Shopify Is A Massive Churn Factory And Is Going To Be Hit By Facebook Hard. Price Target $60

Summary

• Shopify customer churn is monstrous, and we will show it, and hence, SHOP has no way to deserve an EV/REV multiple like other SaaS companies.
• Shopify is buying its massive revenue growth through acquiring very large quantities of low quality customers.
• Shopify will not be able to improve its operating leverage by cutting its massive sales and marketing spend. Otherwise, growth would grind to a halt.
• Facebook is going after low-quality stores advertising in its news feed, and we believe this will have a significant impact on Shopify's GMV and revenue.

[Update - July 1, 2018]: On Friday June 29, we discovered a mistake in the second of the two churn models we used in this report. Because of our conscience we believe it is correct to issue a revision that can be found here. Based on correcting the mistake, we now believe that 77% of customers churn off before reaching their first anniversary as opposed to 85% in the original report. Also, our estimate of customer lifetime of newly acquired customers goes from 14 months to 15.58 months. We also raise our price target from $60 to $65.

Summary

Shopify (NYSE:SHOP) is a high-flying SaaS e-commerce company focusing on the SMB segment which floated in May 2015 at $17 and currently trades at $153 as of June 25. It hence boasts a $16.27 billion market cap and an Enterprise Value of $14.77 billion (it has $1.5 billion in cash), according to Yahoo Finance. Shopify has become a Wall Street darling due to its hypergrowth, showing 65% revenue growth as recently as its last quarterly earnings on May 1, 2018. The company is guiding to revenues of $1 billion this year and is trading at a sky-high EV/2018 rev multiple of 14.77x, according to Bloomberg. The majority of Shopify’s revenue is not subscription license revenue but merchant solutions revenue with a much lower gross margin. It isn’t profitable and spends 61% of its gross margin just on sales and marketing. Wall Street expects Shopify to eventually become highly profitable by cutting S&M spend and spreading its R&D and General & Administrative costs across a much larger revenue base with revenue nicely stable and predictable due to its SaaS subscription nature. Shopify management have been very skillful in supporting this picture.

In this report, we will attempt to prove that Wall Street has gotten Shopify completely wrong. We think that Andrew Left of Citron Research was right about many of the allegations he made against Shopify in his report. He just really didn’t have the data to back them up to be undisputable (but hats off to his intuition). We think we do, and we will also show much more. We will prove our investment thesis with myriad pieces of data nobody has ever published before and show that Shopify’s true value is nowhere near its current market cap. At this point in time, we are quite in the contrarian camp on Shopify. Out of the 28 analysts covering Shopify, 27 have a BUY or HOLD rating on the stock and only 1 has a SELL. We believe Wall Street will take note of our findings, and the stock should start re-rating very soon. These are our key findings for which we will show evidence further on in the report:

• Shopify customer churn is monstrous and hence SHOP has no way to deserve an EV/REV multiple like other SaaS companies. Spoiler: We believe that 85% of newly signed up paying clients are likely dead within 1 year.
• Shopify is buying its massive revenue growth through acquiring very large quantities of low quality customers (basically your average Joe type that happens to fall for various make money online schemes). In the case of Shopify, they are sold the dream that they can make substantial amounts of income from drop shipping items from AliExpress with a 5x markup. Only a minority of newly signed up clients are businesses that are already actually selling something. We believe a surprisingly high proportion of SHOP’s revenue comes from clients that are less than one year old. Spoiler: Through an analysis of the .com zone file, we have found that 753,000 .com domains are hosted by Shopify. 405,000 of these domains are less than 1 year old! That’s 53.7% of all their domains.
• Shopify will not be able to improve its operating leverage by cutting its massive sales and marketing spend, currently running at 61% of its gross margin, because its revenue growth would massively decelerate. This is Shopify’s deal with the devil. As opposed to standard SaaS businesses, which spend a lot of S&M money to acquire a client and then the client brings a constant stream of cash for a very long time and has a low likelihood of churning, Shopify, on the other hand, spends a lot on S&M on clients that have a very short average lifetime. Our analysis shows the average lifetime of a new Shopify client is about 14 months! The Shopify growth model is heavily dependent on a constant inflow of large amounts of new clients. What happens when this inflow eventually starts drying out?
• The number of Shopify paying customers is plateauing. Our research shows that the number of net adds of domains hosted is already very low. Shopify’s problem is that it needs to bring in large quantities of new paying clients just to replace the clients that churn off. Eventually, there will be more customers churning off than being added. Spoiler: Through an analysis of the .com zone file we have analyzed the number of domains that stop being hosted by Shopify. This means either the domain expires or the domain is redirected elsewhere. In either case, it is obvious that the owner of the domain is no longer a Shopify client. Out of a total of 765,000 (please note this report has been in the works for a long time, so total number of hosted domains may vary in different parts of the report) domains hosted by Shopify, 2,309 domains churn off from Shopify every single day on average!
• We believe a significant majority of Shopify customers are not doing very well and are hence very likely to churn. We have run the entire list of domains hosted by Shopify through SimilarWeb and found that only 123,000 of the 765,000 domains are showing ANY traffic at all.
• On June 12, The WSJ reported that Facebook (FB) is offering a new feature that will let users rate shopping that stems from Facebook ads. Facebook will then start banning low quality stores. This will have a very significant impact on Shopify’s GMV and Merchant solutions revenue because a huge slice of GMV is tied to Facebook and Instagram ads. We believe that there is a massive clean-up in Facebook ad space coming and Facebook will make a move on low quality drop shipping stores on Shopify. Spoiler: We have analyzed the top 250 most visited Shopify stores representing 162 million visits in May and found that possibly up to 50% of this traffic could be at risk in a true Facebook cleansing scenario.
• Looking at Shopify’s 250 most visited stores, we have found that 29 have very suspect traffic patterns. Historically, they have been showing 0 traffic and then suddenly in April and May, they shoot up from 0 traffic to several hundreds of thousands or even millions of visits per month. We use these stores to illustrate how fickle some Shopify revenue might be. Here’s a concrete example:

