Illumina Inc (ILMN) 2017 Q1 法說會逐字稿

Welcome to the Q1 2017 Illumina, Inc.

Earnings Conference Call.

My name is Adrienne, and I'll be your operator for today's call.

(Operator Instructions) Please note this conference is being recorded.

I'll now turn the call over to Rebecca Chambers.

Rebecca Chambers, you may begin.

Thanks, Adrienne.

Good afternoon, everyone, and welcome to our earnings call for the first quarter of fiscal year 2017.

During the call today, we will review the financial results released after the close of the market and offer commentary on our commercial activity, after which we will host a question-and-answer session.

If you have not had a chance to review the earnings release and earnings presentation, both can be found in the Investor Relations section of our website at illumina.com.

Participating for Illumina today will be Francis deSouza, President and Chief Executive Officer; Marc Stapley, Executive Vice President and Chief Administrative Officer; and Sam Samad, Chief Financial Officer.

Francis will provide a brief update on the state of our business, and Sam will review our financial results.

This call is being recorded, and the audio portion will be archived in the Investors section of our website.

It is our intent that all forward-looking statements regarding our expected financial results and commercial activity made during today's call will be protected under the Private Securities Litigation Reform Act of 1995.

Forward-looking statements are subject to risks and uncertainties.

Actual events or results may differ materially from those projected or discussed.

All forward-looking statements are based upon current available information, and Illumina assumes no obligation to update these statements.

To better understand the risks and uncertainties that could cause actual results to differ, we refer you to the documents that Illumina files with Securities and Exchange Commission, including Illumina's most recent Form 10-Q and 10-K.

With that, I will now turn the call over to Francis.

Thank you, Rebecca, and good afternoon, everyone.

Q1 was a strong start to the year.

The NovaSeq launch surpassed our expectations with more than 135 orders in Q1.

Revenue grew 5% year-over-year to $598 million, with consumables and services growth more than offsetting the expected declines in sequencing instruments related to the NovaSeq introduction.

We were very pleased with both the level of customer interest in the new platform as well as the positive feedback we received on NovaSeq's attributes.

NovaSeq was designed to be the most powerful sequencer ever made.

Its price points and throughput enable customers to increase experimental scale and complexity from panels to exomes and from exomes to genomes as well as increased sequencing depth.

Like the GA, HiSeq and HiSeq X, we are confident that NovaSeq will also demonstrate the elasticity of the genomics market and redefine its trajectory.

We were pleasantly surprised by the level of NovaSeq interest from our HiSeq X installed base as more than 1/3 of HiSeq X customers placed NovaSeq orders, many of which were multi-unit deals.

We also saw the start of the HiSeq replacement cycle as the majority of our customers that ordered NovaSeqs were legacy HiSeq customers looking to refresh their fleets and run at the more economic price per gigabase.

With an active HiSeq user base of approximately 800 customers, the replacement of older-generation instruments by NovaSeq is just beginning.

Since launch, we have heard from customers the desire to address each lane of the flow cell independently to more easily allow them to load different library pools, applications and samples.

Yesterday, we unveiled a new development program that will bring to market an accessory device and associated consumables to provide this functionality.

Compatible with all NovaSeq flow cells, this updated workflow will allow customers to load different project, library and applications in each lane, increasing their multiplexing capabilities and reducing the DNA input needed for each run.

We expect to begin shipping the consumables and device for this new workflow in the fourth quarter, shortly after the launch of the S4 flow cell.

The response we have seen since NovaSeq launched reinforces our view that as far out as we can see, there is an insatiable demand for high-throughput sequencing.

We're actively scaling our manufacturing capabilities to meet anticipated demand.

During the quarter, we manufactured and installed more than 25 instruments, slightly exceeding the high end of our expected capacity.

Importantly, machines installed during the quarter are already producing high-quality data in customers' hands.

As expected, we will remain capacity-constrained in Q2 and are tracking to exit Q3 with NovaSeq manufacturing scaled to meet quarterly demand.

Focusing now on the rest of our high-throughput portfolio.

We were pleasantly surprised at the level of demand for HiSeq and HiSeq X in the first quarter.

As we've previously shared, we had expected to ship approximately 10 instruments in Q1, which we surpassed as select clinical customers required additional HiSeq capacity ahead of the broad availability of NovaSeq.

Additionally, a few HiSeq X customers added capacity for ongoing projects, and customers in China took shipment of previously ordered instruments at a level that exceeded our expectations as the HiSeq X remains a powerful competitive product in the regional market.

Going forward, our outlook for this family of instruments has not changed and remains muted as we expect a vast majority of customers will wait to access the NovaSeq platform.

Our benchtop portfolio was stable in the first quarter as the trends we shared on previous calls remained steady.

NextSeq demand continued to be generated by commercial customers, including our partners in China, Berry and Annoroad.

Annoroad received Chinese FDA approval for their NIPT assay on the NextSeq AR 550 during the quarter, further strengthening the NextSeq outlook in the region.

New to sequencing customers continue to bolster MiniSeq and MiSeq placements as they accounted for approximately 60% of quarterly shipments.

Additionally, we are adding to our benchtop specialist team to generate incremental demand for our low-end portfolio and library prep solutions as well as investing further in lead generation and inside sales.

In Q1, microarray revenue, including services, grew 17% year-over-year to just over $100 million, driven primarily by consumer demand which increased over 40% versus the prior year.

Total oncology testing demand also remained robust as shipments to this market segment grew 20% versus the prior year.

