The next wave of consumer genomic products need to be cheaper, leaner, and multiplayer

If you work in the genomics field, I have written this to confront dogmas and stretch your imagination.

Michael Jordan can help Direct to consumer genomic testing lift off.

(Part 2 of 2. To see Part 1, click here.)

The association of a person’s genotype to their traits are generally better for traits that are health-related than those that are not health-related. We have a better sense of what variants are implicated in breast cancer than breast size. This makes sense given the focus of funding in biomedical research. However, translating health-related content into DTC products have been difficult because a DTC healthcare market hasn’t materialized meaningfully yet. I expect that it will, and this area will continue to be developed. But I previously wrote that until then, the next wave of DTC products could focus on delivering entertainment, without being held back by doubts of the predictive value of the test. Customers may be willing to engage with “sunshine” genetic tests that focus on the enjoyable aspects of genetics that don’t need to provide any guarantees on its predictive capabilities. With this angle, entrepreneurs can focus on validating new products and markets, rather than validating the science, which for nearly all traits – especially multifactorial and polygenic ones – will almost always have flaws anyway. Therefore I propose this framework to create DTC products in genomics: they need to be leaner, cheaper, and multiplayer.

Leaner

New product concepts in the DTC entertainment category could sound outrageous. The problem with novel ideas are the lack of evidence of whether the product should be built. Therefore lean startup methodologies need to be employed, similar to how new tech products are developed. Frameworks given by Steve Blank, Eric Ries, and Alberto Savoia are useful starting points.

The customer “job to be done” should be clarified and a DTC product shown to meet it, validated through prototypes or other mock-ups. Once proven, the test should focus on delivering only that. Delivering a narrowly-focused product is sometimes counter-cultural in the genomics industry, where there is the temptation to build products that generate more data for future use. This could be useful for more complex business models but for simplicity I won’t address this here, and assume that all data should be deleted as soon as a result is produced for the customer.

Cheaper

Getting data as cheap as possible is essential because customers have a limited frame of reference for DTC genomics products. Starting cheaper will allow customers to give it a go, preferably at prices lower than $100. This has worked for ancestry testing, which incidentally is also priced appropriately for gifting.

Is it possible to sell tests at lower prices without sacrificing profitability? Genomics is sometimes and unnecessarily synonymous with next-generation sequencing. These assays costs 100’s of dollars. While the ability to sequence wide and deep is useful for research and diagnostics, in an entertainment setting where the stakes are low this is not necessary. Until these costs drop significantly, technologies that detect the bare minimum number of variants should be used. This could be through low-plex SNP assays, low content arrays, or low-pass sequencing, using imputation as needed. These assays could cost as low as 10’s of dollars, and a good reminder of why 23andMe might have stuck with arrays. Combined with deleting data after producing a result, the minute amount of data produced may help ease concerns on privacy.

Multiplayer

Taking a key trait that customers might be interested in then producing the result for a low cost allows an entrepreneur to focus on the user experience of how the product is ultimately consumed. For many existing DTC genomic products, this has been a report of some kind to be read in private. This stems from the legacy where genomic results are confidential, serious, and the result themselves elicit a strong reaction from its owner. Because the results for non-health DTC tests have weak predictive values, it won’t be delivered with the same gravity. Therefore if the reception of DTC results remains this lonely, dour endeavour, the customer experience and social utility will be limited.

Entertainment genomics should be fun and associated with doing things with others. Using gaming as an analogy, DTC genomics has mostly been a single player game. Instead, it needs to help the customer engage their friends and make it a multiplayer experience.

Imagine a product where the traits should be shared and misinterpreting them is part of the experience. If social media users are looking to share interesting discoveries with their friends, results of entertainment DTC products should be designed for increasing engagement (e.g. picture or video format), re-posting, and commenting (encouraging disagreement and re-interpretation). Consumers enjoy having using novel content to engage with their friends and family around a dinner table, or friends and followers online; DTC genomics can provide a treasure trove of novel content.