Source: SimilarWeb

• Shopify has been focused a lot on its Shopify Plus offering, which is a solution priced at $2,000+ per month. Shopify management has been spinning this as new revenue from blue-chippy clients with enterprise grade qualities. Our research has found that a significant chunk of these Shopify Plus clients are not those A-class businesses. For example, Fleshlight.com (Shopify’s #4 visited website according to our SimilarWeb data) or SiliconWives.com are peddlers of “sexual hardware”, but worse, a significant chunk of Shopify Plus clients are showing terrible Trust Pilot ratings (2* and less) such as ColourPop.com, KylieCosmetics.com or MVMTwatches.com (all 3 are top20 Shopify Stores according to SimilarWeb) with the majority of reviews saying they have not been delivered product at all or they have impossible return policies. These stores could massively be affected by changes in Facebook policies towards low quality advertisers. Since pricing for Shopify Plus is variable, this could hurt Subscription solutions revenue.
• We believe Shopify Capital is behind the significant improvement of Merchant solutions Gross Margin. We however believe that the cash advance portfolio (origination tripled y-o-y) could end up being quite toxic. Our hypothesis is that these cash advances are taken up predominantly by drop shipping customers whose destiny is tied to their ability to promote their stores on Facebook and Instagram. When Facebook goes after low quality store advertisers, Shopify will have a problem to collect these cash advances since it does the collecting through taking a percentage of their daily revenues.

Shopify churn rates and customer lifetimes

The reason SaaS enterprise companies command very high EV/Revenue multiples is due to the fact that their customer churn rates are remarkably small, their dollar net retention rates most often positive (they manage to expand revenues per client over time), and their customer lifetimes are very long. According to Tomasz Tunguz from RedPoint Ventures, a firm with $3.8 billion assets under management, a typical SaaS enterprise firm has a 6-10% annual customer churn. Think of companies like Salesforce (CRM), WorkDay (WDAY), ServiceNow (NOW) or Ultimate Software (ULTI) in this category. With a 10% annual customer churn rate, the average customer lifetime is 10 years. This comp group currently trades (as of June 21, 2018) at a blended forward 7.1x EV/Revenue multiple according to our Bloomberg terminal. Shopify trades at 13.3x blended forward EV/REV and about 16x 2018 revenues. We assume this is because the market believes that Shopify has to have superior growth rates and similar churn metrics like the comp group.

So, in our research on Shopify, we thought a great starting point would be to analyze Shopify customer churn metrics. Typically, with other SaaS companies that is quite simple – you look into their annual reports. With Shopify, it’s different. Shopify only discloses a certain metric it calls Monthly Billings Retention Rate, which it describes as:

MBRR is calculated as of the end of each month by considering the cohort of merchants on the Shopify platform as of the beginning of the month and dividing total billings attributable to this cohort in the then - current month by total billings attributable to this cohort in the immediately preceding month.

Shopify releases this number once a year in its Full Year results and has always claimed that it is higher than 100%. We think this number is a meaningless metric to describe Shopify’s quality of revenue and customer cohorts because it allows Shopify to cleverly hide customer churn. It covers total Customer Billings – which not only include the most coveted license revenue but also Merchant solutions revenue. This covers up the fact that Shopify loses a ton of clients (and the most valuable Subscription solutions revenue with it) through the fact that the surviving clients start generating much less valuable Merchant solutions revenue with low Gross Margins. The disappearing subscription revenue is replaced essentially with low margin payment revenue. If Shopify would publish a metric called Monthly Gross Margin Retention Rate, it would definitely be negative (i.e. less than 100%). That’s why it invented a metric that isn’t. Since we find the MBRR metric as hiding customer churn, we had to come up with our own way of determining Shopify’s customer churn rates. This was quite a challenge, but in the end, we think we prevailed.

So, what did we do? We did a detailed analysis of the .com zone file managed by Verisign (VRSN). The zone file is a list of all registered domains in the .com universe that are active. There obviously are other domains, but .com is by far the most important and most representative. Getting access to the zone file isn’t easy – only domain registrars have access so we partnered up with one

First, we downloaded the entire zone file, which is 134,500,000 .com domains. Then we attempted to find all the domains within this universe that are hosted by Shopify. Shopify has this following IP address range: 23.227.32.0/19. We came up with 753,000 domains. That’s a lot of domains running on Shopify, in the perspective of all .com domains in existence that’s 0.56%.

The first thing we did was to analyze the age (defined by their creation date) of all these Shopify domains. We came up with this:

Age of 1 year or less	405,000
Age 1-2 years	129,000
2-3 years	63,000
3-4 years	37,000
4-5 years	23,000
5 years and older	76,000

Click to enlarge

Source: Our analysis of the zone file

The fact that domains 1 year or younger comprise 57% of all the domains hosted by Shopify is quite striking. These are all new clients for Shopify less a few that point their domain to Shopify and never start paying after the 14-day free trial. It’s important to note that less serious people starting on Shopify do not even bother to operate on their own domain and select to host on a myshopify.com subdomain. Churn for this group is obviously larger than for clients selecting to operate on their own domain name. We believe that pretty much everyone with their own domain name became a paying Shopify customer. Approximately half of new domains (i.e. freshly registered) are registered through Shopify, so all these are paying customers straight away because they make the payment for the domain to Shopify. Hence, with all likelihood the 405,000 domains represent new Shopify customers or past new Shopify customers (a significant amount of them are already dead).

We then proceeded to measure Shopify customer churn using two different methods.

In the first, we identified all domains out of these 405,000 that have an expiration date over a 5-day period comprising of June 10, June 11, June 12, June 13, and June 14, 2018, and watched what happened to these domains. We also split these domains into 3 categories according to their age: 1-year old domains (experiencing their first renewal), 2-year-old domains (second renewal) and 3-year-old domains (third renewal). We ignored older domains because the sample sizes were getting small. This is what we found:

	1yr old	2 yr. old	3 yr. old
No. of domains expiring	4003	1100	658
No of domains running store after expiration	1312	562	381
Survival rate	32.78%	51.09%	57.90%
Churn rate	67.22%	48.91%	42.10%

Click to enlarge

Source: Our analysis of the zone file

When we looked at the 1-year old category, we found that 1 day after the expiration date, only 1,312 out of the total 4,003 had functioning Shopify stores – only 32.78%. The remainder didn’t – the vast majority of these typically showed a parked page (on the expiration date a domain falls into a 40-day redemption period before it is deleted, in which the registrar points the domain to a parking page to make a little revenue from remaining traffic). These are all dead clients for Shopify that churned off within a 1-year time period. Over a 5-day period, the total number of 1-year old domains lost through the expiration process came to 2,691. That’s an average of 538.2 domains per day.