This strong result was driven by continued growth in commercial molecular diagnostic customers as well as the translational liquid biopsy segment.

Moving now to reproductive and genetic health.

We will continue to see positive NIPT reimbursement trends in the United States.

CAPS, the Coalition for Access to Prenatal Screening, contacted the vast majority of Medicaid offices in Q1 to update their respective fee schedules to be in line with the reimbursement level set by the 2017 clinical lab fee schedule.

In response, more than 20% of states contacted are reviewing their reimbursement practices, and 4 states are in the process of increasing their fee schedules.

Additionally, outside the U.S., we are also making progress in driving NIPT adoption.

With the recent launch of our VeriSeq NIPT solution, which includes a CE-IVD marked library prep and analysis software, clinical labs will now have access to highly accurate, fast and reliable software for analysis of NIPT in-house.

Reimbursement in the region is progressing.

In addition to the Netherlands, Denmark began covering the test during Q1, and a coverage decision is expected shortly in France.

Over the last few years, we have also been focused on ensuring pediatric patients with suspected genetic conditions of access to sequencing, which has been shown to dramatically increase the diagnostic yield in these cases.

There are 15 million children in the United States living with a rare and undiagnosed disease that historically have not had access to sequencing.

This market continues to progress well as recent updates made to certain payer policies now reimburse whole exome sequencing for pediatric patients with suspected genetic conditions.

Today, coverage stands at 50 million lives, a vast improvement made in just a few months.

In closing, I am pleased with our first quarter results, which have set the stage for a strong 2017.

I will now turn the call over to Sam for a detailed overview of our first quarter results.

Thanks, Francis.

As Francis mentioned, Q1 revenue grew 5% year-over-year to $598 million, slightly exceeding our quarterly guidance.

Geographically, revenue in the Americas grew 4%, while we saw a decline of 2% in Europe.

Asia Pacific grew revenue 15% versus the prior year.

Continued weakness in Japan's genomics funding was more than offset by China shipments, which grew more than 35% due to clinical and high-throughput instrument demand.

Given the transition in our high-throughput portfolio, instrument revenue declined 15% year-over-year to $100 million, in line with our expectation.

First quarter consumable revenue represented 65% of total revenue or $387 million, an increase of 7% compared to the first quarter of 2016.

Sequencing consumable revenue also grew 7% year-over-year to approximately $320 million as a result of growth in our installed base.

Utilization trends across the benchtop portfolio were steady as pull-through came in within each instrument's respective guidance range.

HiSeq and HiSeq X consumable revenue was slightly below our forecast as customers ramped down their reagent orders at a rate that slightly exceeded our expectation, instead using inventory on hand given the upcoming integration of NovaSeq into their operations.

We believe our customers continue to sequence in Q1 at typical rates despite the lower reagent revenue, which is an indicator of the expected reagent rebalancing required with the launch of NovaSeq.

This transition impact is expected to continue in Q2 before improving in the second half.

We removed close to 20 HiSeq instruments from our installed base during the quarter, a figure which is expected to increase as customers begin to ramp up production on NovaSeq.

Services and other revenue grew approximately 20% versus Q1 2016 to $107 million.

This improvement was driven by strength in genotyping services due to consumer demand and sequencing instrument maintenance contracts.

Turning now to gross margin and operating expenses.

I will highlight our non-GAAP results.

As previously noted, we will report gross margin, R&D, SG&A and operating margin inclusive of stock-based compensation going forward.

This is consistent with the treatment of the expense in our non-GAAP diluted earnings per share figures.

I encourage you to review the GAAP reconciliation of non-GAAP measures, which can be found in today's earnings release and presentation.

Please note that all subsequent references to net income and earnings per share refer to the results attributable to Illumina stockholders.

Our non-GAAP gross margin for the first quarter was 66.4%, a decrease of 310 basis points sequentially.

Lower instrument margin, given high-throughput promotions and ramping NovaSeq manufacturing as well as mix within sequencing consumables and arrays were the primary drivers of the sequential decline.

Year-over-year gross margin decreased 490 basis points, impacted again by lower sequencing instrument margin and mix within consumables and arrays.

Non-GAAP research and development expenses in Q1 were $140 million or 23.3% of revenue, an increase of $10 million over Q4; and non-GAAP SG&A expenses for the quarter equaled $152 million or 25.6% of revenue, an increase of $7 million sequentially.

The sequential increase in operating expense was primarily driven by higher stock compensation expense, which grew $12 million given a full quarter of expense associated with our annual equity grant as well as increased benefit in variable compensation expense.

Looking to Q2, we expect operating expenses to increase sequentially.

Non-GAAP operating margins were 17.5% compared to 25.1% in the fourth quarter, and operating margin was lower compared to the 23.9% reported in the first quarter of last year due to increased investments in headcount, GRAIL and Helix.

Excluding GRAIL and Helix, operating margin was 21.1% compared to 29.6% in the fourth quarter.

Although the sequential decline was expected, it was larger than anticipated due to the previously mentioned drivers impacting gross margin.

We reported first quarter GAAP net income of $373 million, including the book gain on our partial sale of GRAIL shares, and EPS of $2.52 per diluted share.

Non-GAAP net income was $94 million or $0.64 of EPS, with GRAIL and Helix dilution of $0.03 and $0.04, respectively.

Cash flow from operations equaled $168 million, reduced by 100% of the GRAIL and Helix cash burn of $30 million this quarter.

Q1 DSO totaled 56 days, equal to last quarter.

