Is this the end of the booming genetic genealogy industry?

I have never hidden my opinions or mellowed my criticism on the genetic genealogy industry, even though I am a part of it. I designed the first dedicated microarray (chip) for genetic genealogy – The GenoChip. The GenoChip was designed for National Geographic’s Genographic Project. I also designed the online tests. National Geographic cashed in heavily from this project, but they were too greedy. They asked a blogger to revamp the test (essentially update my original test a bit) and continue selling it through Helix. It is noteworthy that Genographic made much of the data available for the scientific community (pending on signed consent from the users) and donated a small fraction of the revenues to support indigenous communities. They also REFUSED to share the data for profit to the extent that they did not advertise new products to their customers in the hope of gaining their trust. It was not the best strategy. The Genographic Project was the first major DTC service provider to shut down around this year. RIP.

It wasn’t long after we launched the Genographic Project that other companies realized the potential of the ancestry business, not necessarily to study population genetics but to collect valuable genomic data cross it with phenotype obtained through endless questionnaires and sell it to drug companies for profit. 23andme was the first, followed. Their goal was to GAIN OWNERSHIP of your DNA:


From here. The public didn’t get the ploy. They did not get the science (genetics can be a bit tough). They just wanted to hear that they are 2.5% Jews, and 1% Native Americans, and 5% black to get a new identity that they always wanted without understanding the cost or the process. They wanted to be fooled, and those companies were there to do that. When FamilyDNA and MyHeritage figured it out, they joined the medical game along with the slow dimwitted competitor from the UK, Living DNA. During this time the prices for test kits and results fell from $999 at launch to $399 in 2008 to $99 in 2012. The companies promised whatever all charlatans always promise: health and wealth – for themselves, of course, not you.

23andme - headline

The FDA shut down their health business for a while for making false health statements, but they bounced back with a different set of misleading health claims.

In 2014, I developed the Geographic Population Structure (GPS) with my colleagues. The paper describes an accurate algorithm for bio localization (From DNA to Geography) for people who have four grandparents of the same origins. The paper is still #2 most-read paper in Nature Communications. This is how GPS works:

Even 6 years after its publication it is still the most accurate bio localization algorithm:

Figure 1

From our draft paper.

In 2015, DDC licensed the rights to GPS Origins. I designed the DREAM chip for them and the algorithm of GPS Origins, an advanced version of GPS capable of tracing two parental lines and calculate their migration routes.


GPS Origins remains state of the art in the field and has been used by the industry and scientific community and, of course, the general public. Five years after it’s creation GPS Origins is still the best selling test of DDC and I get reports about its performances alongside user feedback, some of which I published in my Twitter (pending user consent, of course).

Let us get back to the genetic genealogy market. By 2015-2016 competition became fierce and it only became worse since then. 23andme and dumped the price of the DNA kits and gambled on the volumes of kits that they could sell. The sequencing cost is about $35. Those companies would usually charge $50-60 and up to $100. Those very narrow margins were supposed to support fancy BioTech offices in California, expert personnel, generous salaries and benefits and of course profits. Tough right? The next thing that these companies did was to invest a lot more in advertising. Remember all the “success stories?” families reunited? OMG, I found my Jewishness; Holy baby Jesus, I found my Irishness. Holy cow, I found my Indian brotherhood and so on and so forth – many of these were paid and bought for. In terms of the Theory of Games, these companies played the worst possible scenario. No cooperation, throat-cutting price competition, and pretty much the same test results. With all the haste and rush to out-compete one another, these companies forgot something important, the customer. The results that people got were pretty much the same as they got five years ago. The tests didn’t keep up with the science, because again, none of the new companies was truly interested in ancestry. It was only a way to lure customers to give up their most precious treasure – their DNA data and, no less important, the annotation gained through hundreds of questionnaires.

Eventually, the inevitable happened and

In November, typically a big month for Christmas and other holiday gift buying, Ancestry’s US website sales were down 38 percent while 23andMe’s were down 54 percent compared to a year earlier, according to the data. The declines continued in December, down 15 percent for Ancestry and 48 percent for 23andMe, marking a year of declines.

