Fixing pre-launch commercial ops mistakes (summarized text version)
Rebuilding commercial ops inside a live startup is harder than building from scratch.
Lori makes a distinction worth sitting with: There's building commercial ops infrastructure from scratch, and there's stepping into a startup that has already launched and finding the infrastructure needs to be dismantled and rebuilt. She's done both. The challenge in the second scenario is you're diagnosing while the business is running.
Her framing around speed to patient isn't aspirational language. It's the practical test she applies to every infrastructure decision. Does this get product to patient faster, or does it slow that down?
In CGT and rare disease specifically, she points to the logistics chain from HCP to site of care as the operational variable that gets underestimated. Long selling cycles, complex handoffs, and a very narrow margin for error on execution. Her observation is that companies often underestimate how much time establishing the right operating model actually takes, and that delay creates friction between corporate strategy and what the business needs to do to execute.
Small biotechs buying the wrong secondary data lose the budget and the launch window.
Lori is clear that she's speaking from small to mid-size biotech, not large pharma. In that context, secondary data procurement is one of the biggest line items in a Comm Ops budget, and the wrong call there is costly in two ways: the spend itself and the absence of useful insight from it.
Her reference window is T-24 to T-12 months pre-launch. The point isn't just timing, it's that the right data set is specific to the product and the market, and that clarity requires early work. Data informs go-to-market strategy, field force sizing, territory alignment, and more. Getting anchored to the wrong data set ripples outward.
When she talks about corporate strategy versus business strategy tension, she locates it in data procurement decisions. That tension, in her experience, is a signal that operating model alignment hasn't actually been achieved yet.
Marketing has a strategy, Ops has no infrastructure to run it. T-18 months is where that gap opens.
The specific pattern Lori describes: Marketing is fully staffed, Market Access has its strategy in place, and Ops and IT are brought in after the fact with no infrastructure ready to operationalize what the other teams have built. She calls this out as a timing problem, not a capability problem.
Her view is that getting the right Ops leader in at T-18 to T-12 months changes what's possible. Not because earlier is always better in the abstract, but because the operationalization work takes time that most organizations don't budget for.
On market access, she's emphatic: "You cannot get that wrong." The business model in market access is foundational, and she puts it alongside Ops as a function where the right expertise, brought in at the right time, is what the launch is built on.
Network hires narrow the pool. Startup launches need people who have built this before, under the same constraints.
Lori names something that's easy to overlook. Leaders hire from their networks, which makes sense under time pressure. The tradeoff is a narrower candidate pool and, sometimes, people who are strong in function but haven't operated inside a startup launch environment with resource constraints. Those are different jobs.
Her specific framing: you need people who understand what the operating model should look like and have done it before in this context. In a startup, there isn't a long runway to course correct at the senior level.
AI is producing real value in MLR, competitive intelligence and patient ID. The board pressure is creating commitments the infrastructure can't support yet.
Lori's take on AI is balanced and specific. She's not dismissing it. She points to MLR/PRC process efficiency, competitive intelligence synthesis, and rare disease patient identification as places where it's working or has real potential. Machine learning for patient identification in rare disease, she notes, has been in practice for decades.
What she's flagging is the gap between board and investor enthusiasm and organizational readiness to operationalize AI broadly. That pressure is producing commitments that outpace infrastructure, and in her experience it's also creating contractual complications when vendor AI capabilities overlap with what a company's own integrated data warehouse can do.
Her position on talent for AI strategy: you need someone who understands life sciences and pharma, not just AI. Domain fluency isn't secondary to technical expertise. She's seen the outside-industry-expert approach and doesn't recommend it.
Fragmented data sources don't fix themselves. Lori's move was to name the problem to the CEO and build workstreams around it.
At Genzyme, Lori walked into fragmented, independent data sources and no functioning data warehouse. Her approach was to name the limitations explicitly, including to the CEO, define what good needed to look like, and build workstreams with clear ownership and timelines. She describes it as bringing everyone along, not just the technical team.
Her closing point is personal and direct: she reaches out to her network when she has blind spots. Not as a last resort, but as a deliberate practice. Her observation over ten years is that the wrong infrastructure goes in when organizations rely entirely on instinct and internal perspective. The cost shows up in patient access, which is where it matters most.