Capital expenditures in Q1 were $83 million, and we reported an additional $27 million increase in property and equipment recorded under build-to-suit lease accounting, where such expenses were paid for by the landlord.

Consequently, Q1 free cash flow was $85 million.

We ended the quarter with approximately $1.8 billion in cash and short-term investments, including $278 million of proceeds from GRAIL's Series B raise, of which $92 million will be used to pay the associated tax liability.

During the quarter, we repurchased approximately 600,000 shares under our previously announced buyback program at an average price of $162, completing the authorization.

Turning now to guidance.

We expect Q2 revenue to grow approximately 7% versus the prior year.

GAAP earnings per diluted share is expected to be $0.56 to $0.61, resulting in non-GAAP earnings per diluted share of $0.65 to $0.70.

For 2017, we continue to project approximately 10% to 12% total company revenue growth, including less than a 1% revenue contribution from each of GRAIL and Helix.

For the full year GAAP earnings per diluted share, we are expecting $5.26 to $5.36, which now includes the onetime impact of the GRAIL Series B raise in Q1, and non-GAAP EPS is still expected to be $3.60 to $3.70.

Our GAAP and non-GAAP EPS guidance assumes no meaningful impact to tax expense from the new stock-based compensation pronouncement.

Thank you for your time.

We'll now move to the Q&A session.

(Operator Instructions) Operator, we will now open the lines.

(Operator Instructions) And our first question comes from Tycho Peterson from JPMorgan.

First question on NovaSeq.

Congrats on the orders.

I guess as we think about what the funnel looks like, I know you said the majority were legacy HiSeq customers.

Can you talk a little bit about any mix dynamics there in terms of low-throughput versus high-throughput labs, 2500 versus older customers and anything, I guess, we should be thinking about in terms of the pacing of GRAIL orders throughout the course of the year?

Sure, Tycho.

Since we look at where the orders came in, in Q1 and today, the majority were HiSeq legacy customers, as we pointed out.

About 40% came from customers who were just HiSeq X customers, so the large genome centers, for example.

And then about 10% were new to sequencing.

And so we were really encouraged to see sort of the broad mix of customers, validating the thesis that there is an upgrade cycle in progress here that's starting from HiSeq but that the NovaSeq platform is also attractive to HiSeq X customers, especially as S4 comes out later in the year.

We're also starting to have conversations with people who are entering the high-throughput space with the promise of what NovaSeq can do around certain projects.

So early indications but it was good to see customers from a number of segments sign up to NovaSeq.

In terms of GRAIL, we talked about the fact that over the course of the year, we expect less than 1% of our company revenue to come from GRAIL, and we still expect that.

And then can you help us on the 2Q guide?

You're coming in $0.20 lower than The Street.

How much -- is that all higher OpEx?

And can you maybe talk about what those incremental investments largely are?

Tycho, this is Sam.

So I think some of the dynamics that we saw in Q1, specifically with regards to gross margin, we expect to see again in Q2 as we ramp up the production of NovaSeq, as we ramp up the manufacturing, as we also see some of the transition in consumables with the customers, whether it's validating their workflows, depleting inventories.

We expect to see a similar gross margin profile in Q2, actually slightly lower in Q2.

So I think that's a key driver.

There is some operating expense growth as well that drives the EPS as well.

So I think those are the key drivers in terms of the sequential Q1 to Q2 EPS.

And not nearly as meaningful, but just Helix will continue to ramp sequentially as well, just more of a small impact though.

Okay.

If I could squeeze one more in because it's an important question we're getting a lot.

It's just on the cross-contamination, bar coding issues.

Francis, can you maybe just comment on that?

And I'll hop off.

Sure, Tycho.

So we've been working with a couple of customers who've experienced elevated levels of this index swapping on the HiSeq 4000.

We don't expect it to have an impact on the NovaSeq demand trajectory going forward.

The background, we have seen low levels of index swapping in the 0.1% to 2% range, and that's been sort of a part of a noise, if you like, in the [ wet ] chemistry now ever since GA, at a very low level.

It's had really no meaningful impact on sequencing results for the vast majority of applications.

And there are a number of best practices, including using washes to remove free adapters from libraries, using unique tool indices that further reduce even this low level of index swapping.

We have been engaged with a couple of customers in the last quarter that have seen elevated levels of index swapping on the 4000.

We've been working with them to understand the specifics of their situation and share best practices that will help mitigate the issues that they're seeing.

As we think about NovaSeq, we said we don't expect this to be an issue, and the offering we have coming out later this year that will help customers address individual lanes will help mitigate this issue even further because it will lower the level of multiplexing that you need to do on NovaSeq.

And the next question -- sorry, our next question comes from Doug Schenkel from Cowen and Company.

Recognizing it's still early, is NovaSeq driving any increased demand from biopharmaceutical customers?

Doug, at this stage, it's too early to call sort of an industry trend like that.

So it's certainly driven a lot of conversations, but again, too early probably to call out a specific trend like that.

Okay.

Regarding the developed -- go ahead.

I'm sorry, Rebecca.

No problem.

I was just going to say outside of Regeneron that we highlighted as one of the initial customers at JPMorgan.

Okay.

Regarding the development of the workflow to allow for multichannel loading of NovaSeq flow cells, is this largely to accommodate some of the smaller core labs that were concerned about number of samples required to make the economics work on the NovaSeq?

And in your -- I guess on your website, you indicated you decided to make this new workflow available based on customer feedback.