If traits should be shared, they could be used in a gaming context between multiple customers. A customer learning about traits A-Z on its own could be bored (especially since their predictive values are limited), akin to reading an astrology report. It would read like a check list of their predicted phenotypes that they probably already knew about. However they may be interested to see how they sit in a spectrum among people they know, and this experience can be gamified.

Product ideas to help you start brainstorming

This set of quantity-over-quality concepts aren’t necessarily any good at all, but is listed here to expand your imagination. If you’re scoffing as you read through this list, ask yourself if would have read the business pitch for Twitter before it was developed, and bought into the idea.

• What do you and Michael Jordan have in common? The basketball legend has been blessed with great genetics along with his work ethics, and you can test to see what traits you have in common – and what you don’t.
• Dating apps that use supposedly shared traits to find what’s in common between two people, or to create some comical conversation points.
• Trivia between friends to find out who is more prone to being sleepy, who can tolerate more alcohol, or who has a larger appetite. Designed for blow-by-blow question and answers to debate fiercely (#myDNAwins or #DNAfail).
• A Nintendo game where your character’s skills are based on your DNA.
• Which siblings do you share more DNA with? Which grandchild has inherited more of a grandparent’s DNA?
• How do your genes compare with others in terms of height or weight? Has your lifestyle helped you beat where you were expected to sit in the spectrum?
• If you’ve got mixed ethnicities, learn which of your facial features are derived from which ethnic group.
• 52 of your traits delivered each week as a post for your social media account to engage your followers.

Conclusion

Genomics is a complex field, with heavy discussion points on how to ensure results are acted on appropriately, managing the possibilities of false positives and negatives, uncertainties around penetrance and expressivity, and with appropriate concern of data misuse. Along with the high costs of sequencing, the subject does not seem designed for entertainment. But this is only the case when dealing with health results and data storage where the stakes are high.

I propose that separate to developments of genomics for improving human health, the stakes could be lowered for entertainment use with an emphasis on social utility. With a plethora of information on the GWAS catalog there is fodder to release a wave of tests that are developed using lean methodologies, with those that establish product/market fit sold at a low cost, and with results delivered for multiplayer user experiences.

DTC genomics needs a new category of entertainment products that do NOT rely on accurate predictions

(Part 1 of 2. To see Part 2, click here.)

The trajectory for direct to consumer (DTC) genomics was looking bright a few years ago. Ancestry testing reached a scale of 26 million consumers being tested, a range of various new products were hitting the market, most notably BRCA testing by Color Genomics, and this culminated in the launch of Helix as an app store platform to provide consumers access to a range of products with just one saliva sample. The promise of DTC products is the ability to access a new and repeatable payment source for the genomics industry, tapping into retail dollars instead of relying on research/block/investor funding or healthcare reimbursements.

But now DTC is going through a quiet period. Color Genomics appear to be investing more into population programs suggesting enterprise markets remain more attractive. Helix likewise has pivoted into similar markets as their original app store business model floundered. Suppliers for collection kits and testing consumables have duly cut their sales forecasts in DTC segments. What happened to the retail dollar?

While demand for ancestry waned, none of the emerging genomic tests took a strong enough hold in their respective markets and crossed the chasm to become mainstream products. There are broadly two types of these tests. There are tests backed by reasonable evidence, including tests for pharmacogenomics, “incidental” findings (though this term doesn’t fit in the DTC context), and reproductive carrier screening. But many of these tests have limited product/market fit in a DTC setting as customers mostly continue manage their healthcare through traditional pathways. Companies in this space have tempered expectations, choosing to focus more on traditional payers. There are also tests where there could be a stronger product/market fit such as reporting on fitness and nutrition optimization, however the results provide limited utility and importantly the evidence for these tests lack strength. These products have mostly been shunned as an embarrassment to the genomics industry.

I propose that until DTC healthcare takes off, the next wave of products need to be more like ancestry, where less perfect science can still achieve product/market fit. Consumers have taken DNA tests out of curiosity, sometimes a bigger motivation than health reasons. They also look for ways to connect and engage with others, both in person and on social media. The “job to be done” assigned to a DTC product may not necessarily be to help consumers live more optimized lives, but to be entertained and connected.

