My reaction to Zulily’s IPO filing and Flash Sales explained

This week Zulily, the ecommerce flash-sale site targeting moms with products for their kids and themselves, filed their S-1 to go public.   While I knew the filing was coming and I had an idea what would be inside, my first reaction was still……HOLY SHIT.  Without a doubt, the 179 SEC page filing is a must-read for ecommerce investors and entrepreneurs everywhere.  The reason: on one extreme it highlights how well they have executed on an incredibly tough business in just four short years, going from a zero to $600 million in annual sales (my estimates for 2013).  On the other extreme, they have limited profits and have lost $56 million to get to this point in their life-cycle.  In between, there is a ton of interesting and pertinent information.

The intent of this post is not to recommend whether I would invest in Zulily on a go-forward basis.  Regardless of what happens in the public markets, they must be commended on what they have built.  As someone whom competed against them while I was the CMO of HauteLook,  they have done an amazing job.  It does bother me that they tried to block HauteLook IPs and HauteLook employees from accessing their site for competitive purposes, but that is a blog post for another day!  Also, as a new investor for Upfront Ventures, I wish my early deals could end-up like Zulily did for my friends at Maveron (we are a co-investor in a different deal) who STILL owns 24% of this business prior to going public.  Maveron is not a huge VC fund, but this could be one of the great investments outcomes in the ecommerce era.

With all that said…..a little background.  I can remember when we first heard about Zulily in the fall of 2009.  It was four months into my job at HauteLook.  The head our kids business at HauteLook was talking about industry chatter about some Blue Nile guys trying to create “HauteLook for moms”.  The industry might have actually been saying “Gilt for moms”, but we at HauteLook would never say that because we always thought (and to this day believe) we had a better business than Gilt.  Regardless, then came the calls over the next 18-24 months from Silicon Valley VC friends asking me about Zulily (and OneKingsLane for full disclosure) and what I thought.  I always gave them the same honest answer re: Zulily: “love the target market, feel the kids/moms retail space lacks a dominant brand and think the value proposition will resonate.  However, have concerns what was assumed was low average order values and low product margins (a flash-sale problem), thus driving to lower than desired LTVs, high customer acquisition costs and eventually low net income margins”.

So right……..but also so wrong!

Zulily has done an incredible job of building a huge business in a very tough space, without blowing thru a ‘ton’ of money.  Don’t get me wrong, they have raised $137 million to date ($94 million primary and $43 million secondary) which is a lot, but to their credit they still have tens of millions in the bank.  While other flash-sale sites like Gilt and Fab have raised $221 million and $336 million respectively (thus far), Zulily has used a more balanced strategy in regards to investment growth and cost containment.  In addition, their non-inventory model has resulted in a negative working capital model which has set them up for long-term success.  Regardless, just as in other ecommerce businesses, the core data points that need to be looked at relate to a) Gross Margins, b) Operative Leverage, c) Marketing Spend as a % of Revenue and d) Customer Productivity Overtime.  These core data points are obviously in addition to things like team, brand, market size, competitive set, etc.

Let me dive in:

Gross Margins: Zulily’s gross margins — pretty much dollars left after you pay for the product you are selling and the cost for it to be shipped — are just under 30%.  This means for every dollar collected in sales, $0.70 is gone before any expenses such as fulfillment centers, business operations, merchandising, employee salaries, marketing spend, office space, etc.  Compare this to businesses like Lululemon with 56% gross margins or Nordstrom with 39% gross margins.  These two cases are the difference between someone who sells their own product (Lulu) vs. somewhere who sells someone else products (Nordstrom).  In short, this is not as high as most investors would like, but is the reality of being a seller of other people products and made tougher when selling other peoples product at low price points and at a discount.  Now you can see why businesses like Bonobos, Combat Gent and Nasty Gal all get such attention from investors as their gross margins are closer to Lululemon than Nordstrom (and Nordstrom is an AMAZING business) and also is a reason why from an early-stage investment perspective, I am a pretty quick pass on any ecommerce business with less than 40% gross margins.

Operating Leverage: Flash Sales are hard.  The business is hard to scale and expensive to operate. Remember when there were several hundred similar businesses? Not only are you playing with less than desirable gross margins, you are effectively merchandising a new shopping experience everyday.  Zulily highlights that they are launching 250,000 product styles per quarter (2,777 per day).  That is A LOT.  Think about the production and operational work and processes that needs to get done every day to get all of these items live to the site.  From the coordination and storage of product samples, to scheduling of photo studio and photographers, to photo creation, selection and editing, to product measurement and product detail description writing, to the other ten or 20 steps in the sales creation process.  Also, that doesn’t include all the pre-work around setting-up the OTS (open to sell), vendor set-up and accounting, post-sale coordination, etc.  While someone like a ShopBop.com or Nordstrom.com might have 20,000 to 50,000 unique SKUs for sale at any one given time, such items on those sites stay-up for long periods of time (usually 3 months or longer) – maybe 200 to 500 new styles go live each day.

Compare this this large flash sale sites where more than a hundred new products go live to the site each HOUR.  To support the relationships with 10,000 brand partners and 1,000,000 product styles per year, Zulily will spend an incredible $100mm on SGA in 2013 (excluding marketing spend).  I would take an educated guess and say that most of their 811 non-fulfillment related employees work on this part of the business, including 302 in merchandising, 89 in technology and 39 in photo studios.  The Zulily reviews on Glassdoor and some of the notes in the S-1 (see page 58) would also suggest these numbers are correct as well.  Assuming $70k/year in fully loaded expenses for 500 people, one can assume they spend $35 million per year just to get those products to the site each year ($35/item). For full disclosure, Zulily reports to have 886 full-time employees, of which 75 are in in fulfillment, but not including the temp labor they use in the their fulfillment centers which I am sure is in the several hundreds.