Factor again this feedback, how much demand do you think this change opens up in terms of market opportunity?

Sure.

So this new offering does help address a number of things, including some of the things you talked about.

What it does is it allows customers to run experiments with samples that have lower DNA inputs.

It allows them to run different applications across the different flow cells.

So it gives you a number of things that customers have wanted that they've given us feedback on since we launched the NovaSeq.

In general, these are sort of incremental improvements to the NovaSeq platform, and so we shouldn't expect a dramatic change in the demand trajectory associated with NovaSeq because of these improvements.

They are good, but they're not going to change fundamentally the demand trajectory.

And also, this comes out in Q4.

And so it really shouldn't have much impact on 2017 at all.

Okay.

One last quick one, and then I'll get back in the queue.

Given the stage where you are in the product roll-out and the manufacturing ramp, I know in your prepared remarks, you talked again about the expected manufacturing constraints and progressing through the supply chain.

Is it fair to assume that with a full month -- I'm sorry, full quarter of production that you could at least double the number of NovaSeq placements in the second quarter?

So I'll start by saying I touched on this in the prepared remarks, but we're very happy with where we are at the end of the -- at the end of Q1.

Obviously, an instrument like NovaSeq is a brand-new architecture from the ground up, and so it was a very big task to scale up the manufacturing process on something like a NovaSeq.

And so as we closed out the quarter, we shipped, as you know, just over 25 instruments.

And not only are we happy with the number, but we are also very happy with the quality of the instruments and the quality of the data that our customers are seeing.

We expect to continue to ramp, as we said, over Q2 and Q3.

And over Q3, we expect to be at the stage where we are -- we expect to exit Q3 at the stage where we're meeting the incoming order demand.

I think that we haven't given specific guidance in terms of the manufacturing for Q2, but something along the lines of what you talked about seems reasonable.

The next question comes from Ross Muken from Evercore ISI.

It's Luke in for Ross today.

I guess as you look at the evolution of the sequencing market and as it's gone all the way back to the GA and now we're pushing on the edge of clinical sequencing, how is the feedback from customers kind of shape your medium-term opportunity?

The -- as you point out, we've seen these step changes in the market before as we put out the GA and the HiSeq and the HiSeq X. And so we come into this with that experience.

And so far, I would say that NovaSeq has played well to our expectations that we expected NovaSeq to drive an upgrade cycle in our high-throughput customer base, starting that with the HiSeq customers and then to the HiSeq X customers.

And that's starting to play out as we talked about in the quarter.

We do think now that this will drive -- with the data we have, this will drive that upgrade cycle in our legacy customers.

We also talked about the fact that we expect NovaSeq, with its power and its price points, to drive the elasticity of the market.

And it is early, but there are indications we're getting from our customers in terms of new projects that have been announced, like the Regeneron Glaxo project for the biobank in the U.K., those kinds of projects as well as ultra-deep sequencing projects in the oncology market as well as liquid biopsy sequencing.

So 3 months -- it's early data, but the early data is very consistent with the hypothesis we had when we embarked on the development cycle for NovaSeq.

And the next question comes from Derik De Bruin from Bank of America.

So you mentioned 20 HiSeqs came out of the mix during the quarter and you said you had [ accelerate ]. Do you have any sort of testaments for where the HiSeq number will end up in the interim?

I'm trying to get to the swap ratios and sort of what's the indication from customers that have bought so far or so far of how many HiSeqs are going to pull out for the Novas.

Derik, we don't provide that number, and so I guess, mostly what we can do to help you is sort of give you the qualitative results that we've talked about that we are seeing and earlier than -- maybe slightly earlier than expected conversation with the X customers in terms of looking at NovaSeq.

We certainly expect as S4 comes out that we will see X customers moving or re-platforming to NovaSeq.

At this point, I don't have any data to give you -- that I can share with you around ratios.

And Derik, part of the reason for that is effectively, some customers are just testing out NovaSeq.

So they aren't necessarily thinking about the ratio for the entirety of their fleet.

Other customers are further down that path, and so the variance between customers is quite wide.

Additionally, for us to take something out of the installed base, it needs to really not be used, right.

And so that's not something that really we can forecast at this point in time because many customers do keep these boxes installed until they're not being used for quarters upon quarters.

And so that's more of a retroactive look and not necessarily something that we can forecast.

Great.

And then a quick follow-up on that.

You said the consumables were -- the consumable pull-down was a little bit more than you thought in terms of slower -- you're attributing that to more inventory reduction rather than delays in terms of the sequencing market.

I agree with you, but I just was wondering if you did any -- found any evidence at all that certainly, the funding uncertainty in the market is having any issues.

We have not heard from anywhere that the funding uncertainty is what's driving the consumable slowdown.

What we do know is that we initiated conversations with our high-throughput customers, and that has caused our customers to do a number of things: One, they are pausing so they can talk to us about NovaSeq.

They are drawing down existing inventory with the understanding that they are likely just sort of re-platforming the future.

They're also not doing any large future buys for the same reason.

And then initially, there will be the smaller purchases as they validate the workflows associated with the NovaSeq.

And so those are expected steps, and so that's what's driving the slowdown that we saw in the consumables.

Derik, Marc here.

Having said that, I'd just point out that Q1 is normally a strong order quarter for consumables, sequencing consumables in general, and we did see some backlog build there as well from our customers.

So those who aren't slowing down at least are still placing orders and expecting those to be shipped throughout the year.

And the next question comes from Dan Arias from Citigroup.