This is exactly what ancestry testing provides. The tests may be inconsistent with hilarious anecdotes of its failures, but ancestry testing delivers discussion and entertainment, on a subject that is important particularly to Americans, and goes beyond giving back a “report” to something that help their customers connect with others. (Their potential relatives.)

The new wave of tests likewise need to go beyond providing a report that has value in its prediction capabilities, instead focusing on delivering entertainment value. This allows an entrepreneur to dip into the GWAS catalog that lists over 300 publications in non-health related areas. To be clear, all these publications need further validation, but I propose that some fraction of these could still be used for creating novel entertainment products. Presenting these results can’t be in the bland manner of a private report as existing DTC products do, but will need to be much, much more creative.

The results of these tests shouldn’t be consumed in isolation but should be used in areas of gaming, expanding social influence, or connecting people over what they have in common. Most importantly, it needs to create discussion points that are totally unrelated to health. I’ll expand more on this in Part 2 of this 2-part series, with an article titled “The next wave of consumer genomic products need to be cheaper, leaner, and multiplayer”. It will no doubt attract condescending smirks from those who think genomics should only be based on well-validated science. But given how the DTC market has floundered, some original ideas are warranted.

Congratulations, you bought the latest NGS system for clinical sequencing. Now what?

Congratulations, you got a new Illumina sequencer! 

You work for a run-of-the-mill lab and institution that normally focuses on research or traditional genetic tests in the clinic, and you’ve now been given money to set up a genomics lab and bring in new revenue.  Your boss and the board have heard that clinical genomics is the future and you are to lead them there.  You already had a lab. You just needed a sequencer so voila you’re in business! You just need the ancillary equipment, and Illumina has thankfully given you a list of the most expensive suppliers to set you up. And last but not least, the clincher for the business case (that was never written) is that some local clinicians have said they’d love to support your lab.  Every now and then they have a patient who should get their cancer sequenced, or are diagnosed with ataxia, Brown syndrome (you had to Google this one), or telekinesis. That’s the beauty of genomics, one lab can get you access to so many different markets.

You know that buying big shiny new instruments is similar to constructing new buildings, neither produce meaningful output on its own.  But both help attract new staff, and thankfully you have some budget for this.  So you go about hiring people.  Grab some lab techs and a lab manager, or someone who at least knows what a quality management system is, and a pathologist or equivalent – oh my gosh they cost more than the sequencer itself!  Don’t forget a bioinformatician, or two, or three, you never really know how many you’ll need.  Consultants will come knocking on your door – they are expensive too – but you should at least entertain them, they’ll give you attention and flattery.

Now put together your quality management system if you don’t have one already and perform your validation for accreditation.  That’s easy huh?  Only that it isn’t if you’ve never done clinical genomics before.  Then you realise how different the somatic and germline pipelines are really going to be. Your bioinformatician spends all day either on some internet forum with weird avatars or on a black screen with colored text, so you don’t know if they are really solving your problem or communicating with a Computer.  Now you realize why companies like Sophia Genetics have succeeded, buying their kits and data/variant analysis licenses are really just an excuse for you to have a someone who knows what they are doing provide you samples and a template for your validation report.

Time passes, you keep yourself busy learning about the umpteenth iteration of the ACMG guidelines, you form opinions on whether copy number variations, gene fusions, methylations, and extrapolations are important variants to report on, and as this is starting to get complicated you attend some patient support groups as well as an AA.  By the time you’ve recovered, your lab’s been inspected and you’ve got accreditation.  Congratulations again!

Take some photos and videos.  This helps create marketing stock images for the instrument manufacturers.  They’ll interview you for a customer testimonial and invite you to their user group meetings, of which you get nothing out of, now that travel isn’t allowed.  Have fun showing stakeholders through your new lab.  Since the samples are bland and colorless, the sequencer itself doesn’t move, and genome browsers and databases look so 1990’s, you just leave them ogling at the pipetting robots.  And let them put on some multi-colored disposable gloves.  Speaking of gloves, I hope you remembered to throw out those blow-up glove dolls, taken down the posters of the DNA codon table and periodic table of elements, and your whiteboard of consumables to order.  Your lab’s supposed to look “clinical” now!

