We built the program to stack, not sprawl.

The core design decision was sequence. Rather than a series of standalone workshops, we built four sessions where each one ended in a specific deliverable that became the raw material for the next. The assignments were deliberately cumulative:

• Session 1 produced a creator mapping project. Each newsroom documented five to ten trusted local creators, capturing platform, follower count, audience demographics, content focus, values alignment, and a fit rating against our framework, along with a note on which communities in its coverage area were not represented.
• Session 2 turned that map into a partnership agreement. Using the creators they had surfaced, newsrooms drafted a working agreement covering the partnership's goal and model, success metrics, scope of work, a compensation model and payment milestones, and editorial standards with platform-specific disclosure.
• Session 3 turned the agreement into an outreach and editorial plan. Newsrooms built their approach to first contact and a fact-checking and review workflow, defining what they would review for and where the creator's voice remained their own.
• Session 4 turned the whole arc into a reflection. Each newsroom wrote a candid case study of where its partnership landed and what it had learned.

Because each week's homework was the input for the next week's work, progress compounded instead of resetting. A newsroom could not arrive at outreach without having done the strategic thinking that made outreach worth doing. The structure carried the rigor, so participants did not have to supply it through willpower or free time they did not have.

We forced strategy before outreach.

Most newsrooms' instinct is to start with a creator they admire and work backward. We designed the sequence to make that impossible. The Session 1 assignment required each newsroom to define what a partnership was actually for, reach, trust, access, or format, and to score creators against that goal rather than against follower count. The payoff was visible in the work: one newsroom, having clarified its goal, deliberately moved away from a curation account it had first been drawn to because it wanted a creator who could make original content. Building the discipline into the structure, rather than merely recommending it, is what made it hold.

We ran the whole program on a single framework.

Every session, assignment, and decision ran on STEPP, our proprietary framework for ethical creator partnerships, built on five commitments: Standards, Transparency, Engagement, Platform-Native content, and Public Service. The framework gave the program a shared vocabulary and a consistent standard, so the work was coherent across five very different newsrooms rather than reinvented each time. It also did real evaluative work: it gave newsrooms a defensible basis for saying no. In one case, a newsroom used the Standards test to screen out a creator with more than 400,000 followers because its editorial independence did not meet the bar, exactly the kind of disciplined decision the framework is meant to produce. And because the method is codified, it travels to the next cohort intact rather than living in any one facilitator's head.

We designed for evidence, not just delivery.

Because the real question was whether the program built capability, we built measurement into it. Every session produced an artifact we could point to, the creator maps, the agreements, the case studies, and we closed with an anonymous exit survey measuring how prepared newsrooms felt and how useful they found the method. We designed the program to prove, or disprove, its own effectiveness, and to surface exactly where the next version should improve.

Over eight weeks, five newsrooms across five states moved through the full sequence. Collectively they mapped and scored dozens of local creators against the STEPP framework, spanning the full range from neighborhood micro-creators to accounts with audiences in the millions, and they did the disciplined work of narrowing those fields to the right partner for a defined goal.

The rigor is legible in the artifacts. One newsroom built its creator map as a custom tracker with STEPP scoring columns built in, evaluating seven environment-focused creators whose reach ranged from roughly 6,600 to over 400,000 followers, and weighing editorial independence, disclosure practices, and mission alignment for each. Another mapped eight creators from a local photographer to accounts in the hundreds of thousands, used the framework to screen one out on editorial-independence grounds, and honestly flagged that its map had surfaced too few creators serving the area's immigrant communities, a finding the process was designed to make visible. A third mapped a slate spanning lifestyle, nature, food, and political voices and used it to work through a genuinely hard question for a nonpartisan newsroom: how a creator's own politics might affect its credibility. It reached a clear, defensible direction. A fourth chose two mid-sized regional creators over far larger accounts because the audience match was stronger, a direct expression of the goal-before-reach discipline the program was built to instill.

The partnerships that resulted range from in production to planned for a fall launch, an expected outcome for relationship-based work seeded in an eight-week window. The point of the engagement was never a set of finished campaigns. It was that five newsrooms with no prior infrastructure for this work ended it with vetted partners, working agreements, internal structures, and a method they could run again on their own.

Two of the five newsrooms did not complete the full arc, and we treat that as data rather than something to hide. The reasons, capacity constraints and the additional burden the program placed on a Spanish-language newsroom working in a primarily English-language format, are precisely the inputs that make the next cohort better, and they point to specific, actionable refinements rather than a flaw in the model.

What made this program work is domain-independent: a scaffolded sequence where each deliverable feeds the next, a proprietary framework that keeps the work consistent and repeatable, a discipline built into the structure rather than left to willpower, and measurement designed in from the start. Those are the components of any program meant to build durable capability in resource-constrained organizations. Journalism was the setting for this cohort. The method is the thing we designed, and it transfers.

The measure that matters most is not the campaigns this cohort produced, but the capability it left behind. Five newsrooms that started with no infrastructure for this work ended it able to run creator partnerships on their own, and the program that got them there was built to be run again. That is what we design for: not a single successful engagement, but a repeatable model that turns the same investment into capability every time it runs.

The most useful thing we did was design a program that produced capability rather than content. It would have been easy to run a series of engaging sessions on creator partnerships, collect good feedback, and call it a success. What we set out to build instead was a structure that moved newsrooms from thinking about this work to actually doing it, and that left them able to keep doing it after we were gone.

The clearest sign it worked was who these newsrooms were at the start. For most of them, this was the first time they had seriously taken on creator-partnership work at all. Eight weeks later, every newsroom that responded said it felt significantly more prepared, and each had the artifacts to prove it: a real map, a real agreement, a real partnership in motion. The program did not just inform them. It built a capability where none existed.

That is the design we can run again, refine, and scale. The framework travels, the sequence holds, and the measurement tells us honestly what to improve. A first cohort proved the model. The next one will go further.