Francis, can you just talk about the back half of the year?

Any assumptions around X system pull-through once the S4 configuration is out there and presumably folks start doing whole-genome mark de novo?

Are you still thinking you can exit the year with the X still in the $625,000 to $725,000 range?

So we are not expecting those pull-through ranges to sort of carry forward as we go through this transition period.

And so I talked a little bit about the dynamics that are playing out in the high-throughput market in terms of people looking to re-platform NovaSeq, and that is affecting the kinds of orders that they will be placing and the work that they'll be doing as they transition onto the NovaSeq platform.

We are also expecting very little in terms of new X placement, obviously, as we look at Q2 going forward, and so that will have some impact in the back half of the year overall as well.

And Dan we said on the Q4 call that we had suspended that range, and so this isn't necessarily a change in messaging.

It's effectively just a continuation of that messaging where we don't necessarily have a HiSeq or HiSeq X guidance range at this point in time because these are -- this transition is dynamic.

Yes.

And so what you can expect to see is that this is going to be a transition year.

And then over -- as we go through sort of a 4-quarter period on NovaSeq, we'll be in a better position to start to talk to you about what you can expect on NovaSeq.

Okay.

And maybe just following up on your comment, Francis, on HiSeq versus X customers in the order book, can you just sort of further that thought by talking about how many of the 85-or-so orders beyond the 49 that you announced in January came from conversion of the X backlog?

I think there were 35 or 40 in the backlog as of 4Q.

Just curious how many of those traded up.

We haven't done the breakdown on the post-49, the walk from, I guess, 49 to the greater than 135.

I think the overall number is generally representative, I'd say.

I don't know.

I haven't done the breakdown for the post, but I don't expect there's incremental insight to come from that breakdown.

And the next question comes from Amanda Murphy from William Blair.

I just had a couple of follow-ups, some of the questions that have been asked.

So I guess, first, on the replacement cycle more broadly, so you talked to some of the demand that might have been a bit surprising for the legacy platforms outside of the X. So I'm just curious, has that -- I guess, has that demand changed your view ultimately on what the magnitude of the replacement cycle could look like for the NovaSeq or the timing?

And then just do you have any updates on your thoughts around supporting, or I guess offering, the legacy platforms?

I think you made comments about 2500 but some of the other HiSeqs.

Sure.

I'll start by saying that we still deeply believe that NovaSeq will drive an upgrade cycle of our HiSeq customer base and, over time, our HiSeq X customer base.

And so we talked about the fact that we have 800 HiSeq customers, and we have north of 30 HiSeq X customers.

We deeply believe that, and everything that's happened in Q1, in my mindset, reaffirms that thesis for us.

What we saw in Q1 was we saw a slightly stronger-than-expected demand, we said, for HiSeq X and for HiSeq.

And what was playing out there were there are customers that are in the midst of running large whole-genome projects and they needed more capacity.

And what they didn't want to do is switch platforms in the middle of a project, and so they added capacity and they added Xs.

We also saw some customers in China that continue to buy Xs because in that region, it's a very competitive platform.

And until we get to the S4, it's going to continue to be a very competitive platform.

And then we have some clinical customers that have their validated workflows that are sticking with the platform that they're using.

We think these are transition effects.

So I think they drop off very quickly even in Q2.

And so it doesn't really change our thesis around the upgrade cycle.

Okay.

That makes sense.

And also change our position, continue to [ work ] our portfolio plan around the instruments we're looking to retire.

That doesn't change.

Again, I just had one other follow-up to Tycho's question on margin.

So just thinking about when you have the NovaSeq sort of up and running, can you give us some at least qualitative commentary on the right way to think about the margin profile on the NovaSeq when you -- from an instrumentation consumable perspective particularly given that you're obviously launching higher-performance cells and single-loading cells going forward just on a sort of more sustainable basis?

Sure, Amanda.

This is Sam.

So I think the way to look at NovaSeq margin profile from an instrument standpoint, I think it's slightly lower margin than our HiSeq platform.

And I think as we go forward in the year and as manufacturing ramps up, it will -- the margin profile on the instrument will improve, but it will still be slightly below the HiSeq X instrument and the HiSeq profile.

So there will definitely be an improvement in the instrument margin as we go through the year, but it will still be below the HiSeq platform, higher than the benchtop platform though.

From a consumable standpoint, I think the important part is that as we ramp up consumables in the second half of the year, we also expect to see a margin profile on the consumables that's on par with our high-throughput consumables.

And the next question comes from Jack Meehan from Barclays.

I also want to follow up on one of Tycho's questions.

You mentioned 10% of NovaSeq customers are new to sequencing.

Is that what you thought the uptake would be like compared to previous new product launches?

And any other data points around new customer interest would be great.

Yes.

I mean, that was a good number.

It wasn't a hugely surprising number.

Given the dynamics that we are seeing, we did expect some new customers to be created in the high-throughput segment.

Some of them are regionally oriented.

So we're seeing new customers, for example, emerge in China, some of them driven by the opportunity created by the Chinese Precision Medicine Initiative.

So the number wasn't really that surprising, and we expect to see some of that going forward.

Obviously, most of the new sequencing customers typically show up at the lower end of our portfolio.

So 60% of the MiniSeq, MiSeq level of our customers coming in there are new sequencing customers.

But at the top end, 10% is not too surprising.

Got it.

And then just as a follow-up.

I want to ask about the service revenue strength in the quarter.

It looks like it was coming through the micro-array line.

Just any color there would be great or broader trends around service revenue.

