Jefferies Global Healthcare Conference, November 17, 2026
Illumina's Ankur Dhingra told the Jefferies Global Healthcare Conference that clinical demand is driving NovaSeq X placements at a pace the company has not seen since launch. He said Illumina has placed more instruments in the last four quarters than it did in the first year the platform was on the market. He tied that to underlying clinical demand and to the company laying out a framework for what NovaSeq X can deliver in throughput and value over an extended period. In his words, this is allowing customers "to make a lot of long-term bets on the platform, especially around the clinical validation."
Dhingra described Q3 as a good step in the return-to-growth framework Illumina set out in 2024, which targets a return to high single-digit growth. He said 2026 is running at steady mid-single-digit growth across the quarters, with the Q4 guide consistent with that. Operating margin was 25% in Q3 and is guided to roughly 25% in Q4. He called a 25% second-half margin a milestone on the path to 500 basis points of margin expansion in the core businesses.
AI Companies: Large Error Bars, Small Dollars So Far
On the question of AI companies emerging as customers, Dhingra was candid about the uncertainty. He described the discussions as "excellent" but said "the error bars in the discussions we're having with most of them are really large." The open questions, he said, include how much data a frontier AI model needs to produce a step change in the understanding of biology, and what the AI companies need to generate themselves versus source elsewhere. He said Illumina will be open about it when those discussions come to fruition.
What is visible today is a group of smaller "neo AI companies" setting up laboratories to generate data. Some are buying instruments, some are asking Illumina to generate data for them, and some are buying into its datasets. For a business of Illumina's size, he said, "that usually turns out to be small dollars. It doesn't move the meter yet." Within the 1 to 2 points of additional growth that Illumina has framed for 2027 from new products, he said BioInsight and AI are expected to be one of the bigger contributors.
Billion Cell Atlas and the Case for Pharma Paying for Data
Pressed on why a build-it-and-they-will-come data model should work this time, given earlier efforts such as Celera, Dhingra said AI tools now do pattern recognition and connection "much, much better than anything we had even 2 years ago or 3 years ago." He said they are significantly better at filling gaps and linking new data to historical data. He pointed to Illumina's own SpliceAI work, which has covered another 10 points of genome coverage in uncoded regions and variants of unknown significance. He called the addition of AI on top of that "really powerful."
Dhingra said Illumina invested ahead of the market in the Billion Cell Atlas but still makes money on it under its pharma agreements. He said the company is iterating with customers on data structure as it goes. After the first 250 million to 300 million cells were delivered and analyzed, customers came back asking for different cell types. He said Illumina sees demand and signals from the data, and sees pharma deeply engaged.
He expects pharma to become a larger customer group than it has been historically, with Illumina working more closely with the industry. He cited two enablers: AI, and the breadth of the multiomic portfolio, which lets pharma characterize samples "in a unified structured way at a level of depth that really wasn't possible before." Spatial was launched a few months ago. Dhingra said people who see the combination of spatial, whole genome and transcriptome data say they "wish I could have been doing this for 10 years."
Multiomics Through a BioInsight Lens
Dhingra said he now applies a BioInsight lens to all multiomics. Illumina can choose to sell kits, as it is doing in single cell, or in specific disease pathways or organs it may lead with data generation and monetize through the data or associated models. Through that lens, he named three combinations as the most interesting: proteomics with spatial, whole genome with proteomics, and methylation in specific disease pathways. He summarized the approach as "generally a play of whole genome plus something based on different kind of disease pathways."
Clinical Pull-Through and the Conversion Tail
Dhingra addressed the relationship between placements and consumables, noting that consumables growth is being driven by clinical and by the X platform. Clinical customers take longer to reach peak pull-through than research customers because they run validation first. Much of their pull-through also depends on the ramp of the on-market tests or clinical trials for which they are adding capacity. Because the above-average placements have been largely clinical, he said, that ramp is still playing out.
On U.S. clinical consumables growth of over 20%, he said the core therapy selection business has held up well, and MRD and the upcoming MCED are layering on additional growth. He said Illumina's ability to sustain the growth rate depends on its largest customers executing against their publicly stated ambitions. Europe and the rest of the world have not yet begun to show a tailwind from MRD or early detection. Dhingra noted that clinical adoption historically starts in the U.S., then moves to Europe and wealthier Asian countries as evidence accumulates....