New skills, new mindset

Every leader we spoke to arrived at the same conclusion, though from different directions: AI makes human capabilities more valuable, not less. The explanation is straightforward. When a model can draft, compile and analyse, the work that remains is precisely what it cannot do.

Kate Cox at Guesty split the future workforce into two camps: systems builders who design frameworks and automations, and tastemakers who understand quality and can see the bigger picture. Pia Hämäri at Futurice described AI as an amplifier of human expertise, most effective when it synthesises perspectives and insights into clearer outputs, but unable to replace the strategic direction that comes from the team. Julia Cames at Hays was direct about the stakes.

The practical implication is clear. Leaders who invest only in AI tools without equally investing in the people who shape, judge and apply the output will find their marketing indistinguishable from everyone else's. The technology is available to all. The thinking is not.

This was the single most repeated piece of advice across all 14 interviews, phrased differently each time but pointing in the same direction.

Julia Cames described it as finding pain points first, then seeing where AI can address them. Kate Cox advised picking a use case you can solve relatively easily, getting it live and measuring it. Elizabeth Maxson at Contentful cautioned against teams stuck in perpetual experimentation, trialling new tools without clear objectives. Her advice: clarify whether you are aiming to improve efficiency, reduce costs, increase pipeline or expand reach before selecting any tool.

Erin Pearson at Evalueserve built agents to solve a concrete problem: meeting preparation that consumed too many hours. Her team created an agent that researches attendees, pulls recent company news and earnings data, and generates tailored talking points before every meeting. Adoption was immediate because the value was obvious. Julien Harazi at Keyloop asked his team to list the repetitive tasks consuming their time before touching any new tools.

The results speak for themselves when the starting point is right. Alex Venus at Personio reduced monthly business reviews from a full week of compilation to approximately two hours by building an AI agent connected to their data infrastructure. Linda Rønningen cut comprehensive content production from two weeks to one. Amir Jirbandey estimates AI now handles 20 to 30 percent of repeatable manual tasks at Personio, freeing capacity for work where human judgement is essential. These are not theoretical projections. They are measured gains from teams that identified a specific bottleneck and solved it.

For 20 years, digital marketing rewarded deep expertise in narrow disciplines: paid search, email automation, content production, analytics. AI is reversing that trajectory. When a model can draft email sequences, generate ad variants and compile performance reports, the premium shifts to people who understand how all of it connects.

Elizabeth Maxson calls this the "full-stack marketer": someone equally comfortable with strategy and execution, who knows how to use tools to extend their reach rather than replace their thinking. Alex Venus at Personio observed that the foundational qualities remain consistent: adaptability, intellectual curiosity and the capacity to work across domains.

Hiring is shifting accordingly. Kate Cox hired someone who had been running an entire marketing department alone using AI, the kind of generalist that lean start-up teams produce. Elizabeth Maxson now prioritises soft skills alongside technical fluency, arguing that curiosity, empathy and judgement are precisely what AI cannot replicate. She is not alone in that emphasis. Five of 14 leaders independently cited curiosity as the defining trait for success in an AI environment, without being prompted. They arrived at the same word from entirely different contexts: hiring decisions, team culture, personal development and leadership capability. When that many leaders converge unprompted, it is worth paying attention.

Yet the survey data tells a more cautious story. When we asked leaders to rate their confidence in their team's AI readiness for the next one to three years, the average was 6.7 out of 10. Personal readiness to lead through the change scored 8.0. That gap between individual conviction and team capability is where the real work lies, and it will not close on its own.

The traditional tension between marketing and sales has not disappeared, but AI is changing the terms of the conversation. When both functions access the same intelligence — intent signals, conversation analysis, account engagement data — arguments over lead quality become less about opinion and more about evidence.

Rebecca Angus at GBG provided the most concrete example. Before implementing intent data tools, her ABM programme influenced approximately £1.3 million of pipeline. That figure grew to £8.5 million in EMEA, built on closer partnerships between marketing and SDR teams where both sides could see which accounts were actively in-market. A cross-sell campaign running for just two and a half weeks generated 10 qualified meetings with existing customers.

Julia Cames described the shift in practical terms: when a high-demand candidate engages with content on the Hays platform, consultants receive context about what that individual has consumed, turning outreach from a cold call into a conversation grounded in the candidate's interests. Raine Pell at Sopra Steria cautioned that faster delivery creates a risk of unsustainable demand. Sales teams may push for more campaigns simply because the capacity appears to exist.