Come up with a name for your professional-looking lab.  Maybe your institution will allow you to use theirs – just add “clinical”, or “labs”, or “diagnostics”, as a prefix or suffix.  But maybe they aren’t sure if you’ll damage their reputation so you need to come up with your own name.  Research has shown that there’s a rapid increase in the number of registered company names including “genet-” or “genom-” in the name.  Cool, let’s try that.  You go with the creative GenetLabs.  Oops, that’s taken.

Time to launch.  There’s tens of millions of cancer patients in the world, and 1 in 20 people have a rare disease.  These people need your lab, so you tell your staff.  You all brace yourself for the flood of samples that’s about to come in.  If things go well, you could run a discount program because it fits your ethics, a bit like what Invitae does with its partner programs for a random selection of disorders.

Only that the samples don’t come in.  For all those cancer and rare disease patients out there, only a thumbprint’s size of that group actually fit under reimbursement criteria.  Furthermore, a large number of cancer patients actually get better through standard of care, and most people with rare diseases get on with their lives without needing a diagnosis.  Come to think of it, you have hundreds of connections on LinkedIn and not one of them have a rare disease unless Asperger’s counts now.  How is it possibly 1 in 20?  Hey, those genomics papers you were reading up on didn’t tell you this!

The sales/marketing/commercial/product/business manager guy/gal that you hired (really late in the process, as the turnaround time from their start date through to results was expected to be quick) reminds you that all along they’ve been needing a team.  You finally believe this person.  Maybe you should have listened all along, this was the only person in your company that showed up with corporate clothes.

You hire product managers and they fuss over your estimates of the TAM/SAM/SOM.  It all seems rather academic, and given that you’re probably an academic, you think that’s a poor expression.  All you care about are orders received, not a sketch of the customer journey.  You hire marketing managers, literally on a first-in-best-dressed selection criteria.  First, they tell you to get a proper logo for your company, not the one you crowd-sourced.  They helpfully suggest that you need a DNA helix in it.  Secondly, they bemoan how they were the last to be hired when they should have been the first.  Not very helpful.  In any case, they aren’t the last to be hired; the sales reps are.  These are either young scientists and genetic counselors who didn’t enjoy practicing what they actually studied for, or grizzled veterans who won’t use your CRM because they can’t touch type.

Illumina at this stage tries to sell you more sequencers.

You start to figure out the market.  The Genetic Testing Registry lists many hundreds of labs in this market, and it’s definitely not updated.  And these labs all use the same technology as you.  But you’re different.  Your staff assure you that your lab provides higher quality, as if the other labs don’t make that claim too.

You’ll also do a data play.  You also know about block chain.  And genetic counseling chat bots.  An Elon Musk poster goes up in the lab.

As you continue to figure out the market, you realize that Invitae are doing what you do, only better.  That’s because they have so much more money to burn that you ever had.  Ah, so that’s why they are able to provide those freebies.  So you write what would be their business case to your boss and the board and they turn it down.

Since you’ve got no access to modern monetary theory like Invitae, you now have to look at other strategies.  Foundation Medicine was acquired by Roche for billions of dollars.  Billions!!  Just think of the ROI!  But they actually had some unique technology.  So they had spent even more time than you did prior to accreditation.  You don’t bother writing up a second business case, because you need to make money now from that sequencer, not after a few more years of RnD.

At this stage, it looks less likely that you’ll be part of the routine healthcare system.  You know, the routine healthcare system where a patient in your neighborhood gets ill, they see their physician, who even though has never gone beyond learning G, C, A and T at school, knows the difference between germline and somatic cancer, is now providing genetic counseling, advice on how to navigate the various panels available, and is willing to make and take multiple phone calls and forms from payers, ordering a test from your lab and explaining variants of unknown significance or a clinical trial on the other side of the country, and telling their colleagues to do the same.  Actually if you wanted to be part of the system, you should have set up a PCR for COVID-19 testing.

