one plan, every market and milestoneNames anonymized
The Problem
A pharmaceutical launch is not one project. It is dozens of them running in parallel — regulatory submissions per country, manufacturing and supply readiness, market access and pricing, field-force enablement, medical affairs, promotional review — each on its own clock, in its own market, with its own dependencies.
The planning artefacts that hold this together are usually spreadsheets and slide decks. They are built from scratch for every launch, they go stale between steering meetings, and they make the one question leadership actually asks — are we ready, and where are we not? — expensive to answer.
Every launch was being planned as if it were the first one. The hard-won structure of the last launch never carried forward.
The Approach
I designed the platform around a single idea: a launch plan should start from a position, not a blank page. The system treats modality — oncology biologic, CNS small molecule, rare disease, vaccine — as a first-class concept. Modality selects a baseline plan and a set of country playbooks, so a new launch begins with the workstreams, milestones, and dependencies that its kind of product actually needs.
From there the plan is one object viewed many ways: a task hierarchy by workstream or country, a Gantt across the full timeline, a market-by-market readiness view organised into launch waves, a risk register, and a set of phase gates that evaluate their own pre-conditions.
On evidence: I designed against Trinity’s launch consultants rather than the global launch leads who would run a plan day to day. They know launch planning intimately — but they weren’t the people who would have to keep twelve country rows current between steering meetings. That gap is the one that worries me most about this product, and the closing section says why.
Evidence base — internal SME review · domain-expert working sessions · stakeholder walkthroughs. No direct end-user testing.
1 / 5One plan, many views
Key Decisions
Modality as a first-class concept
An oncology biologic and a vaccine do not launch the same way. Making modality a structural choice — rather than a label — lets the system seed the right baseline plan and country playbooks, so teams start from accumulated practice instead of an empty grid.
One plan, many views
Task hierarchy, Gantt, markets, RAID, and dashboards are views onto the same object, not separate trackers. A date moved in one place is moved everywhere — which is what kept the old spreadsheet-and-deck workflow permanently out of sync.
Markets organised as launch waves
Countries are grouped into sequenced waves rather than listed flat. Each market carries its own lead, submission state, next milestone, and readiness score — so the sequencing decision and the readiness picture are the same screen.
Phase gates that evaluate themselves
Go / no-go reviews check their own pre-conditions against live plan data. The gate stops being a meeting where status is assembled and becomes a decision where status is already known.
Risk surfaced at the top, not buried in a register
Flags, watch items, and critical counts sit in the plan header with their movement since last review. The RAID register holds the detail; the header answers whether anything changed.
Outcome
12countries sequenced into three launch waves in a single plan view
4modalities with their own baseline plans and country playbooks
Launch Accelerator turns launch planning from an artefact teams rebuild each time into a system that carries structure forward — where the question "are we ready, and where are we not?" is answered by the plan itself rather than assembled for a meeting.
What I’d Validate Next
The failure mode for a planning system isn’t that people reject it — it’s that they keep it open in one tab and run the launch in a spreadsheet anyway. These are the three signals I’d watch for that.
Does the modality baseline survive contact?
The core bet is that a seeded plan beats a blank one. I’d measure how much of the baseline gets deleted in the first week. Heavy gutting would mean the templates encode Trinity’s practice rather than the client’s — useful to know before shipping a fifth modality.
Do the phase gates actually get trusted?
A gate that evaluates its own pre-conditions only pays off if leadership stops assembling a deck. The honest measure isn’t whether the gate renders — it’s whether a pre-read still gets built by hand the night before. I’d go looking for that deck.
Who keeps the market rows current?
The old spreadsheet failed because it went stale between steering meetings, and nothing in this design guarantees a country lead updates their row. I’d track update recency per market before claiming the staleness problem is solved rather than relocated.
