorithms are trained on historical behaviour, leaning on them alone risks reproducing safe, backward-looking choices. Hattie’s counter to this is deliberately building diverse perspectives into the creative process, not as a box to tick, but as what she called “a primary weapon against predictable work.”

“Big data only codifies the past. They don’t invent the future. Doing that requires moral imagination, something only us humans can do.”

Uncharted was one of the first ten agencies to sign the Campaign Inclusion Charter, created by the Diversity Standards Collective, which works with agencies to check creative concepts for bias before they go out, particularly useful when an idea feels bold enough to be nerve-wracking to pitch.

Rapid prototyping

Testing unusual ideas cheaply and quickly, rather than committing to full production first, has become far more accessible than it used to be. Hattie described using AI platforms to mock up early concepts before presenting them internally, including a campaign for fintech client Insignis, built around the idea that leaving cash sitting idle was “the most dangerous thing you can sit on.” Prototyping the concept with AI and CGI let the team get buy-in from the CEO and board before committing to media spend, and the campaign significantly outperformed previous work.

Pre-test and de-risk

Pre-testing platforms let a team measure likely emotional impact and audience response before spending on media at all, shifting the conversation from instinct to evidence. Hattie mentioned tools including Dsvi (for moving image content, testing emotional spikes and likely brand impact) and Loops (for fast qualitative feedback from real audiences globally), while noting simpler routes exist too, a social media poll or a direct conversation with existing customers works just as well if budget doesn’t stretch to paid tools.

“It shifts the conversation from gut feel, ‘I like this idea’, to something with more certainty and objectivity behind it.”

One caveat she raised directly: pre-testing should inform a decision, not make it for you. Systems can be gamed once people learn how they work, and the value lies in what a team does with the results, not in treating a score as the final word.

Stakeholder nervousness featured heavily in the research Hattie referenced, 52% of clients named convincing internal stakeholders as a serious challenge. Her advice: map stakeholders’ individual goals and KPIs so you can speak their language (commercial impact for a CFO, brand health for a CMO), and bring people along incrementally rather than presenting a finished concept for a single sign-off moment.

She also shared a reframe passed on to her by a CMO: if 60% of a media budget risks being wasted because work fails to trigger any emotional response, the real financial risk sits with playing it safe, not with trying something different.

“The first lesson of branding: memorability. It’s very difficult buying something you can’t remember. The real financial risk here is being entirely forgotten.” — John Hegarty

Her example of a genuinely bold campaign going through this full process was for eSIM brand Airalo, an offbeat, humour-led ad very different from typical category messaging. Backed by pre-testing and staged stakeholder buy-in (including the founder, board, and private equity owners), it delivered 80% growth in its US test region before rolling out to multiple markets, and was shortlisted for Campaign’s Global Film of the Year.

On juniors not being invited to the table

Asked how someone earlier in their career builds stakeholder relationships when they aren’t yet included in key conversations, Hattie’s advice centred on building visibility and relationships proactively, rather than waiting until a decision is needed. She referenced something described to her as the Ben Franklin effect: asking someone senior for a small, genuine favour (advice, a coffee, their perspective on something) tends to make them more invested in your success afterwards, not less.

Does too much testing kill good ideas?

Asked whether excessive testing dilutes bold creative work, Hattie described a shift in her own view over her career. Older-style testing done right at the end of a production process (after a TV ad was already made) tended to result in ideas getting the edges knocked off, with little room left to genuinely improve the work. Testing earlier and more often in the process, she argued, lets a team push an idea further rather than merely validate or reject it.

“It’s what informs you and gives you something that you can use to make a decision… but if you’re using it to make the decision for you, unfortunately people may game it.”

Can data and creative intuition coexist?

Responding to a question about whether it’s possible to override test results and back your own instinct, Hattie was clear that data should support a recommendation, not replace judgement entirely. She pointed to the (possibly apocryphal) story of Cadbury’s Gorilla ad, reportedly resisted internally for some time before going on to become one of the most fondly remembered ads in the category, as an example of a decision that wasn’t purely data-led.

Ideas versus performance

Asked how to avoid being seen as “just” a creative thinker rather than a credible performance marketer, Hattie pushed back on the framing itself, arguing that treating brand and performance as a zero-sum trade-off is the wrong lens entirely.

“Be an ideas person. Deliver results.”