Affymetrix, still shaken from a series of forecasting and manufacturing missteps, last week offered additional insight into the problems that have made its genotyping business difficult to forecast in recent quarters.
Specifically, Affymetrix's Chief Financial Officer Greg Schiffman for the first time identified a problem with the firm's software for its 500K Mapping Array Set genotyping product as a reason for scale-up issues for customers that affected reorders.
Schiffman, who spoke at the Deutsche Bank Health Care Conference in Boston last week, said that Affy is now distributing a new algorithm to 500K users, and divulged for the first time that Affy will launch two new genotyping products this quarter. During his talk, Schiffman also conceded that Affy is facing more competition in the genotyping market, noting that it is likely to increase.
Explaining Affy's lackluster Q1 performance, Schiffman reiterated the company's position that its weak first quarter in genotyping had little to do with its technology, and more to do with the effectiveness of its organization.
"Where we've struggled over the last nine months — and this has been something new to the company — is getting the products we've developed cleanly through manufacturing into the customer's hand and then sort of the post-sale and customer support," Schiffman said.
"One of your competitors has obviously enjoyed some nice success. I think some people in the Wall Street community would suggest that it is almost a zero-sum game."
Schiffman added that this structural weakness, which depressed revenue for the first three months of the year to miss expectations by $5 million, was exacerbated by the scale of the company's initial launch of the 500K product. He explained that Affy usually has between five and 10 early-access customers, but in the case of the 500K product, Affy released it to 50 early-access customers.
"This really strained our ability to [deliver the 500K arrays to customers] in the early phase of the release," he said. Schiffman said Affy has a "relatively small number of important customers that are pursuing experiments that do equate to the [genotyping] revenue stream."
"As they had challenges ramping up, the reorder rate and the speed of them being unable to complete their research has been one where we overestimated — and they themselves overestimated where they are at," he said.
Algorithms and Assays
When problems with the 500K first arose at the end of September 2005, Affy attributed its inability to meet guidance to manufacturing constraints — constraints that led Affy to expand its Sacramento facility's capacity by 60 percent over the past six months, and to invest $5.5 million in a new plant in Singapore (see BAN 10/5/2005, BAN 1/31/2006).
However, as Schiffman said at the Deutsche Bank conference, Affy's performance in genotyping last quarter was due to the timing of reorders for the 500K product, which in turn was slowed by the complexity of the assay for the 500K.
Part of that problem was Affy's original data-analysis algorithm for the 500K, which Schiffman said "wasn't fully optimized" upon release. "[It] didn't allow the power of the array to be fully realized and as a result we saw customers experiencing delays in scaling up to their 500K product," he said.
Schiffman said that to address this challenge, Affy has taken two steps: The company has invested more in training for its sales and support team and has introduced a newer algorithm for the 500K that will be in the hands of all Affy 500K customers by the end of this month.
"[With the older software] there was a larger degree of disparity between heterozygote calls and homozygote calls," Schiffman told investors. "The new software algorithm takes the same information from the chip that we've already processed but enables them to extract all the information off the chip. It really has corrected the challenges we've seen in terms of the software," he said.
In addition to rolling out the new algorithm, Affy will also launch two new SNP assays to supplement its 500K product, Schiffman said.
"We've got two that we are releasing this quarter," Schiffman said. One of the products is a targeted 50,000 SNP assay that uses SNPs from the HapMap project to "extend the coverage" of Affy's 500K product to 550,000 features. Affy's chief genotyping rival Illumina's latest array, the HumanHap550, offers customers 550,000 SNPs, although both companies deny the existence of a numbers game.
Affy's other new product is a different 50,000 SNP array that offers genetic HapMap coverage and includes 20,000 of the non-synonymous SNPs from that project.
The new products are just the beginning of a suite of genotyping products that Affy has promised to release this year. In April, Affy CEO Stephen Fodor said that the company was entering a "phase now where [it is] going to increase dramatically the number of new products in genotyping and clinical applications as well as diagnostic applications."
"Over the next year and a half expect to see a large number of products roll out," Schiffman said last week. "Some will expand the breadth of the 500K and others [will target] very specific subsets in application and development processes," he added.
During that time, the company also plans to shrink its feature size from 5 micron to 1 micron, which he said will help bring the cost of Affy's chips down while providing the same, if not more, content. "This will allow Affy to penetrate new markets that are more price sensitive, including emerging areas such as pathogen detection, biodefense, and water purity [testing]," Schiffman said.
At the conference both Schiffman said that the bulk of Affy's challenges are behind it. Still, he said that the company will not resume forecasting its performance in the near future. The company suspended providing guidances following the release of its Q1 results.
"I don't think we are at a point where we feel there is accurate clarity to provide a range to the market over the next quarter or two that is reasonable for specific revenue guidance," Schiffman said.
Not a 'Zero-Sum Game'
Although Schiffman assured investors at the conference that Affy's problems were in its execution, not its science, the company's worse-than-expected performance last quarter, compared to Illumina's better-than-expected performance, prompted one investor to ask whether Affy is losing genotyping business to its rival in San Diego (see BAN 4/25/2006).
"One of your competitors has obviously enjoyed some nice success. I think some people in the Wall Street community would suggest that it is almost a zero-sum game," the investor commented.
"We don't believe it's a zero-sum game," Schiffman responded. "We placed twice as many instruments in Q1 as the nearest competitor did. We sold more consumables than anybody in the marketplace did. We feel that we are making great progress in this space. I don't know if we've necessarily conveyed effectively that we still did see good growth in the genotyping business," he added.
Schiffman said that in "the academic arena the majority of customers have always used both products and from that standpoint they continue to use both products."
"Traditionally Illumina has been used in targeted SNP studies and Affy has been used in whole-genome association studies. We are introducing products in the targeted subset space. Illumina is introducing products that are going after our space. From that standpoint, we are starting to see each other in competitive fields that we didn't use to see [each other in]. But we haven't heard from customers that [they] are moving from the platform," Schiffman said.
Schiffman added that Affy expects the genotyping market to become increasingly competitive and predicted that it could swell to a $1 billion market in the future.
Illumina spokesperson Bill Craumer declined to comment on Schiffman's remarks.
— Justin Petrone ([email protected])