PW Puts the Media Transformation Summit in the Spotlight

The latest Media Transformation Summit in July 2026, cohosted by The Publishing Signal, and knkMedia attracted more than 500 registrations, approximately 200 live participants and an audience drawn from across the publishing world.

But perhaps the clearest indication of its success came a week later.

Publishers Weekly devoted a substantial article to the discussions, under the headline: “AI May Be Making Publishing Easier, But It’s Still Not Easy.”

Technology is lowering barriers. It is making content faster and cheaper to produce. It is creating new formats, new channels and new ways of reaching readers. But it is not making the difficult decisions of publishing disappear.

In many cases, it’s making those decisions even more important.

Transformation Is Not a Project

The Summit opened with Amy Jones, Chief Transformation Officer at Emerald Publishing, challenging common assumptions about digital transformation.

Transformation is not a project with a fixed beginning and end, and success can’t be measured simply by whether a new system was installed, or a workflow redesigned. The real test is whether the organization is better equipped to handle the next change.

Did the project create better information, stronger processes and more capable people? Can the organization make its next change more confidently? That is a much more demanding definition of success – and a much more useful one.

AI Can Produce More. But Should It?

Fred Zimmerman of Nimble Books and Big Five Killer supplied one of the Summit’s most deliberately provocative sessions.

Zimmerman used AI to publish 322 books in six months and reconstructed a 128-volume nineteenth-century history of the American Civil War, largely as a one-person operation. The example inevitably raised questions about quality, oversupply and AI slop, but it also demonstrated that production capacity is becoming less scarce.

The ability to create a book, audiobook, video adaptation or other content product is no longer necessarily the limiting factor. The harder questions are what should be produced, for which audience, in which format and which channel.

AI may reduce the cost of production, but it doesn’t automatically create demand.

More Formats Do Not Guarantee More Revenue

That distinction emerged in the discussion between Andrew Weinstein of Bookwire and Mohit Srivastava of Dashverse.

Falling production costs are making it increasingly viable to convert backlist titles into audiobooks, short-form video, and serialized mobile content. Backlist rights now have genuine commercial potential; however, cheaper production doesn’t mean that every title should appear in every format.

Publishers must still understand the audience and the release strategy, which rights are available, maintain quality, and determine how the resulting product will be monetized.

Technology may make adaptation possible, but publishing judgment determines whether it’s worthwhile.

Discoverability in the AI Age

Another important theme, highlighted by Sarah Arbuthnot of Supadu, Emma House and publishing analyst Thad McIlroy, concerned discoverability.

Books are increasingly being found and recommended by AI-generated answers and recommendations. That changes the importance of metadata.

Publishers need rich metadata on their own websites. Much of this information already exists inside their ONIX feeds and internal systems, but it is not always exposed and maintained publicly in a form that AI engines can locate and understand.

In the new discovery environment, good writing alone is not enough. A book also needs to be understandable to machines. It makes metadata strategic.

AI Adoption Is a Management Issue

Arantza Larrauri of De Marque brought the conversation back inside the publishing organization.

AI adoption is frequently presented as a technology decision, but in practice, it’s an organizational challenge.

Senior leadership must understand what technology can and cannot do. Middle management must be involved in redesigning workflows. Employees need clear guidance, incentives and permission to experiment. Governance and quality controls must be established.

Without that coordination, isolated pilots may generate enthusiasm but rarely produce lasting operational change.

There is also a competitive danger in using AI without a distinctive strategy. When competing publishers feed similar external information into the same tools, they may receive very similar recommendations.

The technology itself is unlikely to provide a sustainable competitive advantage. The advantage comes from proprietary knowledge, clean data, publishing expertise, integrated operations and the ability and the commitment to act on the results.

AI Does Not Have a Crystal Ball

The Summit recognized that AI could analyze patterns, identify connections and help publishers interrogate their data, but it can’t predict the next unexpected bestseller, or shifts in culture or reader behavior. Publishing remains a business built on uncertain demand, and that uncertainty cannot be engineered away.

A Practical Conversation About Change

The July Media Transformation Summit brought together nine speakers across seven sessions covering operations, technology, formats, revenue, metadata, SEO, GEO, content and audience discovery. Its success was not simply a matter of attendance.

It succeeded because the discussions avoided easy answers.

It recognized that publishers can’t ignore AI but equally can’t automate everything.

Importantly, they presented a practical path: identify specific problems, run controlled experiments, measure the results honestly and retain the knowledge created along the way.

AI may be lowering publishing’s technical and economic barriers and is moving the real bottlenecks elsewhere: to rights, quality, governance, structured data, workflow design and audience understanding.

That may make publishing faster, but it does not make publishing easy. Perhaps that is precisely why conversations such as the Media Transformation Summit are becoming so important.