The situation

Product Leadership knew the roadmap wasn’t landing. What they hadn’t seen was why: it was inconsistent, contested, and ownerless, because the practice underneath it didn’t exist.

• The business problem. The roadmap was a primary trust signal, and it was reactive and inconsistent. Leaders put on a brave face with external partners while scrambling behind the scenes, and it cost them countless hours a quarter re-litigating the same decisions.
• The org problem. Seven leadership groups — Kroger Precision Marketing, eCommerce, Merchandising, Digital Experience, PR & Comms, Operations, and Strategic Planning — each held a different version of the roadmap and shared no decision rights. One leader put it perfectly: “prioritization is often decided by who can speak the loudest.”
• The audience problem. Internal teams needed full detail and ownership context. External partners needed to know what’s shipping and when it affects them, without the internal detail. One artifact was never going to serve both.

What I built

The brief I was tasked with pointed at templates and cadence. I bet that fixing the stated problem would fail unless I fixed the one underneath it, so I scoped the work as an operational reset.

A broadcast is only as credible as the decisions behind it. So I built the rigor behind how product-driven work gets positioned, inside and outside the company. That opened three kinds of value at once: commercial (better monetization, more supplier trust), operational (fewer scrambles and re-litigation meetings), and cultural (product marketing that happens on purpose).

The operating model

Five layers, in order, every time

1PrioritizeHow work earns its place before anyone talks about it.

2RoadmapThe artifacts we build and maintain. Two of them. Never one with views.

3PositionWho decides what, and how the roadmap gets governed.

4BroadcastWhen, where, and how we market the work.

5ActivateWhat a launch requires, and how we close the value loop.

How I operated

The defining decision was scope discipline in two directions: I expanded scope where the evidence demanded it, and refused to expand it where ownership didn’t exist.

• I let the evidence expand the brief. A structured discovery sprint (a survey plus tailored interviews with 13 leaders across 7 groups) surfaced the same seven themes independently. Prioritization has no standard. Decision rights are unclear. Value gets lost in translation. Visibility depends on who you know. When the same root problems show up across teams that never talk to each other, the problem is systemic. That’s what earned me the right to hand back an operating model.
• I shipped the whole kit. Roughly 14 frameworks across all five layers: work definition, intake, and a tradeoff rubric; internal and external roadmap formats; a DACI decision-rights matrix and confidence definitions; broadcast deck, digest, and triage templates; a launch playbook, tiering, and FAQ. I chose a complete system over one beautifully polished artifact, because the value was in how the pieces connected.
• I pressure-tested, then deliberately stopped. I validated with the three leaders who’d have to enforce it, documented what was Locked, Drafted, and Open, and paused before operationalizing. Cadence mechanics without a dedicated owner produce a system nobody runs. So I wrote a handover guide and a role justification, and flagged the hardest unresolved problem (“altitude,” the threshold for what earns roadmap placement) as the first thing that owner should solve.

What happened

This was scoped as a discovery-and-build sprint, so the outcomes are a locked foundation, an endorsed direction, and a justified role.

• Leadership locked the foundation. The five layers, internal-first sequencing, and the two-separate-artifacts rule were endorsed by senior leadership as “the spine of everything going forward.”
• A comms task became a strategic capability. The reframe from “roadmap communication” to “product marketing operating model” stuck, repositioning the work as a commercial, operational, and cultural unlock.
• ~14 reusable frameworks, plus a validated share-out back to the original interviewees that closed the loop on their input.
• A new role, justified and pre-onboarded. A documented case for a dedicated Product Marketing Manager to own the model, with a 30/60/90 plan and prioritized open questions, so whoever takes it starts from a validated foundation instead of a blank page.

Who owned what

This one is unusual: I was the whole team. Seven leadership groups owned the inputs — their prioritization calls, their audiences, their commitments. The three enforcing leaders owned the decision to adopt. I owned discovery, synthesis, all five layers, all ~14 artifacts, and the validation. My final deliverable was the system and a role case, not a running program.

What I’d do differently

• Solve “altitude” inside the sprint. The threshold for what earns roadmap placement is the keystone everything else depends on, and I left it as the future owner’s first task. A rough v1 answer during the sprint would have made the foundation complete.
• Bank the org-change win as a real outcome. The hardest and most valuable thing I did was get leadership to back a bigger idea than the one they asked for. I let it read as a byproduct of the deliverables. It was the point.