Sure, Jack.

Yes, absolutely, it was coming in from the micro-array line, and it's driven by the real strength we're seeing in the consumer market.

And so it's related to the services that we offer to the consumer market, which now, for many quarters, has been a strong segment for us.

And the next question comes from Bill Quirk from Piper Jaffray.

This is Alex Nowak on for Bill today.

So the 135 orders surpassed your expectations, but you also reiterated the top line guidance.

So I was just wondering if you could reconcile the 2.

Yes.

The 135 was slightly better than we were expecting for Q1, but it's still very early in the year.

We still have -- the majority of the work to do for the year is still in front of us.

This is still a back-end-loaded year for us.

And so while we are happy with the way Q1 played out, we feel it's too early to change anything for the whole year.

Okay.

And then just second question from me.

What gives you confidence that there's enough sequencing demand that NovaSeq's consumable sales pull-through per instrument will be about the same, if not better than HiSeq?

So a few things.

One is our experience now in this market across multiple -- the introduction of multiple high-throughput platforms has given us a track record now of seeing what happens with our customers that as we give customers high-throughput capacity, they are able to embark on more ambitious sequencing-intensive projects.

And so in previous platforms, customers are able to go from panels to exomes, from exomes to genomes with the X. And what we are hearing from customers now is that more customers want to go from exomes to genomes, so not just the X customers, but now our HiSeq customers want to do that.

And so the 4000 is good, but the NovaSeq really enables that.

In addition, they're giving us examples of projects in areas like single cell or wanting to do much, much deeper sequencing for tumor-normals or liquid biopsies or other applications where they're effectively looking for the uncommon needle in the haystack type of events.

And so we're getting feedback from our customers around the projects that they want to embark on, which is consistent with the experience we've had in this market before.

And now as we come out of Q1, it's clear that customers are buying a high-throughput platform because they intend to use the capabilities that come from a high-throughput platform.

So as they purchase, they have plans around what they want to utilize the platform for, and they have expectations that it will be a reasonably utilized platform, too.

And so now we have a few months of data now coming to us with customers reacting to a specific platform.

And that's very consistent with our thesis although it's early.

And the next question comes from Isaac Ro from Goldman Sachs.

Francis, I had a question for you on the consumer market.

I think the FDA recently took a more dovish stance with regards to consumer testing for things like risk assessment.

And so I'm just wondering if just given that, that was a relatively new development if there is anything baked into your guidance this year with regards to an acceleration in those types of markets.

Yes.

That was a really good step forward, we think, for the consumer market.

We are really happy about the impact that's going to have definitely and immediately on 23andMe.

But we think it's a very good sign for the entire consumer market, so -- [ Ancestry ] Helix and everybody else.

And that obviously is good for us, too.

Now we have known that 23 has been working on this for a while.

And so while it was news obviously when it happened, it's not entirely unexpected from our perspective.

And so it's consistent with our view of how the consumer market is likely to play out over the course of this year.

Got it.

And then maybe just a follow-up.

I think all the questions around NovaSeq were pretty well covered.

So I wanted to maybe focus on the other parts of the portfolio on the instruments side, specifically Firefly.

I think you guys gave a little bit of an update in January regarding performance specs and time line for end of year.

Just wondering if that's still on track and any evolved thoughts on what that product line ultimately means for your expansion of the opportunity.

Yes.

The team is making really good progress on Firefly even since the update we gave in January.

As we talked about in January, the intent is to bring out a sequencing box end of this year, beginning of next year, and then bring out the library prep box about a year later.

And the team is on track to do that.

This is one of the fortunate situations where we announced a product well before bringing it into the market.

And one of the reasons to do that was to allow us to engage with prospective customers.

And the team has been doing that and so -- and starting to get a good sense for the customer segments that it will be applicable for.

And what's particularly exciting is they're identifying customer segments that may be new to sequencing that are attracted to Firefly, and so good progress on track with the time lines that we talked about in January.

And our next question comes from Steve Beuchaw from Morgan Stanley.

So first question is on consumables and sequencing consumables.

Sam, could you put any more granularity around what you saw in the first quarter and the impact of some of these dynamics around Nova and the way customers are thinking about the transition?

I mean, given that you have, I would imagine, a pretty good handle on who the customers are, do you have any idea of what sequencing consumables growth might have looked like on an apples-to-apples basis in a scenario where we didn't have the Nova transition?

Yes.

It's hard to give you kind of the apples-to-apples comparison that you're asking for, but let me give you a little bit of color at least in terms of what we saw in Q1 and our expectations going forward.

I think what we saw is basically almost what we expected, which is as Frank has mentioned before, customers essentially validating their workflows, pausing, depleting some of their inventory, and we think that played into the lower sequencing consumable volumes that we saw in the quarter.

As we look -- and we think this is expected given the rebalancing to a new platform, which is NovaSeq.

And as we look forward specifically to the second half, I think the expectation is that we would see a ramp-up in consumables.

Now as we said, going into Q1, that consumable growth is going to essentially decelerate given some of that time that customers will need to ramp up on their consumable purchases of NovaSeq.

But essentially, it's playing out as we pretty much expected, maybe slightly lower consumables on the HiSeq, HiSeq X compared to what we had expected.

But as we look forward, we think that that's -- our expectations are that we will see consumables ramp up with the NovaSeq uptick.

In the back half of the year.

In the back half.

And one thing to add to that, Steve, is there is absolutely nothing that we've heard from our customers that would lead us to believe that there is anything going on with regard to sequencing consumables beyond this natural transition.