In the 2-year-old category, the survival rate was 51.09%, which means a huge 48.91% of domains still churned even in by the end of year 2. In the 3-year-old category, the survival rate goes up only marginally from 51.09% to 57.09% and the third-year churn is still a very high 42.1%.

Bottom line: Using this method, we have found that only 32.2% of clients survive their first year at Shopify, implying a 0.678 churn rate in new clients over 1 year. If we would solely rely on this method to measure Shopify churn, we can derive the average customer lifetime, which would be calculated as 1/0.678 = 1.47 years, which is 17.7 months. By itself a very low number in comparison to real SaaS companies.

However, we had a hunch that the number of surviving clients might still be too high and the churn too low because we didn’t capture all those domains that were hosted by Shopify and had already left before their expiration date (Us downloading and analyzing the .com zone file just gave us a current snapshot of the situation, not historical data of the past). So, we looked at churn through a second method.

Again, we started out with the total number of domains hosted by Shopify (753,000) and watched how many domains stop being hosted by Shopify on a day to day basis – domains that churn off. We did this for a 10-day period between 10 June and 19 June. We have found that, on average, 2,365.8 domain names leave Shopify every day on average (We know this number looks shockingly high, but Shopify still manages to replace them with incoming new domains. This just shows what a churn factory Shopify is) and looked at how many of these were less than 1 year old and it came in as follows:

Date	Total domains lost	domains less than 1 yr. old lost
2018-06-10	1991	863
2018-06-11	2379	1111
2018-06-12	2351	1081
2018-06-13	3213	1552
2018-06-14	2387	985
2018-06-15	2304	931
2018-06-16	2144	930
2018-06-17	2391	1009
2018-06-18	2275	1000
2018-06-19	2223	898
Average per day	2365.8	1036

Click to enlarge

Source: Our analysis of the .com zone file

If we multiply the average number of domains 1 year and younger leaving Shopify every day by the 365 days, we get 378,140 domains. Now, we know that Shopify has approximately 405,000 domains hosted that are 1 year or younger. If the base of 405,000 domains would be the same over time, this would mean that 93.4% of domains churn off before reaching their first anniversary. However, it is fair to note that this method is vulnerable to changes in the base number. If we would rely on this method to calculate new customer Shopify lifetime, we would come to 1/0.934 = 1.07 years, which is 12.85 months.

We think the truth about the churn rate of new customers is somewhere between both methods so between 67.8% and 93.4%, but closer to the higher bound. Let’s assume it is 85%. That would mean the average customer lifetime is 1/0.85, which is 1.17 years, which is a little bit over 14 months. Enterprise grade SaaS companies boast 10-year customer lifetimes.

Looking at the number of domains Shopify adds and loses in .com, we obviously also think the total number of Shopify clients is plateauing (in the 10-day period we observed that the number of lost and added domains was basically the same number) and that’s probably why Shopify opted to start disclosing the total number of active merchants only once per year (at year end). The number of clients growing is now very dependent on Shopify’s international expansion. If you would look at the various country code domains of countries Shopify is focusing on, you would see the number of added domains outnumbering the domains lost, we believe we don’t have access to the various country code domain zone files, so we can’t provide exact numbers for that though. But it’s obvious that .com is very dominant for Shopify.

What is the really bad news for Shopify’s business model and its investors is the fact that domains that are between 1 year and 2 years are also showing very high churn (i.e. a sizeable chunk leaves Shopify every day). Using the expiration method above, only 51% survive their second-year anniversary. Taking into account that quite a few domains leave before their expiration, this percentage is still higher. Data for domains between 2 years and 3 years old show just relatively marginal improvements in churn than their 1 to 2 year old cousins.

Do you seriously still want to pay 16x forecasted 2018 Revenues for Shopify?

You may argue that the churn in customers is not necessarily so important if Shopify’s revenue doesn’t churn. Shopify claims that revenue from every cohort of customers they acquire is stable over time, which they illustrate by a metric called Monthly Billings Retention Rate (disclosed once per year) and have always reported this to be over 100%. As we already said, we find this metric a very flawed indicator of the quality of cohorts for one simple reason: It mixes together high value subscription solutions revenue (gross margin 76.9% in Q1 2018) with low value merchant solutions revenue (gross margin 41.1% in Q1 2018). We believe this metric is marketed by Shopify management because it smartly hides the fact that the gross margin contribution from these cohorts declines over time. By how much? It would be nice if management starts disclosing a new metric called Monthly Gross Margin Retention Rate, so we would know.

Before Shopify management take us up on this, we have to go for our own estimate. We bring in our customer churn model that shows that 85% of Shopify clients are dead in 1 year or less. So, in an example, let’s assume Shopify signs up 100 new paying customers on Day 1 that pay Shopify $52 a month average subscription (this is roughly the average price Shopify customers pay. We got this number by taking Q4 MRR and dividing by active number of clients). At this point, these are completely new stores that are not doing any revenue so they are not doing any merchant solutions revenue for Shopify (Merchant solution revenue is predominantly revenue generated through payments). At this point, all the revenue Shopify is getting from this cohort of 100 stores is 100% Subscription revenue with a 76.9% Gross Margin. In one year's time we know that Shopify is left with only 15 stores (15% survival rate based on 85% churn rate from our churn model above). Now, the big question is what the subscription revenue from these stores is. Let’s assume 3 different scenarios. One is ultra conservative in which the subscription fee is the same at $52/month per store. In this case, Shopify didn’t manage to move any of these stores to higher paying plans. The second one is optimistic and assumes that Shopify has managed to extract 65% more subscription revenue from these stores so they are paying $85.8/month/average. And then, we also include a super optimistic scenario in which Shopify extracts 130% more so the average subscription is $119.6/month. We also know that Shopify claims a 100% or slightly higher Monthly Billings Retention Rate, which means the remaining revenue has to be coming in as Merchant solutions revenue. So, let’s assume that in 12 months’ time, the revenue from this cohort of stores is 110% of what it was on Day 1. And now look at this table. It says a lot:

	T=0	T=12 conservative	T=12 optimistic	T=12 super optimistic
Number of stores	100	15	15	15
Fee per store	52	52	85,8	119,6
Subscription revenue from cohort	5200	780	1287	1794
Gross margin from subs revenue (76.9%)	3998.80	599.82	989.70	1379.58
Merchant Revenue from cohort	0	4940	4433	3926
Gross margin from merch revenue (41.1%)	0	2030.34	1821.96	1613.58
Total revenue from cohort	5200	5720	5720	5720
Gross margin from cohort	3998.80	2630.80	2811.67	2993.17
Difference in gross margin after 12 months		1368.64	1187.13	1005.63
%		-34.20%	-29.60%	-25.10%

Click to enlarge

Source: Our own model

If we assume that the optimistic scenario with a monthly $85.8 is the most likely, then this shows that the same cohort of stores is contributing 29.6% less Gross Margin than 12 months ago! In our customer churn model above, we have shown how many stores are expected to die over the following 12 and 24 months so you can do the modelling yourself of what gross margin this cohort brings over the next couple of years.

In the world of real SaaS companies (We hope that by now you are with us that Shopify isn’t one them), there is this one super important metric called dollar-based net retention rate which is used to measure revenue churn. We’ll borrow AppDynamics’s description:

“To calculate our dollar-based net retention rate for a particular trailing 12-month period, we first establish the recurring contract value for the previous trailing 12-month period. This effectively represents recurring dollars that we should expect in the current trailing 12-month period from the cohort of customers from the previous trailing 12-month period without any expansion or contraction. We subsequently measure the recurring contract value in the current trailing 12-month period from the cohort of customers from the previous trailing 12-month period. Dollar-based net retention rate is then calculated by dividing the aggregate recurring contract value in the current trailing 12-month period by the previous trailing 12-month period.”

In real SaaS-land, only subscription revenue is counted towards the Dollar based net retention rate. The most successful companies have Dollar based net retention rates of 130% across a 12-month period. This means that the subscription revenue of the cohort actually increases by 30% over a 12-month period – these companies are skilled at moving their clients to higher paying plans, up-selling other SaaS services etc. If we would look at Shopify’s pure subscription revenue, this metric basically turns out terrible for Shopify. If we would take into account new Shopify paying customers and the total subscription revenue of the cohort after 12 months, we would come to these following dollar revenue retention rates using the pessimistic, optimistic and super optimistic scenarios we provided:

	Dollar based retention rate
Shopify conservative scenario $52/month	15.00%
Shopify optimistic $85.8/month	24.75%
Shopify super optimistic $119.6/month	34.50%

Click to enlarge

Source: Our modelling

I don’t think it is completely fair to Shopify to completely ignore the merchant solutions revenue because it does behave relatively predictably (however less predictably than the street thinks as we will show further in this report) but is definitely less stable than the subscription revenue so it should have a haircut to its value. Because of the different margins for Subscription and Merchant solutions revenue, we will use the respective gross margins to calculate this Dollar-based retention rate, and we will assign the Merchant solutions gross margin various weightings – we use 30%, 50% and 70% (meaning that Shopify Merchant solutions gross margin is only 30, 50 and 70% as valuable compared to subscription gross margin). For Shopify’s sake, we even include 100%. These are the Dollar based expansion rates we get under the various scenarios and weightings:

Shopify conservative scenario $52/month + 30% merch GM	30.23%
Shopify conservative scenario $52/month + 50% merch GM	40.39%
Shopify conservative scenario $52/month + 70% merch GM	50.54%
Shopify conservative scenario $52/month + 100% merch GM	65.77%
Shopify optimistic $85.8/month + 30% merch GM	38.42%
Shopify optimistic $85.8/month + 50% merch GM	47.53%
Shopify optimistic $85.8/month + 70% merch GM	56.64%
Shopify optimistic $85.8/month + 100% merch GM	70.31%
Shopify super optimistic $119.6/month + 30% merch GM	46.61%
Shopify super optimistic $119.6/month + 50% merch GM	54.68%
Shopify super optimistic $119.6/month + 70% merch GM	62.75%
Shopify super optimistic $119.6/month + 100% merch GM	74.85%

Click to enlarge

Source: Our modelling

For the sake of comparison with all the other real SaaS names, let’s settle for the optimistic scenario (Shopify manages to extract 65% more subscription revenue from the stores that survived 1 year) and attributing 50% value to the Merchant solutions gross margin, we get to a Dollar-based retention rate of 47.53%. I am calling this rate Shopify Adjusted in the table below. That is indeed a very very low number. Pacific Crest have put together a fantastic table with the respective Dollar retention rates of various publicly traded SaaS companies at the time of their IPOs.

We’ve decided to select a couple of them along with their Dollar based retention rates and we add their 2018 EV/REV multiples using June 21. stock prices.

Company	Annual retention rate	2018 EV/Rev multiple	Stock price June 21.
Chanel Advisor (ECOM)	>100%	2.66	14.82
Cloudera (CLDR)	142%	4.89	15.28
Everbridge (EVBG)	112%	9.63	48.72
Hubspot (HUBS)	90%	10.45	139.4
Mimecast (MIME)	108%	9.91	45.19
Mindbody (MB)	109%	7.33	42.75
Okta (OKTA)	123%	21.06	54.03
Paycom (PAYC)	91%	11.75	109.43
ServiceNow (NOW)	96%	12.05	181.45
Ultimate Software (ULTI)	96%	7.41	274.87
Zendesk (ZEN)	120%	9.82	57.7
Shopify Adjusted	47.53%	15.83	166.76

Click to enlarge

Source: BlueCrest and Bloomberg terminal

As you can see, Shopify’s Adjusted Annual Dollar Retention rate is a paltry 47.53%. On the Pacific Crest list, there isn’t a single SaaS name with a lower dollar retention rate than 80%. In this perspective, it seems absolutely crazy that Shopify is trading at a forecast 2018 EV/REV multiple of 15.83. Since Shopify has lower gross margins than the other SaaS players, the difference would be even more glaring if we would look at 2018 EV/GM multiples. True, Shopify is growing its revenue 68% y-o-y, but this revenue can be classed as absolute garbage compared to the other SaaS players. Okta is growing over 60% y-o-y, but it has a huge dollar retention rate of 123% (and also a better gross margin) and hence trades at a 21.06 multiple. If we would compare Okta and Shopify, Okta could be called a complete bargain. Never in our lives would have we thought that we would be calling a company trading at 21x revenue a bargain. Only next to Shopify!