No, you’re not part of that system.  You’ll have to fight over other genomics markets that are looking less like your original plan.  There’s population sequencing, research grants where clinical results are provided back, screening / direct-to-consumer testing (you haven’t yet found out how much it costs to advertise on Instagram), and maybe you can outsource out whatever you’ve come up with in your lab – at the very least the sequencing!  By the way, the list of sequencing providers is even longer.

So while you aren’t broke, this is all pretty disheartening.  Illumina have now released the SuperSeq, which gives access to the $599 genome.  But you’ll have to pay for the new machine, which they don’t even give you a trade-in option for as their reps have made their bonuses for the year already.  To add insult to injury your reps go over and work for them, as do your bioinformaticians.  Around about now, Ion Torrent comes calling.  They are cheaper, but you’d have to start all over again.  And you might not ever report indels but at this stage you just don’t care, and you’ve taken down that poster of Elon.

If the proverbial hits the fan and you do go broke, just remember that some years ago, one of Illumina’s customers allegedly sold their fleet of sequencers on eBay.  Aren’t you glad you took those great marketing photos?

Can germline testing be monopolized by one lab?

I ramble on about all things genomics. If you want to banter, follow me @GenomicsCow on Twitter. Here are some of my thoughts on a very interesting space in genomics: germline testing by next-generation sequencing.

Summary:

• There are several challenges to consolidating the germline testing market.
• Comparisons with consumer tech companies business models are not appropriate.
• There is no technological moat, and pricing doesn’t give a quick win.
• Variant interpretation doesn’t scale well, and a lot of stakeholders could resist this change.
• Providing a full range of pathology services could be the key if this genomic testing becomes mainstream.

The theme around labs like Invitae that diagnose hereditary diseases depend on their ability to consolidate the market for next-generation sequencing of hereditary diseases. By aggregating tests from around the world, this improves their platform to refine services, lower costs, and to monopolize this market. The lab leading the pack is Invitae – so I’ll refer more to this lab in particular, especially as much discussion is oriented around how willingly they lose money to aggressively pursue this goal. Can this goal be achieved?

There is a lot of complexity to their business plan, and the upside is summarized articulately by Ark’s genomics analyst here. Invitae has given its critics reasonable measures to track their success, by reporting volume and COGS routinely. By looking at these charts, it appears that Invitae is executing successfully, even with blips caused by COVID-19. So, if the market can be consolidated, they’d be in the pole position to benefit.

(Taken from Invitae’s 2020 Q3 earnings presentation)

However, these measures are insufficient on their own. Other labs like Fulgent Genetics, and privately held labs too, are growing their volumes and reducing their COGS too, albeit at a lower scale. More importantly, these charts don’t tell you whether the market is monopolizable, the key outcome of their “go for broke” strategy. I’m looking to see how these 7 challenges are being overcome, one way or another, to show that the market can be consolidated around them.

1. Monopolies are often established by creating or owning a new distribution model

Comparisons between Invitae with platform-oriented tech companies are rife. The reason for the dominance of tech companies has been described by Stratechery; in a nutshell a key component is to find new ways to control distribution, often by getting closer to the customer.

I don’t see this pattern with diagnostic labs, where little has fundamentally changed with how any lab reaches its various customers. The current distribution model would change if direct-to-consumer testing takes off and drastically changes how healthcare is delivered. This is one reason why Invitae, and seemingly, every other lab is considering their strategy here. However, the paradigm has not meaningfully shifted yet.

While there are other similarities between Invitae and tech companies, by not sharing this crucial component, I believe comparisons are misleading. Until a lab can structurally change the way it reaches its customers to give it at a sustainable advantage, it won’t monopolize its market the same way that other tech companies have.

2. It is difficult to achieve a technological advantage

Labs generally do not need to change the business model. They instead need to have a technological edge. But unlike other genomic tests where new methods are unique to a company, germline testing relies on open-source information. It still takes high levels of expertise to produce diagnostic reports, but this barrier is not insurmountable, as described in more detail here. For example, one lab could be switched for another to perform a test. The second lab, if credible, should be able to provide the same result back. While there is variability between labs, this has more to do with inconsistencies in using variant interpretation guidelines rather than a technological difference between competitors.