We have seen very, very large customers work down their inventories on hand, which led to the sequential decline in the high-throughput component of the portfolio.

So I think that is absolutely on par with our expectations, to Sam's point, maybe a little bit quicker than we had expected but that is aligned with HiSeq X customers exceeding our expectations with NovaSeq.

So it's all one in the same in terms of the drivers.

And on an apples-to-apples basis, ex the NovaSeq transition, we would expect nothing beyond the traditional formula of what sequencing consumables would do in any given quarter.

Okay.

I appreciate all the color there.

And just one follow-up here and that relates to the operating expense.

There's been some commentary on the prepared remarks on adding new feet on the street, if you will, new focus throughout the organization.

I wondered if you could spend a minute just talking about where these people are headed.

Then maybe if you could compare headcount growth that you're thinking about for the company in 2017 relative to the last couple of years just as we think about the modeling, that would be a big help.

Yes.

So one of the areas we are going to be adding feet on the street is around focusing on the low-end part of our portfolio, the benchtop instruments.

So we're looking to add an inside sales team, expand the inside sales team that focuses on that part of the portfolio, additional lead generation capability to drive leads into that part of the portfolio.

You, as a customer, would have already seen the investments we've made in e-commerce.

So that's come online in the past few months, and we're going to continue to invest there because we think there's a real opportunity to improve the customer experience as well as improve the velocity at which we do deals and our cost to serve by moving the part of the consumables, part of our business as well as the low-end portfolio through e-commerce.

And so those are some of the areas that you can expect to see us expand our capacity.

And I would just add geographically in China as well, a major investment for us this year.

And the next question comes from Puneet Souda from Leerink Partners.

Just one briefly.

I just wanted to understand a little bit.

I think it's been covered -- parts of it have been covered before, but as the installs that are -- already happened, some of them have gone through the validation.

If you could talk about and maybe give color on how labs are designing experiments.

I'm sure there are some labs that are transitioning from validation to designing their own experiments on this.

Are they simply waiting for S4?

Or are they, in the meantime, using the S2 flow cell to try to do some of these experiments and does that -- and taking on some of the production samples onto S2?

And does that change your view at all on timing of S4 and maybe potentially moving that up?

So I'd say that if you think about the makeup of the people who bought NovaSeq, I said the majority of them were legacy HiSeq customers.

And so for them, the move to S2 is a beneficial move on its own.

And so they're certainly not waiting for S4 and nor do they need to.

We are -- and so we expect the people who bought NovaSeq will start to sequence on S2 and very few, if any, will be waiting for S4 before they start sequencing.

Obviously, if you are HiSeq X customers, you're designing very large experiments, you know the S4 is a really exciting flow cell for you.

And it was actually that feedback we got from customers when we launched the NovaSeq that caused us to pull the S4 in from Q4 to Q3.

And so that's our plan of record, and that's the plan that we're going to stick to, but it was very much driven by the higher-than-expected interest we got in S4 from X customers certainly.

Okay, got that.

And then just one quick follow-up.

Could you update us on the manufacturing ramp?

I mean, there is -- obviously, the HiSeqs are getting essentially reduced in the -- some of the facilities, the San Diego, the Bay Area, the Singapore and Cambridge facility and the Wisconsin facility that you have.

So I'm just trying to understand how are you ramping up NovaSeq and retooling some of those facilities to ramp up on both flow cells and the instrument.

Yes.

The -- as we said, Puneet, NovaSeq is every -- is a completely brand-new architecture, and every major component is a brand-new component in NovaSeq.

And so it has been a really big effort on behalf of our operations team to ramp up our manufacturing capability of NovaSeq and transition that product from product development into our operations team.

And I am incredibly proud of the terrific work that team has done in getting to where we are and allowing us to be slightly ahead of where we expected to be as we got out of Q1.

The way we think about our facilities broadly is a lot of our new introduction work happened here in the U.S., close to the development centers.

And that's consistent with how we're thinking about ramping up NovaSeq.

And that's sort of how we're thinking about it.

And the next question comes from Dan Leonard from Deutsche Bank.

Just wanted to clarify, well, 2 questions.

The first one, clarification on gross margin.

Sam, I think you mentioned that the promotions in the high-throughput sequencing market were part of the depression on gross margin in the quarter.

Can you clarify what you're doing in the promotional space in the high-throughput market and why that would continue in Q2?

Well, as we ramp up NovaSeq and for some of the customers that have ordered HiSeq in the past, we've said that we were going to have a promotion where we would match some of the prices.

And so we have -- essentially, that's had a negative impact on ASP on some of our high-throughput instruments.

So that's primarily the reason.

In the first quarter, that's just not expected to continue into the second quarter, Dan.

The second quarter impacted more along the line of the ramping NovaSeq percent, effectively mix in the sequencing instrument line.

Yes, right.

So that's the second part, which is as we move forward, it's really a question of NovaSeq ramping up in terms of production.

And until the volumes catch up in Q3 and Q4, we're going to continue to see some negative impact on gross margin in Q2 because of that.

Okay.

That's what I wanted to clarify.

And then my follow-up question, can you offer us an update on what percentage of your revenue is China?

And you're going to be rolling up against some pretty tough comps in the second half from China and whether you expect the growth rate continue at the levels you've been seeing.

Yes.

We remain bullish on the Chinese market opportunity.

It is roughly, you should think about it as around already 10% of our business.

And the secular drivers of the market there continue to be positive.