As we have mentioned before, we truly believe that Shopify cannot be compared to other players in the SaaS sector and hence cannot deserve a SaaS multiple. Shopify has made a complete mocking of the term Monthly Recurring Revenue (MRR). Actually, how can it even be called monthly recurring revenue if it’s clearly not recurring but churning like crazy.

Did Square just do a deal of the century by acquiring Weebly for $365 million?

Or did Square (SQ) just pay fair value and the market is valuing Shopify at insane levels if we use Weebly as a benchmark? On April 26. Jack Dorsey’s Square announced that it is acquiring Weebly for $365 million in a cash and stock deal. More details about the acquisition such as revenue or EBITDA multiples were scarce. Why was the news scarce? Our belief is that Square was basically arbitraging its own crazy 2018 EV/REV multiple of 18x. But the press release did mention Weebly has 625,000 paying subscribers. That got us thinking and digging.

Because 625,000 paying subscribers is quite a lot. Shopify itself claims to have something like 600,000+ clients. Obviously, Weebly is much more of a website builder than a store builder but it does have its own e-commerce offering as well. So, we went searching for a benchmark to Weebly to figure out how much revenue could it possibly be generating per paying subscriber. Fortunately, there is WIX (WIX) which is publicly traded. And it’s exactly and predominantly a website builder that also has an e-commerce solution just like Weebly. Perfect match. In WIX’s Q1 2018 earnings release, we found that WIX had 3,500,000 paying subscribers as of March 31, 2018. They generated $137.77 million of revenue in the quarter. That would be $13.12 per user per month of revenue.

If Weebly would generate the same amount per user per month, its 625,000 subscribers would be doing $8.2 million in revenue per month. That comes to $98.4 million annualized. So, we think Jack actually bought Weebly for something like 3.7x annual revenue, and Weebly was also most likely profitable. Not kidding. He just didn’t want to talk about it. So, people would keep on paying that crazy 2018 EV/REV multiple of 18x for Square itself.

We understand people look at website builders as a little old school and that website builder valuations cannot be compared to sexy e-commerce SaaS providers like Shopify. But website builder revenues are really recurring license revenue whereas the bigger part of Shopify revenues are Merchant solutions revenues that we argued are less valuable than the subscription sort, right?

Since we are a curious bunch we kept on asking ourselves that there must be some kind of catch that would explain why Jack could pick up Weebly so cheap. Well, it’s obvious Shopify is growing its revenue faster than Weebly but maybe the relative cheapness of Weebly could be explained by its subpar churn rates and Dollar-based retention rates. Definitely worse than Shopify’s, right? How could we find out?

You guessed it. We did the same analysis of the .com zone file for Weebly just like we did for Shopify. And we were really surprised by the results.

On June 25. There were 738,386 .com domains hosted by Weebly (very similar number to Shopify actually) according to our analysis of the .com zone file. This was their age split:

1 year or younger	159584
1-2 years old	122019
2-3 years old	89969
3-4 years old	87442
4-5 years old	65798
More than 5 years old	213574
Total	738386

Click to enlarge

Source: Our analysis of .com zone file

Seems like a much healthier split than Shopify. Shopify had 57% of its total domains 1 year or younger. In the case of Weebly, it’s only 21.61%.

We also took a look how its domain base is churning. This is the number of domains we observed churning off over a 10-day period:

2018-06-10	272
2018-06-11	900
2018-06-12	354
2018-06-13	756
2018-06-14	433
2018-06-15	579
2018-06-16	381
2018-06-17	640
2018-06-18	203
2018-06-19	922
Average	544

Click to enlarge

Source: Our analysis of the .com zone file

The differences in churn were huge. Whereas Weebly was churning off 544 domains average per day from a base of 738,386, Shopify was churning off 2365.8 domains off a base of 765,000 domains. Why is Shopify churning off 4.3x more domains off a portfolio of a very similar size? We think the reason is that people starting stores on Shopify believe they can become rich, as the army of affiliate marketers, or "partners" (that's how Shopify calls them), tell them is absolutely possible and just a few clicks away. Whereas when they launch sites or stores on Weebly, they actually start them because they want a nice functioning website or store that they want to keep.

The Shopify new born merchants find out the hard way, spending money on Shopify subscriptions, and most of them never even make a sale and leave shortly. That’s why Shopify cohorts churn so bad and Weebly’s don’t, we believe.

They say a picture is worth a thousand words. Here are just some examples of how Shopify’s partners are acquiring customers for Shopify on YouTube:

Source: YouTube screenshots of videos promoting Shopify as a source of massive income

No operating leverage

Shopify simply signed a deal with the devil, and it's going to unravel soon. The devil always wants payback in the end. Shopify became completely addicted to its continuously rising valuation which it was getting for showing hypergrowth. It delivered and is still delivering it by acquiring huge amounts of low quality customers that churn off very quickly. But the problem with this deal with the devil is that it has to keep on increasing the number of new customers it signs up to keep growing because the huge churn of its customer base is catching up with them. And as every online marketer in the world knows, there are diminishing returns on your marketing spend. Each marginal customer you acquire becomes more expensive. And because Shopify has to keep its hyper growth to support its stock price it is trapped with spending a ton of money on sales & marketing. It can’t cut the S&M spend because it’s growth would start grinding to halt and its valuation with it. Actually, as a percentage of revenue, S&M spend is even slightly edging higher year over year as you can see from this table from Shopify’s latest quarterly results:

Source: Shopify Q1 2018 earnings release

Looking at this table, it looks like Shopify overall has very little operating leverage. The roughly 2 percentage point improvement in its loss from operations y-o-y can be explained by the improvement of gross margin in the merchant solutions segment.

Shopify Capital

How was this improvement in Merchant solutions achieved? We don’t really think it came from better terms with card companies, payment providers etc. We think it’s coming from Shopify Capital. Shopify Capital provides Shopify customers with merchant cash advances, which is basically a form of lending, but with a different name (and less regulation). Shopify boasted in its Q1 2018 results that it tripled the amount of cash advances y-o-y to $60.4 million from $18.9 in Q1 2017. The total amount of outstanding cash advances as of March 31 was $63.5 million.