Invitae’s advantage isn’t technological but instead relates to their vast infrastructure and deepening connections across the market. This is a strength, but it is replicable.

3. Pricing doesn’t wipe out competitors in the short term

Due to the first two points, a lab that wants to monopolize the market would want to aggregate as many samples as possible to drive down costs, which Invitae has succeeded at.

There are various price points for tests. Institutions and insurers, who, according to this analyst, make up the dominant proportion of payers, are going to be attracted to the lower prices that are offered. These contracts are renegotiated over time presumably to the advantage of a well-priced vendor, but it is not as dynamic as a transactional market where prices are decided on a day-to-day basis. In the latter scenario, a company like Amazon can choose to ruin a particular competitor in the short term based on driving prices down and withstanding it.

A number of labs have folded, and Invitae can crush more of their competitors with its pricing, it will just take time to play out. Notably, some of its competitors are not exactly strapped for cash, as they also can tap into similar or other sources of funds, and recently, even by performing COVID-19 testing.

Invitae will be willing to wait this out. But as they are well-known for its cash burn, it would be good to have some certainty on a significant number of their competitors will fold.

4. Variant interpretation doesn’t scale alongside other components of testing

While waiting for competitors to fold, any lab aiming to maintain high growth will have to take on a larger range of tests across all hereditary diseases.

Next-generation sequencing for hereditary diseases is premised on volume, where larger numbers can reduce cost per sample. Samples can be collected, prepared, sequenced, and analyzed in larger and larger batches, providing economies of scale. But variant interpretation, the key medical component of their test, is not as easily batched. While software and intelligent use of databases have improved automation, selecting variants for reporting is still a highly manual process.

So far, the more well-known diseases to report on have been driving the growth of this market, such as familial cancers. One could argue that interpreting cancer variants is streamlined because their genes and variants are well annotated. Therefore, a lab that focuses on just these indications is able to scale.

But, for more complex and rarer cases, especially where there is no clinical diagnosis, the time taken is much longer due to clinical ambiguity and less information out there for the genes and variants involved. Historically, these cases have been managed by centers of excellence that each focus on particular disease areas. Each indication takes experience and feel, developed by a curator or lab who, with time, just gets to know the disorder and its genes.

This is troublesome for a lab like Invitae expanding across these broader disease areas. Their case mix will go from being concentrated on a common family of diseases to a diluted stretch of diseases with lesser similarities, akin to moving into the “long tail” of a market. To maintain their high quality levels, the average time to curate variants could take significantly longer, disrupting their ability to manage the increasing volume cost-effectively.

Because of this, I am uncertain how Invitae will continue to scale and drive down costs as their caseload in hereditary testing expands.

5. A lot of smart people lose if one lab consolidates the industry

Is it in everyone’s interest to centralize testing? Certainly, not for local lab staff around the world, nor for Illumina.

The germline testing market has often been a paradigm of individual, local labs supplying to their neighboring referrers. They have built up these relationships over the years. This isn’t transactional. It has resulted in research collaborations between lab and clinic, and some professionals have studied together and known each other a long time. In many places, testing is outsourced to labs like Invitae, but often with a view that this arrangement would be changed the moment the technology is simpler and more cost-effective to set up locally.

If this market consolidates, the efficiencies will cut down lab jobs; losses will be felt by those who have spent over a decade in training. These stakeholders form incredibly strong advocacy groups. Doctors and their medical societies are quite unlike the taxi drivers and retailers whose livelihoods were upended by Uber and Amazon.

These local labs have a strong ally, however. A non-consolidated market has also been beneficial to another key stakeholder: Illumina. Illumina supplies to each of these labs, and it benefits them to keep things decentralized. They watch instrument utilization rates to see if there is “too much centralization of the market“, allowing them to sell more instruments and a higher number of reagent kits per sample sequenced. It may not help their cause if a single lab is getting its samples shipped from all over the world, because it ruins their ability to sell instruments to local labs. Illumina will remember their problems years ago with the Beijing Genomics Institute, one of its largest customers at the time, giving great short-term sales results, but made it difficult for Illumina to maintain control over the market and empower other customers. If any single lab consolidates this market, Illumina will lose a lot of its bargaining power.

