A launch plan is built on a picture of the market. The picture gets assembled, signed off, and then defended in review meetings for the next year and a half. The market does not hold still while that happens: a competitor reads out earlier than expected, a standard of care shifts, a payer policy changes, a filing lands six months ahead of anyone's model.
Pre-launch competitive intelligence is the function that keeps the plan aimed at the market the asset will actually enter rather than the one that existed when the plan was written. That is a different job from the landscape work that supports early development, and it runs on a much harder clock.
The only question that matters
It is not "what does the competitive landscape look like". It is this: on the day we launch, what will the prescriber already have, what will the payer already cover, and what will we be able to say that the incumbents cannot?
Everything else in pre-launch CI is instrumental to answering that. Note that it is a forecast rather than a description, which means it carries uncertainty, and that uncertainty has to be stated rather than smoothed away. A landscape that presents the competitive position at launch as a settled fact is the most dangerous document a launch team can be given.
The four pre-launch deliverables
- Launch sequencing — who lands before you, who lands during your first two years, and in what order
- Anticipated labels — the claim set each competitor will plausibly be able to make, inferred from trial design
- Access starting conditions — what coverage, contracting and channel look like on day one, not today
- Surviving differentiation — the subset of your claims still unique once the above is accounted for
Working backwards from the launch date
The useful way to structure pre-launch CI is against the decisions it has to inform, not against a reporting calendar. Roughly:
| Window | Decision being made | What CI owes the team |
|---|---|---|
| T-24 to T-18 | Positioning frame, target product profile refresh, pricing research design | Competitive set defined at indication and line level; anticipated competitor labels; first sequencing view with explicit confidence |
| T-18 to T-12 | Access strategy, contracting assumptions, forecast lock | Payer starting conditions; incumbent contract exposure; erosion of claimed differentiation under each competitor scenario |
| T-12 to T-6 | Message testing, field training, congress plan | Competitor readouts and filings monitored live; objection handling built from what competitors are actually saying, not what you expect them to say |
| T-6 to T-0 | Launch execution, first-90-day response | Fast-turnaround monitoring with a defined escalation path; a pre-agreed view of which competitor moves require a plan change |
The dates matter less than the ordering. The recurring failure is that CI is commissioned at T-12, which is after positioning and access strategy have already been set, so it becomes evidence for decisions rather than an input to them.
Competitor launch sequencing
Order of entry is one of the few competitive variables with a well-documented commercial effect, and it is largely forecastable. The inputs are public: readout timing from registry estimated completion dates and sponsor guidance, filing pathway signals such as breakthrough or accelerated-approval designations, and regional regulatory lag between the US and Europe.
What makes sequencing hard is not gathering those inputs but resisting the urge to state them as a single date. A competitor's approval is a distribution, not a point. The honest form of the deliverable is a range with a stated basis — "most likely H2, could pull into H1 if the interim reads out and they file on it" — because that is the shape a launch team can actually plan against. A single date invites the plan to be built on it and then quietly invalidated.
It also pays to be specific about what "first" buys. In a crowded class with an established standard of care, arriving second with a better label in the population that matters is frequently the stronger position, and a sequencing analysis that only counts order is not telling the team anything useful.
Reading a label before it exists
The most valuable pre-launch deliverable is usually the anticipated competitor label, because almost every downstream decision — positioning, message, access strategy, field response — depends on what the competitor will be permitted to claim.
You can bound that from registrational trial design well before approval. The enrolled population sets the outer limit of the indication. The comparator determines what superiority claim is even available. The primary endpoint and the statistical testing hierarchy determine which secondary claims can be made without a multiplicity problem. A trial that tests overall survival only as a descriptive secondary is not going to produce a survival claim, whatever the press release implies.
Two cautions. First, the design gives you the ceiling, not the label — regulators routinely narrow indications relative to the trial population, particularly where a subgroup drove the effect. Second, the claim set and the commercial message are not the same thing; a competitor with a narrow label will still promote to the edge of it, and your field force will meet the message rather than the label. Track both. For an indication-level view of how competitor programmes are structured and where the claim boundaries sit, a maintained pipeline analysis is the underlying input this all rests on.
Access is a competitive variable, not a downstream one
Clinical differentiation converts into uptake only if the access position permits it. In practice this means pre-launch CI has to cover the competitive situation on the payer side with the same seriousness as the clinical one:
- Channel. Whether the incumbent sits in the pharmacy benefit or is bought and billed changes who makes the substitution decision and what leverage is available to whom.
- Incumbent contracting. An entrenched contract with volume commitments can make a payer economically uninterested in a better product, and that is a knowable constraint before you launch into it.
- Utilisation management. The step edits and prior-authorisation criteria in force for the class are usually published, and they tell you what a new entrant will have to displace.
- Provider economics where the product is administered, because reimbursement mechanics can outweigh modest clinical differences.
Teams that treat access as a function that receives the launch plan rather than shapes it tend to discover the constraint at the point where nothing can be done about it.
War-gaming, and when it is theatre
Competitive war games — internal teams playing competitor roles and responding to the plan under realistic constraints — are a reasonable way to surface reactions the plan has not accounted for. They work when the people playing a competitor have been genuinely briefed on that competitor's pipeline, cost position, contract exposure and stated strategy, and are held to it.
They are close to worthless when they are run as a facilitated workshop with no prepared intelligence behind them, because what gets produced is a room's collective assumption about competitors, dressed up as an exercise. If a war game is on the plan, the CI input to it is the work; the day itself is the presentation layer.
How pre-launch CI usually fails
- It arrives after the plan is locked. The single most common failure, and it is a commissioning problem rather than an analytical one.
- It describes today's market. A landscape of currently approved products is a description of the market you are not launching into.
- It assumes competitors behave as stated. Public guidance is a negotiating position as much as a forecast; the trial design is the more reliable signal.
- It stops at approval. The first ninety days generate more decision-relevant competitive information than the preceding two years, and most teams have no monitoring capacity left by then because the budget was spent on the pre-launch deck.
The related methods are covered separately: primary versus secondary sources for where the forecast inputs come from, and congress intelligence for the readout cycle that drives most of the sequencing revision.