The survey paints a nuanced picture. Twelve of 14 respondents said marketing and sales are aligned on what constitutes a qualified opportunity. Yet when asked how effectively they can act on buying signals and intent data, the average score was just 5.7 out of 10 — the lowest of any question in our poll. The alignment exists in principle. The operational ability to convert signals into timely commercial action is where most organisations still have ground to cover. Only half of participants said they can currently measure marketing's impact on revenue with the precision they need.

Part of this gap is vocabulary. As scoring evolves from manual lead models to AI-driven account signals, both functions need to understand what the new data means and how to act on it differently. The shift from traditional lead scoring to account-level intent scoring requires shared definitions, agreed thresholds and a common understanding of when an account is genuinely ready for a sales conversation versus simply showing early-stage interest.

AI has made content production cheaper and faster. That is simultaneously an advantage and a threat. When volume becomes easy, mediocrity becomes the default.

Kate Cox at Guesty was blunt about the early experience. As a premium provider, her team could not afford to let AI-generated content out the door without serious quality control. That meant building knowledge bases, training models on what good looks like, and accepting that the review process would take longer than before.

Alex Venus at Personio identified a shift that few organisations have fully absorbed: content creation has accelerated, but decision-making and review approvals now take longer because teams are not always confident in AI-generated output. The bottleneck has moved. It no longer sits in production. It sits in judgement. Organisations that staffed for execution speed now need to staff for editorial quality, and the skill set is different.

The leaders seeing the strongest results have accepted this trade-off. They are producing fewer, better assets and measuring success by conversion quality rather than traffic volume. Where that higher-quality traffic comes from, however, is changing too.

Fewer prospects start their research journey on Google. More are turning to ChatGPT, Claude and Gemini to evaluate vendors, compare solutions and shortlist options. Rebecca Angus observed this shift directly among her audience. The implications are practical: an LLM needs enough detail to understand and categorise your offering. If human visitors to your website struggle to grasp your value proposition, AI systems will have the same problem and you will not appear in their recommendations.

Julia Cames offered a counterpoint worth sitting with. Her team discovered that their existing SEO investment was contributing to how Google's Gemini surfaces Hays content in AI overviews. She had initially encouraged her team to reconsider their SEO-heavy approach given the changing environment. The data told a different story.

Linda Rønningen is already optimising for AI-generated search results and noted that longer-form, comprehensive content performs well in these environments, consistent with how language models process and surface information. 6sense's own data confirms this pattern: their longest-form content consistently ranks highest in AI-generated results. The logic holds. Language models favour depth and specificity, which means the investment in comprehensive, well-structured content is carrying forward into the new environment rather than being wasted.

Elizabeth Maxson at Contentful discovered something that reshaped her team's content approach entirely: 26% of website visitors were clicking the login button. More than a quarter of traffic was existing customers. That single data point changed how her team thinks about who is arriving at their site and what content they need.

Despite declining overall traffic, the visitors who reach the site now carry higher intent because they have already completed their research elsewhere. Mid and bottom-of-funnel content matters more than ever as a result. The shift from keyword optimisation to conversational search is real, but the foundation remains the same: clarity about what you do, who you serve and why it matters.

If there is one message that cuts across all 14 interviews, it is this: technology is the smaller part of the challenge.

And here is the uncomfortable corollary: AI does not level the playing field. The expectation was that it would narrow performance gaps by lifting lower performers. The evidence from these interviews points the other way. AI amplifies existing differences. People with a growth-oriented mindset extract greater value from the tools, extending their advantage. Those who hesitate fall further behind, not because the technology is difficult but because the willingness to adapt is unevenly distributed.

This makes the change management challenge more urgent, not less. Julien Harazi described the pattern most organisations will recognise: a group of natural adopters, a middle group that needs encouragement and evidence, and a smaller cohort that resists. His approach was to find champions first, let them demonstrate results, and use peer evidence to bring the middle group along. Kate Cox embedded AI ambassadors across her organisation after a centralised unit in customer support proved the model. Amir Jirbandey credited executive commitment as the catalyst at Personio, where a dedicated GTM AI council with the CRO at the helm examines the end-to-end lifecycle to identify opportunities.

Raine Pell raised a point that often goes unspoken: there is sometimes a gap between how organisations position AI to their customers and how readily they adopt it internally. Talking about AI in market-facing content while struggling to use it in corporate functions creates a credibility problem that eventually surfaces. The organisations moving fastest share a common pattern: top-down clarity about expectations, combined with bottom-up freedom to experiment.

6sense sees this pattern across its customer base: the organisations that succeed with signal-based programmes are not necessarily those with the most sophisticated tools, but those who have managed the change piece effectively. When new AI-driven processes require teams to work differently, the technology only delivers value once people are bought in and understand why it matters. That principle holds whether you are implementing a predictive intent platform, building internal AI agents, or restructuring workflows around automation.