Rebuilding a Genomics Commercial Engine

Executive summary

When I joined NanoString as SVP of Sales & Marketing in July 2017, the headline number was 4% growth. The more important number was the trend behind it: product and service revenue had grown 42% in 2014, 27% in 2015 and 22% in 2016. 3 years of slowing growth is not a timing problem, so the question was what in the commercial engine had stopped working.

The answer was in how the business made its money. Growth depended on placing new instruments, and placements were falling. Reported growth was supported by collaboration payments that don't recur. And the consumables channel set up earlier in 2017 had consumable reps and inside sales covering the same accounts, both paid on all of the revenue in them, so no one clearly owned growing consumable use in the systems customers already had.

Over the next 4 years we gave every account one owner, widened the applications nCounter served, and launched 2 new platforms without pulling the nCounter team off its core business. Growth recovered to 16% in 2018 and 24% in 2019, and product and service revenue doubled from $72.0 million in FY2017 to $144.0 million in FY2021. The core business returned to growth in 2018 and 2019, and GeoMx, launched into that base, accounted for most of the increase after 2019. Revenue also became more predictable, averaging 99.4% of the initial guidance midpoint from FY2018 to FY2021.

1. Where we started

NanoString sold nCounter analysis systems and the reagents that run on them, and earned additional revenue from research collaborations. When I arrived, the business had 3 connected problems:

• Dependence on placements. Revenue growth relied on new instrument placements, which fell roughly 40% year over year in the third quarter of 2017, the low point. Instruments are capital purchases, so this revenue was uneven and hard to forecast.
• Revenue mix. Collaboration revenue was 37% of total revenue in 2017. It does not recur and sits outside the commercial team's control, so this case study reports product and service revenue, the line the sales organization owns.
• Installed base not fully used. Each placed instrument could produce years of consumable revenue, but most of the channel was still organized to win new placements rather than to grow consumable use in accounts that already owned a system.

There were 3 main changes:

• Rebuilt the consumables channel. In 2018 we added roles and gave every account one owner: instrument reps on placements, consumable reps on the highest volume accounts, and inside sales on the long tail, with pay plans weighted to each role and a Global Consumables Director accountable for consumable revenue.
• Widened the applications. We extended nCounter into immunology and neurology (about 15% of system sales in 2017, about 40% in 2018), bringing in new buyers.
• Built spatial biology sales without distracting the nCounter team. GeoMx launched in 2019 and CosMx in 2021. From 2018 we added a small team of technical sales specialists alongside the instrument reps, who kept the customer relationship and their focus on nCounter. In 2019 nCounter instrument revenue held at about $21 million, as guided, while GeoMx added $10.0 million.

2. Results, FY2013 to FY2021

• Growth. Growth fell from 42% in 2014 to 4% in 2017, then recovered to 16% and 24% in the first 2 full years of the rebuild. The 29% growth in FY2021 was the highest since 2014.
• Margin. Product gross margin reached 57.5% in FY2019. The increase came from higher consumable volume rather than from price.
• Planned margin decline. Margin fell to about 53% in FY2020 and FY2021 as GeoMx and CosMx scaled, trading some near-term product margin for a larger future revenue base.

3. The installed base, measured on a net basis

The gross base grew about 75% from FY2017 to FY2021, while the net base levelled off near 600 systems. On the 5-year basis, nCounter consumables per system rose from $78,000 to $86,000 between FY2017 and FY2019, fell to $61,000 in FY2020 when research labs closed, and recovered to $82,000 in FY2021.

4. Predictability: revenue against guidance

3 of 4 years landed within or above the initial range, averaging 99.4% of the midpoint, or 103% excluding FY2020. That matters for valuation because a buyer applies a lower discount rate to revenue it can forecast with confidence. FY2020 guidance was issued in February 2020, before COVID-19 closed research labs; product and service revenue still grew 7% that year.

5. Lessons for an instrument and consumables company

In order of impact, these are the steps I would take first in any company that sells instruments and consumables:

1. Give every account one owner. Pay consumable roles on consumable revenue, and make sure no two people are paid on the same dollars.
2. Measure consumables per system on the systems in use. Use a net installed base or actual usage data, not the gross count, and track it by year of placement.
3. Forecast instruments and consumables separately. Track accuracy against initial guidance every year, so the record is there when a buyer asks for it.
4. Launch the next platform into the existing base. Add specialists alongside the existing reps rather than moving the reps onto the new product.
5. Widen the applications the platform serves. New applications bring new buyers, and every new system adds consumable revenue.