Episodes

Dan Reeves: The Findable Story

Dan Reeves, Co-Founder of Subjct, on why great content now has to be built to be found, and why the human storyteller still sits at the centre of an AI search world.

Apple Podcasts Spotify
 ·  The Business of Marketing  · S5 E102  · 32 min
Read the interview

"in the last sort of 12 months particularly, has been a watershed moment."

The Conversation

More than two decades across publishing, journalism and content marketing, from selling advertising space at Centaur to co-founding an AI content platform. Reeves argues that great storytelling remains the job, but that in an AI search world content only counts if it can be found and trusted.

Reeves left university looking for a permanent role and started at the B2B publisher Centaur selling advertising space, his first entry into magazine media. He moved to EMAP in the mid-1990s, working commercially alongside the editorial teams on Q, Arena and The Face during what he calls the halcyon days of magazines, when music titles behaved like an extension of the record labels and drove cultural narratives. He later launched his first company with his business partner Ben, built on the premise of helping brands become publishers by taking editorial skills, content and networks to a suite of clients. That agency incubated a natural language processing product, Loyal AI, focused on richer archive search for publishers and funded through Google Digital News Initiative and innovate grants. Stripping everything back, Reeves and Ben then launched Subjct to serve sharp pain points around content optimisation and discovery.

In this conversation with host Justin, Reeves argues that search has moved from Google's deterministic crawl, rank and index monopoly to a probabilistic system of retrieval, synthesis and citation, and that this has reshaped both how content is discovered and how it must be produced. He believes brands had to learn to publish like publishers to earn a savvy audience, that AI can augment human work but never replace the storyteller who makes it sing, and that the whole system depends on an ecosystem of trustworthy content, ideally with publishers remunerated for what they create. He warns against AI slop, backs human-first optimisation over machine-generated volume, and insists niche communities and first-party relationships are how publishers survive and thrive. Great stories, he says, will always find a home.

Key Takeaways

  • Reeves has spent more than two decades moving from the commercial side of magazine publishing into content technology. From selling ad space at Centaur, to EMAP in the mid-1990s working alongside Q, Arena and The Face, to co-founding an agency that helped brands become publishers, then Loyal AI for archive search, and now Subjct for content optimisation, his career tracks the changing economics and mechanics of getting content made, found and paid for.
  • The web and then social media broke the copy-sales model and changed who produced content and how it reached people, and Reeves argues AI search is the next fundamental change of the same magnitude. Search has moved from Google's deterministic crawl, rank and index to a probabilistic world of query fan-outs, passage-level extraction, synthesis and citation, and he says the pace of that change in the last 12 to 18 months has been a watershed moment.
  • Reeves is emphatic that AI augments rather than authors. He warns about AI slop and insists that great storytelling comes from a human, pointing to examples where live footage is expanded with AI while a human stays at the core. Subjct, by his account, is not a platform that generates content but one that optimises human-first content so it can be discovered in both classic and AI search.
  • Trust is the load-bearing idea in his argument. Because AI systems must return answers with authority and truth, he says they depend on an ecosystem of trustworthy content, and without it the systems risk collapsing in on themselves. He hopes for a transactional model in which publishers and creators are remunerated for the content that feeds those machines.
  • On survival, Reeves says publishers are resilient and that the answer lies in owning and monetising the first-party relationship through subscriptions and recurring revenue as arbitrage ad models disappear. He believes niche communities, from Kerrang! readers to marathon runners, will only grow in importance, and that whatever the delivery mechanism, a great story still creates a multitude of opportunities.

Topics Covered

  1. 01 AI search and content discovery
  2. 02 Brands becoming publishers
  3. 03 Content optimisation at scale
  4. 04 Publisher revenue models
  5. 05 Human-first storytelling

Key Exchanges

05
01 Why is the shift in search so important, and what do these AI systems need?

So it's hugely important for these machines to ensure that there is a ecosystem of trustworthy content.

Where Google owned the deterministic crawl, rank and index monopoly, that has shifted at great speed in the last eighteen months. These machines use query fan-outs, passage-level extraction and synthesis, but they have to bring back content that answers a question and intent with truth, authority and trust. So it is hugely important for these machines to ensure there is an ecosystem of trustworthy content. I really hope we see a transactional agreement so publishers and creators are remunerated for the content they produce.

02 How can AI help journalists, creators and marketers without replacing creativity?

it's the human interaction, it's the human element that really makes that then sing.

AI slop is the risk. I genuinely believe in great storytelling, and that comes from a human. AI can assist, but it is the human interaction, the human element, that makes it sing. I have a photographer friend, now at an agency called Enmatic, who shoots live footage and then augments it with AI. The human is absolutely at the core, and it is expanded using AI.

03 You co-founded Loyal AI. Was optimisation the problem you were trying to solve?

it's great that you can produce content, but it needs to be found and discovered.

We incubated Loyal around natural language processing for richer archive search, funded through Google Digital News Initiative and innovate grants rather than equity. What we learnt brought us to Subjct. We stripped everything back because we could identify sharp pain points around content optimisation. It is great that you can produce content, but it needs to be found and discovered, and publishers were struggling to scale the optimisation needed to be found in both classic and AI search, as search moved from deterministic to probabilistic.

04 What new commercial models could sustain quality journalism?

the arbitrage ad, ad models are disappearing.

For publishers the key is driving the first-party relationship, building recurring revenue and subscriptions to replace revenue lost from copy sales. With traffic decline, the arbitrage ad models are disappearing, but new models are emerging and publishers are very resilient. They have an amazing relationship with their readers, viewers and listeners, so monetising that relationship is how publishers survive and thrive.

05 Looking forward, how will content creation change?

I wouldn't want to go ahead any more than a year, if I'm honest.

I would not want to go ahead any more than a year, it is moving so fast, and the last 12 months have been a watershed moment. But there is an innate sense of what people want to interact with, and that will drive it. How it is delivered and consumed will change, but it is about great stories and the human condition.

S5 E102Season & Episode
32 minDuration
12 Months he calls a watershed for discovery
18 Months search shifted at great speed

"We, we, we're not a platform that generates AI content."

Hear Dan on
The Business of Marketing
Season 5 Episode 102 32 min
Apple Podcasts→ Spotify→ Dan's full profile→

Keep Reading