Because Illumina has the scarcity advantage, it would be an oversight if they allow Invitae to grow to a point where Invitae can dictate terms to Illumina, rather than the other way around. Instead, as the market grows as discussed in the next point, Illumina can develop their products to empower a vast network of local labs to set up tests with ease and run them cost effectively.

6. Esoteric tests should be centralized, not everyday tests

These first few barriers are premised on genetic testing for hereditary diseases being a relatively small and highly specialized market. But will it always be small?

Many people believe that genomic testing for inherited disorders could become a regular pathology test, performed as a first-line diagnostic test or for even for screening. If current trends continue, these tests will not remain esoteric.

As a test becomes mainstream, more local labs will want to perform the test too. The disincentive has been the low volumes and also the effort to set up laboratory-developed tests, commonly referred to as LDTs. LDTs for germline testing are particularly hard to set up due to the high levels of expertise required.

But I speculate that as the market grows, these tests could become available as in-vitro diagnostics (IVD) tests, making it easier for inexperienced labs to enter the market. Illumina has deployed a strategy of helping labs set up non-invasive prenatal screening through creating an IVD product, enabling a wider range of customers to enter the market. My guess is that they could do this for hereditary disease testing too.

For post-sequencing, they have purchased or developed capabilities to reduce the burden around bioinformatics analyses and variant interpretation. For pre-sequencing, this process is currently labor-intensive, but one day, Illumina could possibly develop cartridge-oriented kits seen with many other clinical assays.

If a true end-to-end IVD workflow is developed, the barriers of entry become less. This potentially unwinds the effort that large, centralized labs like Invitae have put in to set up their infrastructure and gives a hand-up to local labs that can now provide this test to their existing referrers.

If the market grows and Illumina succeeds in developing this type of IVD workflow, this could keep the market fragmented, and with Illumina continuing to supply to each local lab. Think about the chemistry market and the fully automated analyzers provided by Roche; and now, imagine if Roche didn’t have to compete with Abbott and Siemens. That vision should provide a lot of incentive for Illumina to commoditize genetic testing, but where the instrument vendor maintains power.

7. To take the market, labs need to provide convenience in pathology testing

If these tests become mainstream, how will the “late majority” customers be reached?

(Adaption of Geoffrey Moore’s technology adoption curve.)

Like many products, customers for germline testing might be distributed along the bell-shaped technology adoption curve. There are those who will purchase a product early, followed by the early majority who are more practical, such as waiting on guidelines and reimbursement.

However, this only takes germline testing to half the market. To capture the late majority customer group, using the test will need to be convenient. This could mean that a doctor should be able to work with one vendor and bundle it with other pathology tests for their patient.

This is in the realms of mainstream labs providing a large range of pathology services like Quest and LabCorp. Understandably, these labs are slower to establish cutting-edge genomic tests. But expect them to develop capabilities, as Quest did with its acquisition of Blueprint Genetics. Blueprint Genetics did not have a particularly dominant market position, but presumably, they were attractive for their testing pipeline that could be plugged into the networks of this giant corporation. It may be the case that, when genomic testing for hereditary diseases becomes commoditized, these are the labs that could benefit as they integrate hereditary genomic tests alongside their already expansive reach and test menu.

Conclusion

These are 7 challenges for Invitae in consolidating the germline testing market.

I have immense respect for Invitae’s infrastructure and efforts in the market. But until someone else can explain to me why these challenges don’t impact their ability to sweep the market, or how they are being overcome, I remain neutral about whether they can achieve their long term goals.

I should add that Invitae has now moved into somatic testing and non-invasive prenatal screening, all well beyond the scope of this write-up. But, in my opinion, any success in these markets does not meaningfully address these challenges for germline testing.

If you want to banter with me, follow me @GenomicsCow on Twitter.