From here

Interestingly enough, it happened to the entire industry at the same time, again, one of the prediction of the Theory of Games. When you play in a certain way, sometimes everyone lose. 23andme had a lot more to lose. They are NOT profitable, they own billions of dollars to investors and they have no other revenue streams (except selling your data). They HAD to get the DNA data so that they can monetize on it. are in a much better position. They already have a very high income from their genealogy business ($1B a year in revenues), but they had major competitors like MyHeritage and needed to diversify. Most importantly, they wanted a big IPO, but things did not go as planned. The planned 2017 IPO was postponed to 2019 and would probably be postponed again.

The drop in sales was followed by down-scaling the businesses. “23andme will fire 14% of its staff (100 people)“. In 2019, Forbes reported that:

Nearly 5 million customers did so last year alone, generating an estimated $475 million in revenue for the company, which has yet to turn a profit. It’s also made Wojcicki (No. 33 on this year’s list of Richest Self-Made Women) worth an estimated $690 million, almost entirely from her roughly 30% stake in 23andMe, which is valued at $2.5 billion by investors.

475M/5M is $95 per person. Recall, that the company sells an ancestry kit (official price: $99) and a health kit (official price: $199), they later added a rich-people kit for $499. Confusingly, enough reports that the number of kits that 23andme sold is twice higher, 9M kits. Obviously, someone doesn’t want us to know the exact numbers, so let us go with the lower estimate of 4.5M. The reported decline of 50% equals 2.25M kits. Let us now assume that 23andme has a profit margin of $10 per test (after the cost of test, labor, office equipment, and A LOT OF advertisements). They lost 2.25Mx10=$22.5M. For a company that is not profitable yet it stings. Would it help to fire people? Let us assume that these 100 people make $200K/year (salary, benefits, and health insurance). 100x200K = $20M. At the same time, Ancestry also fired 6% of its workforce. Other companies should follow up with the firing since the decline should be common to all of them. Now the question is what is going on?

Both 23andme and companies blamed the decline on privacy concerns. This is quite a lame excuse. Had it been a major concern, FTDNA wouldn’t have made their data available to the FBI. FTDNA reported that only 2% of its customers left following this action. I was one of those who left. In an era where people share everything online, it is unlikely to be the major reason. It seems more of an excuse for their investors. Some people think that they have exhausted the “early adopters” pool. This may be so because the decline was throughout the market, but evidence for that has not been produced and recall that most companies provide very similar results. I believe that there are other reasons for this trend:

  1. Low accuracy. DNA tests were never able to overcome their inaccuracies. This has been reported by myself and other researchers as well as public members (see: same parents different ancestry)
  2. Data manipulation. Companies continuously tweaked and fixed the results to please (some of) their users. I reported a case where 23andme was caught red handed fixing the results for Jews following an article of the Forward. This was my most-read article in that year.
  3. Ignoring criticism. Companies kept their algorithms secret. No one had access to these black boxes. 23andme even refused to an external analysis of their algorithm on genomic data, I reported it in a paper. I am always laughing when people contrast my findings (on hundreds of samples) with their personal results (sample size=1). For science to be done right you need the numbers. You don’t have it and it is unlikely that you will ever have it unless you will do a proper time-consuming and costly genetic study.
  4. Dishonest scientists. All the large companies hire scientists as consultants and pay them. Those scientists then promote the services of those companies directly or indirectly. They never declare a conflict of interest on their scientific papers, as they should. For some scientists on the the board of FTDNA, is perfectly normal to work for FTDNA and write papers glorifying the Jewish genome (sometimes using FTDNA samples) without declaring a conflict of interest. Here is another example of a scientist appearing in a commercial of Living DNA who never claimed a conflict of interest on his papers. They are hardly the only ones. By the way, Living DNA claims to work with hundreds of scientists. I doubt that there are hundreds of human population geneticists in the word and that a small UK startup can afford to pay them all (conveniently enough, Living DNA never listed their names), but you can see how companies use the names of scientists to authenticate their products. In parallel, the same scientists URGED everyone to SHARE THEIR data for their greater good, of course. In 2019, the Sanger Institution has been accused of misusing African genomes for profit.
  5. Charlatans and so many of them. In some ways, genetic genealogy is the perfect business. Every product is the same but everyone gets different results. It is nearly impossible to know if the test works right or not, let alone, denounce the test as dysfunctional because it COULD HAVE worked well for other people. This unique property of this business allowed the existence of so many charlatans as well as bloggers and other self-anointed “experts” who “recommended” this or other tests in hope to cash in on some of those millions of dollars industry. A few of them got the IBD concept and even fewer managed to make a career out of it. But those were a handful of people. The most common way of making money was to advertise one of the large companies. Do you know these “10 best DNA tests” websites? The first one was a clever idea but then more popped up. They are not a charity organization. Drop them an email sometimes and ask how much it costs to be #1. Do you know those YouTube videos of people opening up their DNA test and start screaming and crying because they are 2% English? They usually have hundreds of thousands of viewers – ask yourself when was the last time you screamed and cried when you opened a letter that didn’t tell you that someone you know died or you lost your home/job? Who dresses up and puts on make up before opening their mail? I know I don’t. I never get anything good. It’s either a bill or an advertisement or something else in Swedish that challenges my linguistic skills at a time that I want to rest.
  6. Stagnation. With the narrow margins, companies did not have the cash to invest in innovation. Developing algorithms is extremely challenging and requires knowledge from multiple fields. Computer science, Math, Statistics, Biology, History, and Archaeology or Linguistics. Not a lot of people can do it. The costs of major genetic tests in the magnitude of Family Finder, GPS Origins, or Chromosomal Painter can easily reach $250K just in the first year. With the money going towards finding new genes for being a morning person or a night person or whatnot, it is not surprising that no new tests are being developed. Eventually, the public got bored and people stopped recommending their friends to buy these kits.
  7. Privacy. Of course, this is very important and the only thing that these companies DID NOT and COULD NOT guarantee.
  8. Bad business model. The idea was that crossing the DNA with personal information would aid in the identification of genes that would be as profitable as the BRCA ones. To the best of my knowledge, it did not happen yet. The problem is that the DNA is only a small part of the picture, and the DNA data obtained from chips is even smaller. Going back to the people to get more details is probably very difficult or unfeasible so the whole concept of let’s keep going until we’ll find another BRCA is deceiving.

