Case study  ·  Product org design  ·  July 2026

Product Owner to Product Manager: what rebuilding the ladder twice taught me

Two product org rebuilds, five years apart, both starting from the same title change. What building the career ladder taught me about designing for the org you're becoming, not the one you have.

Twice in five years I've inherited the same job under a different name: turn "Product Owner" into "Product Manager" and make the new title mean something. Iress in 2021, a scaling multi-product business with 30+ product folks. Tyro in 2025, a payments and banking fintech with a product team under 10. Same transition, opposite starting conditions, and the frameworks we built couldn't have looked more different.

Option 01

Iress, 2021: designing for the org you already have

30+ PMs spread across a multi-product portfolio need one shared language, not a dozen local ones. Growing@Iress deliberately built something light: a five-rung, dual-track ladder (people track and practice track) from Analyst through to Head of Product, run through self-assessment paired with a People Leader, reviewed at least once a year. One line does a lot of work in that deck: "the capabilities framework is NOT a performance benchmarking tool." At that scale, nuance had to live in the coaching relationship, not in the document. The framework's job was consistency. The manager's job was judgement.

Iress career path (2021)

TrackProgression
Practice focusAnalyst → Associate Product Manager → Product Manager → Senior Product Manager
People focusProduct Lead (leads a group of APMs/PMs) → Head of (leader of leaders, portfolio or region) → Global Leadership (Global Heads of, CPO)

Movement wasn't strictly one-directional. A Senior Product Manager could step into Product Lead or, in some cases, straight into a Head of role, depending on where their strengths actually sat. The two tracks existed so craft and people leadership weren't forced into a single, one-size ladder.

Option 02

Tyro, 2025: designing for the org you're building toward

Sub-10 PMs, and on paper, a rubric that looks built for a company ten times that size: a near-full Google-style ladder, seven IC levels from APM through Distinguished PM, mirrored manager titles, four craft pillars including End-to-End Ownership that folds in GTM, risk, and banking-rule obligations directly into the PM role. The detail runs deep: audience size you can credibly present to, whether founders trust your judgement unprompted, whether other PMs cite your strategy docs as reference cases. That looks like overkill for ten people. It wasn't, because the team wasn't staying at ten. The mandate from day one was to scale the product org, and it did, to 30+. The rubric was built for the destination, not the departure gate.

Tyro IC ladder (2025)

LevelIC titleManager-track equivalent
1Associate Product Manager (APM)n/a
2Product Managern/a
3Senior Product Managern/a
4Staff Product Managern/a
5Principal Product ManagerGroup Product Manager
6Senior Principal Product ManagerSenior Group Product Manager
7Distinguished Product ManagerHead of Product

Seven levels for a team that started under ten people. The manager track only branches off from level five onward, which was deliberate: below that, the craft ladder and the leadership ladder are the same conversation.

The pattern underneath both isn't "big org, light rubric; small org, heavy rubric." It's that a framework can describe two different things, and confusing them is the actual failure mode. Iress needed a document that described where the org already was, so 30+ people could have the same growth conversation in the same language. Tyro needed a document that described where the org was headed, so a team of ten wouldn't outgrow their own framework before the company caught up to them. Read backwards, either choice looks wrong. Read against intent, both were right for the job they were doing.

The part neither deck shows: in both rebuilds, the actual work wasn't picking rungs, it was mapping people. Before either framework meant anything, I had to understand each PM's own ambition and where it genuinely lined up against the craft the rubric described. Some wanted the title without the stretch behind it. Some had the craft years ahead of whatever title they'd been given. A rubric only becomes useful once it's translated for one specific person's next conversation. Read cold, it's just an org chart with adjectives.

What I actually think

A role title looks like an HR artefact. Functionally, it's identity. People carry it into how they introduce themselves, how they negotiate, how they picture their own next two years. Get the framework wrong for the stage you're actually at, and you're not making an admin error. You're telling someone their growth is further away than they'd hoped, or so undefined it doesn't count as real. Both rebuilds taught me the same thing from opposite directions: the document buys you consistency and a shared language. Retention and trust are still earned one conversation at a time, with the actual human standing on the rung in front of you, not the rung the document assumed they'd be on.

A rubric describes a destination. Whether someone gets there is still a conversation, not a document.

Where this is heading

Both rebuilds happened before agentic AI became part of the everyday product toolkit. Looking back, the dual-track structure matters more now, not less. As AI takes on more of the mechanical craft work (drafting specs, synthesising research, running the first pass of an analysis), the distinction between someone who wants to go deep on practice and someone who wants to lead people gets sharper, not blurrier. Understanding what an individual is actually trying to learn and achieve doesn't get automated away. It becomes the harder, more valuable half of the job.

What I haven't fully worked through yet, and what belongs in its own case study, is how AI agents actually sit inside a product organisation. Not as a tool a PM uses, but as something closer to a team member with its own place in how the org is designed. What does an AI-native product team's structure look like, and does a rubric like either of these still hold once part of the craft is being done by an agent rather than a person? That's the next one.

Emily K Chen
Emily K Chen VP of Product  ·  ~7 min read
← All work