Now the question is what kind of Annual Percentage Rate (APR) Shopify is getting on its total outstanding cash advances. Well we think it could be around 30%, possibly even higher. That’s 7.5% a quarter. If we take into account Shopify’s $63.5 million outstanding balance, that would come to $4.76 million of margin per quarter. We think this explains most of the significant improvement of Shopify Merchant solutions gross margin y-o-y from 34.4% to 41%. Shopify did $114.14 million of merchant solutions revenue in Q1 2018.

Ethics of lending at high rates aside, it seems like a good business financially. But we know nothing about the credit quality and to what type of customer they are lending. We don’t think the credit quality is very good though. Our hypothesis is that these merchant advances are taken up mostly by the lowest quality Shopify customers – the drop shippers. Genuine businesses don’t borrow for a 30% APR. Drop shippers do because they don’t have a business history and credit history, and that’s the only way they can fund their growth. This whole system works until something changes. Then you’re just left with a toxic non-performing loan portfolio and you’re nursing losses.

Facebook almighty

Well we think this "something" will be Facebook cleaning up its advertising feed. These drop shippers are in most cases totally reliant on acquiring traffic through Facebook and Instagram. If this traffic source is cut off, so is their revenue. This affects Shopify significantly through reducing its payment revenue from these clients and also Shopify’s capability to collect its outstanding cash advances because they are taken from the customers' revenue. Typically, 10% of daily revenue. If a drop shippers revenue drops 90% overnight you know what that means to Shopify’s ability to collect these cash advances.

We think this cleanup of the advertising feed is just around the corner. On June 12, The WSJ reported that Facebook is offering a new feature that will let users rate shopping that stems from Facebook ads and Facebook will start banning low quality stores. Facebook is obviously acting in its long-term interest here because otherwise it risks that people will click less on the ads in its newsfeed, which is the bread and butter of its business. And we believe blue-chip advertisers also can’t be too happy to appear next to products promoted in the news feed like this:

Click to enlarge

Source: BlueCrate.com

The soap dispenser in the shape of a nose is our team favorite.

We think these changes are going to hit Shopify pretty hard. How hard? How possibly could we find out? Well we ran all 765,000 the domains hosted by Shopify through SimilarWeb (expensive!), a service that tracks visitor numbers. We compiled a list of the 250 most visited stores on Shopify within the .com universe (ranked by the amount of monthly desktop visits). We then went through every single one of them, looked at their traffic patterns, categorized them, retrieved total visits for May 2018 (desktop + mobile) and also looked at their Trustpilot and Facebook reviews. A lot of work indeed.

This is what we found out…

These top250 stores recorded 71.79 million desktop visits in May 2018 according to SimilarWeb. After manually checking these 250 domains on SimilarWeb, we found out that their total visits (desktop + mobile) was 162 million visits (162/71.9 = 2.25).

Now how big is this sample in relation to all the stores hosted by Shopify in the .com universe? The total desktop traffic for all 765,000 stores according to SimilarWeb was 371 million and this also includes 40.4 million from MyShopify.com (the domain on which a lot of very small merchants without their own domain run on). If we multiply by 2.25, we get to 835 million visits on desktop and mobile combined for all domains hosted by Shopify in the .com universe.

So, these top250 stores we selected represent 19.4% of all Shopify traffic on .com domains.

So how much of Shopify traffic could be affected by the Facebook advertising feed clean up? First, we scanned these for stores with suspicious traffic patterns. We classed stores that saw their traffic more than triple in 6 months as suspicious. Most of the stores actually went from zero traffic to hundreds of thousands or even millions. Most of this new traffic pretty much fall into Shopify’s Q2. We found 29 stores with these traffic patterns representing 19 million total visits in May. That’s 11.7% of all the visits of the top250 stores. We find these to be the slice that is almost certain to be chased off by Facebook because it’s basically all drop shippers selling stuff from China or other low-quality impulse buy types of products shipped locally. We include the traffic patterns of all of these 29 stores. Feel free to look at what these stores sell so you get a good idea. And also look at their reviews on Trustpilot or Facebook and make sure to ignore the fake ones. Well, it won’t be that easy on Facebook because, not surprisingly, all these stores have pretty much banned Facebook reviews on their Facebook pages.

Click to enlarge

Source: SimilarWeb screenshots of individual Shopify stores

2sbest.com alone goes from absolutely 0 visits in February to almost 3 million in May. As of June 22, it has 54 ratings on Trustpilot with a rating of 1*. People are reporting that the stuff they ordered and paid for never arrived, that they got completely differently looking dresses etc.

So that’s 11.7% of traffic from the suspicious looking traffic category. That’s just the beginning.

Let’s continue with FashionNova.com, which according to SimilarWeb data is Shopify’s most visited store with 14 million total visits in May. Now FashionNova.com isn’t probably shipping from China like the drop shippers but it gets terrible ratings on Trustpilot – please just ignore the ones that are obviously fake (even Trustpilot tells you that they have detected a large number of fake reviews). So, we think FashionNova is quite at risk and FashionNova itself represents 8.6% of total visits in the top250 category!

If we stay in the category of the biggest of the big (top25) with tragic ratings, you have colourpop.com with 4.1 million visits, 2sbest.com with 2.87 million, mvmtwatches.com with 2.15 million, hawkersco.com with 2 million, kyliecosmetics.com with 1.98 million, morphebrushes.com with 1.86 million. This group represents another 14.96 million visits, which is another 9.2% of the top250’s total visits!

And we just looked at the 25 biggest out of the top250. It’s true that these stores are not getting 100% of their traffic from Facebook or Instagram but we think around 80% would be a fair estimate. If Facebook cuts them off, they still have their mailing lists and a few other sources of traffic.

Taking a quicker look further down the list we come to the conclusion that roughly 50% of Shopify store traffic could be vulnerable to Facebook cleaning up its advertising feed! It’s obviously a question how aggressively Facebook will deal with this issue and that will determine the level of discomfort of Shopify.