So we're seeing growth in the clinical market associated with both NIPT and oncology.

And that played out last year.

We expect that trend is continuing to play out this year.

We're also seeing demand driven by the government's Precision Medicine Initiative.

So money has started flowing.

Projects are being funded, and that is causing customers to purchase our instruments, that's causing, in some cases, new customers to be created.

And all of that is going to continue to play out certainly over the course of this year but likely for a few years.

In this quarter, Annoroad just got clearance from the CFDA around its assay, and that will continue to give it momentum in the market.

And so it's a market we are excited about.

And just one thing to highlight with regard to our out-performance in the region last year.

Obviously, we'll have a challenging comp as we go throughout the course of this year and the benefit of HiSeq X instrument sales, in particular, that we saw last year.

So just to take that into account, I completely echo Francis' statement with that caveat.

And the next question comes from Tim Evans from Wells Fargo.

I don't want to neglect the desktop lines -- the benchtop lines rather.

Can you talk about whether your pull-through there was within the guidance ranges for each of those?

And also comment on the level of placements.

Where they kind of up, down or flat relative to the back half of 2016?

The pull-through numbers were consistent with the ranges that we provided, and the placement numbers were roughly in line with where we expected.

Okay.

And then the weakness in Europe, can you just talk -- maybe give a little color around that.

The weakness in Europe was primarily driven by the NovaSeq transition and the (inaudible) they had a good Q1 shipment number last year.

I wouldn't call it weakness in Europe.

I would call it more they met their forecast and plan.

And the next question comes from Bryan Brokmeier from Cantor Fitzgerald.

So you mentioned the solid 20% growth in oncology in your prepared remarks.

Is a large piece of that growth coming from the TST 170?

Or if you elaborate on the underlying driver of that.

TST 170 is a piece of it, but it's certainly not a big piece of it.

I mean, that's -- it's a fairly new product to the market, and so you should expect to see that ramp.

But in Q1, it's certainly not a big piece of the growth we're seeing.

We are continuing to see strength in the translational market in oncology driven by the large cancer centers.

We talked about the growth we're seeing driven by the adoption in the liquid biopsy market of sequencing.

So those are some of the variables that played out much more so than TST 170.

All right.

And would you expect that -- this piece of the market to adopt the NovaSeq?

And therefore, could there be any sort of a hiccup in oncology as those customers transition over to that system?

No, I certainly expect over time, given that some of the applications we're talking about include things like very deep sequencing for tumor-normal, for liquid biopsies -- so I expect over time, NovaSeq will be attractive to a segment of the market.

But we're seeing enough growth from other parts of the market that I'm not sure that you should expect to see a pause in this market at all.

I don't think it will be just a wholesale adoption of NovaSeq.

I mean, there are so many different types of customers that are buying that I think you'll just continue to see growth in that market.

And the next question comes from Catherine from Baird.

Catherine Schulte here.

Are you accepting S4 orders in advance of that launch?

Are those included in the 135 number?

And then will you have initial manufacturing constraints for the S4 when it launches?

Or will it immediately be able to meet demand?

The way to think about the 135 is people are buying the platform.

And initially, they'll buy [ S2 ] because that's the flow cell that's available, and then they will buy -- some of them will -- may buy a flow -- S4 flow cells when it becomes available.

But they're not buying them and waiting for S4 and all that.

It's not how they're thinking about it.

What was the second part of your question?

Will there be initial manufacturing constraints for S4?

I don't think so.

I think the bigger constraint would be just on the platform.

That's the constraint that we're working through.

I expect that as we bring S4 out, we should be able to ramp up to meet demand.

Okay.

And then for NovaSeq orders, can you just describe the cadence of those throughout the quarter and entering the second quarter?

I know you had a lot of buzz around the JPMorgan conference, but just talk about how those progressed over time.

Yes.

I think it's interesting.

It's been nonlinear, sort of -- probably the best way I'd describe it, in the few days before JPMorgan and in some cases, actually in the few hours before JPMorgan, actually, I personally reached out to a small set of customers, and a couple of other people reached out to a couple of others.

And so that's what got us the first set of customers.

And so we were able to come out of the gate and talk about the commitments we had from that first set of customers.

For everyone else, including everyone in the Illumina commercial organization, the first time really they heard about NovaSeq was when it was announced on stage at JPMorgan.

And so in some sense, that was sort of the starting gun for our entire commercial organization.

And we had immediately planned that day a set of training webinars and they -- a marketing campaign ready to go, our roadshows that literally took off that week around the globe to start to explain to our field and to our customers what NovaSeq was.

And so we got that initial bolus of customers in the few days, in the few hours before JPMorgan.

And then we started the sales process.

And so there was a period when we weren't getting any orders because we were still teaching our own team and teaching the rest of the world what NovaSeq was.

And so that started to play out in the weeks after we launched.

And so it really started to come to fruition, if you like, in March primarily, where those conversations started to play out into orders.

I mean, roughly obviously, there are exceptions, but that's roughly how it played out.

And that concludes our question-and-answer session.

I'll now turn the call back over to Rebecca Chambers for final remarks.

Thank you, everyone, for joining us this afternoon.

As a remainder, a replay of this call will be available as a webcast in the Investors section of our website as well as through the dial-in instructions contained in today's earnings release.

This concludes our call, and we look forward to our next update following the close of the second fiscal quarter.

Thank you, ladies and gentlemen.

This concludes today's conference.

Thank you for participating, and you may now disconnect.