In addition and amazingly, Zulily manages 1.1 million square feet of their own leased fulfillment space at facilities in Nevada and Ohio (size of almost 20 football fields). There they can handle over 100,000 items per day in volume per day which if you back of the envelope is about $750 million in volume/year (100,000 items at an ~$20 per item value times 365).  That might actually mean they need to soon add even more sq. footage for fulfillment, but this is not meant to be a takedown of their future cap-ex needs.  With that said, order-to-ship times for customers is 10.6 days down from 12 days just 18-months ago.  This is actually very good for a “pick, pack and ship” model.  The best-in-class flash sale providers like Gilt, HauteLook and Rue La La are all in the 10-11 day range for such deliveries.  Again, a long-winded way of saying that this is not a cheap business to execute on and not easy.

Marketing Spend as % of Sales:  As a marketing guy this was very interesting.  Zulily has been very aggressive on the marketing spend side of the business from the start. Out of the gate (2010) they were spending 27.8% of revenues on marketing and today spending just under 10% of revenues on marketing….wish I was at those Board of Directors meetings in the early days! In real dollars, they will have spent almost $120 million over the past four years to build-up their customer base, including what will be around $55 million in 2013 alone.  This may surprise you, but this level of spend sits between where Fab was spending (no comment) which was reportedly 35% of revenues and the other leaders in the flash-sale space whom are closer to 5-8% of sales.  The reality is while Zulily has shown incredible YoY sales growth, most of those gains are being driven by spending more money each year.  With that said, there is little doubt they are best-in-class marketers and you may even see their TV ads on the air today.  However, it is fair to say that Bill Gurley is probably not be a big fan of the flash sale category.

Customer Productivity: Zulily reported that average revenue per customer (defined as a buyer) per year is $210.  In addition, AOV (average order value) is $53 (which includes shipping cost charged to the customer).  Therefore, an average customers orders from Zulily four times per year.  Without divulging too much, the number of orders per buyer per year is similar to the other major flash sale sites.  What makes Zulily stand-out relative to others is that it appears (at least in the Q1 2011 customer cohort) that average spend per customer per year does not decline on an annual basis.  Like I said, this is a unique positive characteristic and one that is rarely seen in any ecommerce business, let alone a flash sale business.  Hopefully, you have read Jeremy Liew’s blog posts on LTVs, but an increasing annual spend curve does wonder to the long-term business value creation.  In the Zulily case, they have determined contribution margin per order to be 18% (meaning $18 of every $100 spent invariably drops to the bottom line).  In this case, in a single year a customer spends $210 with $37.80 “dropping to the bottom line” and they can afford to spend roughly $40 to acquire a buyer that is paid for in one-year.  At HauteLook we used to target a six-month payback period and were much more conservative on determining contribution margin.  Not saying what is right or wrong, but it there is no one way that all marketers define how to profitably acquire customer.

Other Zulily Data Tidbits and Conclusion: 

• Highlighted the recent changes to Gmail as a risk factor as it could hurt the ability to message to members.  This will be a future blog post and something that is keeping all Flash Sale CMOs up at night.
• Only charge sales tax in Washington, Nevada and Ohio.  This will change, just not sure if this year or in three years.
• 2.2 million active customers (customers who had purchased at least once in the last year), an increase of 93.1% from 1.2 million active customers during previous 12 months.
• $214 of revenue per active customer per year, a 10.9% increase from the $193 of revenue per active customer generated during the 12 months prior. This is most likely due to improvement of shipping times, decline of shipping prices driving more sales and increase in the number of live sales per day and categories of those sales (ie: not just kids clothes, but also gear, apparel for moms, etc.
• Approximately 42% orders were placed from a mobile device.  This is great, but pretty indicative of all flash sale sites (a great mobile commerce category).  With that said, Zulily has a great mobile app that consistently is a Top 20 Lifestyle App in the Apple Store.
• $2.5 million of net income on $272 million in sales during the first six-months of 2013.
• Class B investors have 10x the voting rights of Class A investors.
• Zulily does not have meaningful inventory.  A huge plus for the balance sheet, but it creates a more difficult consumer value proposition (ie: items can take 10 days to get to buyers).   In addition, their choice of categories is not a big utilizer of returns. The end result is an attractive negative working capital model which has set them up for long-term success.

In conclusion, to go to the next level Zulily must a) continue to acquire new customers, b) attempt to maintain the wallet share and engagement from new customers and c) grow by introducing new categories and new markets (ie: international).  All this while the law of large numbers might say this will be difficult as the ability to continue to grow at that pace and the ability to maintain such loyalty to a single brand is getting more and more difficult.  Regardless, Zulily has emerged in a short-time to be a generational defining ecommerce business.  With well over a billion dollars invested by VCs in the flash-sale category over the past five years in the US, they are the ones that got public.  HauteLook and Rue La La were sold large nine-digit numbers and Gilt and One Kings Lane have built great brands and solid businesses, but it is Zulily who may be the company that defined the post Zappos/Diapers.com ecommerce market.

Note:  I know there is a lot of material here and I know a lot of it is inside baseball.  I did my best to highlight the parts of the S-1 I personally thought were interesting.  I know missed things and didn’t go deep enough in some areas.  There will be more posts to come.  Hopefully, you get the idea.