So, what now?

Currently, there are over 200 ancestry companies in the world. Most of them sell the old generation tests, i.e., STR and low-resolution haplogroups wrapped up in some fantastic historical story that has nothing to do with the test. Those companies are typically very small, they charge a lot, do not sell too many kits, and may or may not require their owners to hold another job. They would survive. There are very few large companies that need to sell a large volume of tests to pay high salaries and hold expensive offices. They would continue to struggle. Firing all the ancestry team is like taking out the gas tank of the car to reduce its weight and make it go faster. It won’t work. Unless these companies have identified their new cash cow gene and are restructuring their business, they need to rethink their strategy. The solution, in my opinion, has to be investing in innovation, develop new tools, and change the business model.

Two companies would probably be the immediate victims of this looming catastrophe. that planned an IPO on the background of the “booming DTC industry.” This is not happening anymore. would have to slice their rosy projections that they had for 2023. The second victim would be Living DNA, the charlatan UK company known in telling everyone that they are of English heritage. The company has one product and operates within a troubled market. Considering their customer complaints and the level of service that they provide, I predicted months ago that they would be in trouble.

Update to this post

BlackStone, a private equity copmany, bought for $4B and it’s entire database. BlackStone owner not only supports Trump and provides him with information obtained from the company they owned but are also accused in aggressively evvacuating people out of their homes. And now they have your DNA.


This entry was posted in BLOG and tagged , , , , , , , , , , , , . Bookmark the permalink.

3 Responses to Is this the end of the booming genetic genealogy industry?

  1. frank says:

    Just for interest a couple of years ago I decided to waste some money and sent a cheek swab to livingdna who then sent me results of their analysis. I gave them inaccurate information of my birth place and age to see what happens. I was born in north west iran and consider myself totally iranian (infact mede as opposed to purely persian) on both side of my family. The results showed me with ancestors from central asia (which didn’t overly surprise me as my father’s ancestors were turkic) however, what was a big surprise to me was the over 32% ashkenazi jewish dna on my mother’s side. Her family lived in the mountain towns of the Alborz region and still do. I always thought this dna business was a bit hit and miss and just happened to chance on this site and find this article interesting.

  2. frank says:

    …also interesting bit on the title of this site, i haven’t looked at the site in depth but the farsi for the caspian sea is darya-yeh khazar….

    • eelhaik says:

      Thanks for your comments. If you’ll keep digging in you’ll see why you came up as AJ on this test. Essentially, Caucasus and AJ ancestries are very similar due to shared origins.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s