50% of traffic doesn’t have to necessarily mean 50% of GMV as well because some of the culled traffic could be less valuable traffic than the un-culled etc. But still we think a lot of GMV is at risk here and it will hurt Shopify a lot. Not only in payment revenue. But paradoxically also in subscription revenue. Because Shopify charges Shopify Plus clients 0.25% of each transaction on top of the $2,000 monthly fee for Shopify Plus. The majority of these Shopify top250 stores are Shopify Plus clients we believe. And then don’t forget about that merchant cash advance portfolio turning sour.

Doesn’t it all remind you of a company story we all saw play out in the past? Somebody that grew up with Facebook only to be mercilessly executed by Facebook. Yes, we believe Shopify is very similar to Zynga (ZNGA) in this sense. Good luck.

How is the average store doing?

As we mentioned above, we ran the entire list of domains hosted by Shopify through SimilarRank to find the top250 largest Shopify stores. This exercise also enabled us to take a look at how the average store is doing. When Tobias Lutke, CEO of Shopify, was responding to some of the accusations Citron Research made against Shopify in the Q3 2017 earnings conference call this is what he said about Shopify’s client base: "So even though some entrepreneurs have failed and stop paying $29 per month this floor, hundreds of thousands are thriving and we thrive alongside with them."

So, we went looking for these hundreds of thousands of merchants that are thriving on the Shopify platform.

We ran all 765,000 .com domains hosted by Shopify through SimilarRank.

Out of the 765,000 domains we ran through SimilarWeb, only 123,400 showed any traffic in May 2018. The remainder showed zero. The total amount of desktop visits all 765,000 Shopify stores recorded in May by SimilarWeb was 371 million, but this includes 40.3 million from MyShopify.com (the domain on which a lot of very small merchants without their own domain run on) so we take MyShopify out and we are left with 330.7 million desktop visits, which means 744 million total visits on desktop and mobile combined.

So, we find that the average store has 799 visitors per month and the median store has 0. The averages are supported by some very large stores, just the top250 are doing about 20% of total traffic.

So, we took a look at the visitor numbers of the 123,000 that showed traffic and split them up into categories based on the amounts of traffic they received in May and found this:

Shops with traffic
More than 100,000 visits	1021
More than 10,000 visits	11679
More than 5,000 visits	20499
More than 1,000 visits	57128
Less than 1,000 visits	66367

Click to enlarge

Source: Our zone file analysis data combined with SimilarRank data

Now we understand that most people are not versed in the value of a visit to an e-commerce site, neither are we, so we took the amount of GMV Shopify reported in Q1, which was $8,000 million ($8 billion). We divided that by 3, to get to an average monthly GMV for Q1, which is $2,666 million. Well and then we just divided this monthly GMV number by the total amount of visits Shopify stores received (we include MyShopify.com visits this time) in May – 835 million. We got to $3.1 per reported SimilarWeb visit.

So, by this logic 1,000 visits would mean $3,100 in GMV for a store. GMV also includes shipping and taxes, so to get to store revenue from sale of products, you would need to deduct this.

If we stick to GMV, this logic would mean that only 57,128 stores on .com domains do over $3,100 of GMV per month and 20,499 stores do over $15,500 in GMV/month.

We think that $15,500 in GMV per month is a fair level to use to call a merchant as thriving. It is fair to say that the stores receiving less traffic than the stores receiving hundreds of thousands of visitors could have a higher GMV per SimilarWeb visit than $3.1 because the most visited stores focus on cheaper items, but even if it would be double, it doesn’t really change that much.

In any case, it seems that "hundreds of thousands" of Shopify customers are seriously not thriving, Tobias. We think you should start disclosing the number of customers doing over $10,000 in GMV per month in your regular quarterly filings by the way.

Taking into account that 83.9% of 765,000 domains hosted by Shopify didn’t show any visits in May SimilarWeb) we can obviously state that the customer base of Shopify can’t certainly be classed as healthy. And it explains why the churn numbers of Shopify customers are so astronomic.

Shopify Plus

Shopify Plus is spun by Shopify as a premium service for large merchants and brands that comes with more possibilities to customize storefronts, dedicated account management etc.

As of December 31. Shopify Plus had 3,600 merchants and generated MRR of $6.3 million according to its Q4 2018 results and accompanying conference call.

In the Q1 2018 earnings release, Shopify said Monthly recurring revenue grew 57% and ended the quarter at $32.5 million. Shopify Plus’s share of this ticked up in the quarter to 22% of MRR, compared to 17% in Q1 of last year.

Shopify here is spinning the story that Shopify Plus is growing ultra-fast, it’s MRR at a much higher rate that the SMB business. Since Shopify Plus did $3.5 million in MRR in Q1 2017, we can calculate that Shopify Plus MRR grew 100% y-o-y.

In every earnings release, Shopify talks about the Shopify merchants it adds. For example, in Q1 2018 they mentioned The UGG Company out of Australia, LeSportsac, Monster Electronics, HarperCollins UK, Vega, Colgate-Palmolive (CL) and some great new shops from the likes of Nestle (OTCPK:NSRGF) and PepsiCo (PEP).

Shopify is doing all it can to spin Shopify Plus as a service for blue-chip and enterprise grade clients. Because it knows that this kind of revenue is valued by Wall Street differently than SMB revenue. Looking at the way Shopify presents information to investors about Shopify Plus, we feel it really wants you to sort of think that this fast MRR growth is driven by adding new Shopify Plus customers.

Well, we think that blue-chip clients are a very small minority of Shopify Plus MRR and the MRR growth has more to do with extracting more money from existing Plus customers (Please note this is the wilder hypothesis we make in this report, the others we are much more certain about). We have quite an interesting hypothesis when it comes to Plus MRR size and growth. In conference calls, Shopify has talked about pricing for Plus. We have learnt the Plus subscription fee is $2,000 per month + 0.25% of revenue (with a cap of $40,000). We think that it is quite possible that Shopify is sort of bending the definition of MRR and counting this variable license revenue towards this Plus MRR. We think that the definition of MRR Shopify uses would enable this (not too many people look into the footnotes):

Monthly Recurring Revenue, or MRR, is calculated by multiplying the number of merchants by the average monthly subscription plan fee in effect on the last day of that period and is used by management as a directional indicator of subscription solutions revenue going forward assuming merchants maintain their subscription plan the following month.

