Work
Building a research-led intelligence product for a fast-moving market
A new B2B media and events brand was trying to build authority with mid-market leaders in one of the noisiest markets imaginable. We helped create a weekly intelligence product that combined operator interviews, original research, events, external evidence and market monitoring to explain what was changing in AI and what leaders could actually do about it.
The need
Give the audience something more useful than another AI news roundup
Mid-market leaders were already surrounded by AI launches, vendor claims, headlines and predictions.
The client wanted a different position: practical, evidence-led guidance for leaders with less time and fewer specialist resources than large enterprises.
Each issue needed to answer the questions behind the most important stories: Why does this matter? What happens when organizations actually try it? What breaks? And what should a leader do next?
The starting point
Useful content existed, but the editorial product needed a defined center of gravity
The client already had access to news, research, events, advisors and a growing network of AI operators.
Early newsletter formats brought many of those ingredients together, but risked becoming too long and diffuse, mixing company updates, playbooks, news, research and product coverage without a clear hierarchy.
At the same time, the wider relationship was producing increasingly valuable evidence: interviews with leaders implementing AI, original survey research into the UK mid-market and conversations taking place at events and roundtables.
The opportunity was to turn those separate inputs into a repeatable editorial system rather than fill a weekly publishing schedule.
What we did
Built the weekly product around originated evidence and practical interpretation
We helped shape the editorial position, recurring structure and production process, alongside a rolling research and interview program.
Original material moved to the center of the newsletter. The lead story was typically built from an operator or expert interview, proprietary research or another firsthand source rather than a rewrite of the week's news.
A separate research section allowed us to go back into the evidence. Sometimes that meant recutting the client's own survey data around a question raised by an interview. Sometimes it meant returning to earlier interviews and looking for a pattern that had only become visible after months of conversations. At other times we interrogated an external study, including its methodology and limitations, before translating it into practical implications for mid-market firms.
That produced very different kinds of stories from conventional AI coverage.
One issue paired an interview on unapproved AI use with the client's survey evidence to examine whether higher levels of shadow AI were associated with more reported failures.
Another revisited a growing archive of operator interviews. Six conversations from a corpus of 19 were brought together around one question: AI was making individual tasks faster, but was the organization actually becoming more productive?
In another case, an operator explained an AI sales workflow that initially looked successful because it generated far more trial bookings. Conversion then fell and the total number of paying customers did not improve. The article followed what the team changed and why rather than turning the project into a simple success story.
Events fed the same system. Demonstrations and roundtable discussions generated follow-up analysis, while speakers were interviewed in greater depth so their experience could reach people who had not been in the room.
Product monitoring provided another recurring layer. Emerging tools were assessed around the problems they solved for mid-market businesses, how accessible they were and how strong the evidence behind vendor claims appeared.
Behind the weekly output sat a rolling editorial calendar and defined production workflow, allowing the subject matter to change quickly without losing the same practical editorial standard.
What they got
A weekly audience product backed by an expanding body of proprietary intelligence
The newsletter became the regular expression of a wider research and content program rather than a standalone publication.
Original surveys supplied data that could be revisited after launch, while interviews created both immediate articles and a growing library of operator experience. Events produced questions and cases for subsequent coverage. External research added evidence where it held up, while product monitoring kept the work connected to what leaders were being asked to buy and use.
The same broader program also produced playbooks, research reports, role-specific analysis and material for events and other channels.
By mid-2026, the weekly newsletter had more than 2,500 subscribers.
Why it mattered
The client could stay current without becoming reactive
A fast-moving market normally forces publishers toward speed at the expense of depth. The system created another option.
New developments could be interpreted through evidence the client already owned, people it had already spoken to and questions emerging from its community and events.
That also made the underlying research work harder. A survey finding or interview did not have to peak once at publication. It could be returned to as the market changed, combined with new evidence and used to answer a different question months later.
The result was a recognizable editorial position built around practical AI adoption rather than the volume of AI news.