Subscription plan fee could mean the amount a Shopify Plus client paid in total – fixed $2,000 + 0.25% of revenue, not just the $2,000 base.

Why is this important? Because if true, Shopify could be growing its Plus MRR mainly through growth of some of the clients we talked about in the section of this report about Shopify’s top250 clients – because it collects 0.25% of their revenues as license revenues.

0.25% of revenue can actually be a lot of money for Shopify. We are almost certain that FashionNova.com, a Plus client, is paying Shopify the max cap of $40,000 per month and a couple of other customers are at that level or very close. And possibly up to maybe 200 customers could be paying around $10,000 per month.

So, clients where the 0.25% fee of revenues in effect could be doing a very significant chunk of the $7 million of Plus MRR Shopify reported in Q1 2018. If our hypothesis that this revenue is counted towards MRR is true obviously. If not, this revenue is obviously counted towards the subscription solutions revenue.

From a previous section of this report, we know that a lot of this revenue is tied to Facebook and its willingness to tolerate low quality stores to freely roam in its news feed via its advertising platform. But we know that the point when Facebook starts cleansing its advertiser feed is most probably just around the corner. This will obviously have a strong negative influence on Shopify licensing solutions revenue (by lowering the revenues of stores on which Shopify charges 0.25% of their revenue) and possibly to Plus MRR (if our hypothesis is correct).

At that point Shopify’s investors and Wall Street will learn (the hard way) that Plus revenue was not deserving a higher valuation than its regular SMB revenue.

Since Shopify always mentions some of the blue-chip clients it has signed up in its quarterly reports to illustrate how enterprise grade its Plus revenues are, we finish off this part of the report with mentioning some Plus clients which they don’t mention. Fleshlight.com, a peddler of infamous pleasure hardware, is a top10 most visited Shopify store according to our analysis and pays a lot to Shopify for its Plus license (possibly close to the $40,000 cap per month). SiliconWives.comis doing a service to those seeking a replacement to their flesh and bone wives. And HoneyBirdette.com is quite similar.

Valuation

We’ve went through all the arguments we wanted to put forward, so it’s about time we tell you what a fair valuation for Shopify is.

But before we do that we would like to acknowledge that there also are real and legitimate businesses that run on Shopify and there is quite a lot of them. Not just drop shippers peddling cheap stuff from Alibaba (BABA). Not a bunch of stores which are almost totally reliant on Facebook advertising. But also, a ton of small stores selling handmade products, cool inventions, specialty foods direct from the producer, interesting online courses, a lot of charity shops, you name it. Often with a vibrant fan base and following. These are all merchants which have a lot of value and the revenue from them could be deserving of something closer to real SaaS company revenue multiple. These are the stores Shopify would normally have been focused on acquiring had it not caught the Wall Street bug.

But it did and it signed a deal with the devil. The devil would provide hypergrowth by providing masses of low quality customers with high churn and a constantly growing valuation. Shopify used this constantly rising valuation (up to a 1,000% increase since IPO) to tap the public markets for an unprecedented 3 times since its IPO: in August 2016 (management was happy to sell at $38 here), June 2017 and February 2018 for a total of almost $1.5 billion. Why has management been so keen to raise so much money? We think for two main reasons. Either as some sort of insurance policy if the stock price would collapse (the cash on the balance sheet would be a buffer for the stock to collapse further) or to raise ammunition for acquisitions arbitraging Shopify’s massive valuation multiples (the sort like Square acquiring Weebly), since buyers would probably not be willing to take Shopify stock.

But the devil has now come for payback (we apologize to Shopify for maybe wakening him a tad too early) and we think all these facts mean that the stock is worth much less than the public market thinks.

We think Shopify is worth about $5.1 billion in today’s market, which is about $60 a share.

$1.5 billion is what the cash on the balance sheet is worth. That’s undisputable.

We think the operating business itself is worth 4x 2018 revenue, so approximately $4.1 billion.

Why only 4x revenue?

• The terrible churn metrics of its client base, where 85% of clients are most likely dead within 1 year and churn is still substantial in following years.
• The deteriorating amount of gross margin every cohort of clients brings over time. Real SaaS companies have positive dollar retention rates of high margin license revenue.
• More than 53.3% of Shopify revenue comes from Merchant solutions (with a 41% gross margin in Q1) and only 46.7% comes from the high margin Subscription solutions (gross margin 76.9% in Q1). Real SaaS companies like Salesforce do almost all of their revenue (93.5% in Salesforce’s case in Q1 2018) in Subscription and support services, where Salesforce has a 77.2% gross margin.
• We don’t think much value should be attributed to Shopify’s revenue growth because it is all just a function of massive S&M spend. Shopify spends 61.1% of its gross margin on S&M and it acquires customers with very short customer lifetimes in comparison to real SaaS companies.
• Square most likely acquired Weebly for about 4x revenues. Weebly’s revenue has a higher gross margin. Weebly’s client base shows less churn.
• We think a lot of Shopify revenue is reliant on Facebook tolerating low quality advertisers in its advertising. We think Facebook is going to make a move on these advertisers, which will have a significant impact on Shopify GMV.
• We think Shopify Plus revenues are not as enterprise grade as Shopify would want you to think.
• We think Shopify could be developing a toxic portfolio of merchant cash advances that could be hard to collect. Also, we believe that Shopify’s gross margin improvement in merchant solutions y-o-y is largely due to Shopify Capital.

Questions for Shopify management

We acknowledge that we could be wrong with some or all of our allegations and hypotheses we make in this report. A significant portion of them could be disproved by answering these five questions.

• Please disclose 3 different quarterly cohorts of clients that became new to the platform and disclose the amount of gross margin these cohorts generated for Shopify in every quarter up to Q1 2018.
• What percentage of traffic to all Shopify stores comes from Facebook and Instagram?
• How many of your customers do at least $5,000 of GMV per month average?
• What is the average APR Shopify is getting on its outstanding merchant advances of $60 million? To what extent was the improvement of merchant services gross margin between Q1 2018 and Q1 2017 a result of Shopify Capital?
• Do you count the variable portion of Shopify Plus license revenue towards the Shopify Plus MRR you disclose and have been disclosing in recent quarters? What is the current number of Shopify Plus customers?

Disclosure: I am/we are short "SHOP". I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.