Format for Printing

Format for printing

Request Reprints


By howardroark
February 27, 2006

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

Normally, I'd consider this unnecessary, but I know enough about this situation to realize it's probably worthwhile to start by saying I have no position, long or short, in Overstock. I've been thinking about the company because I'm fascinated by the business (just as I was with Amazon) and because I know some smart people who come out on opposite sides of the coin here. Below is a list of four questions* relating to things I would care most about as an Overstock investor. Most of them relate in one way or another to understanding the fixed and variable costs of the business a little better, given how hard it for GAAP to reflect the actual economics of a rapidly growing online retailer. I've put some comments after each question in the hopes of providing some context and sparking some discussion. Some of the questions may be superficially addressed toward the company, but I am posting here hoping for comments from anyone who has given these issues some thought. It would be nice if this thread (assuming anyone responds) could be a temporary oasis from the naked elephant in the room, and instead focused exclusively on the business.

1. What is the company's best estimates of "maintenance marketing" costs, meaning the rough amount of marketing expenses necessary to keep sales in line with inflation? What information and approach do you use to think about this sort of thing?

As a starting point, it seems clear GAAP expensing of all marketing expenses is punitive to the ongoing profitability of an online retailer growing over 50% a year. How punitive is a harder question. Overstock is not a membership business, like Netflix, where maintenance marketing is a bit easier (but never a science) to estimate: pretty much SAC times Churn rate of mature customers times Customers. A retailer can give you things like repeat customer rates, CPA and frequency, but a customer just isn't the same as a subscriber. Still, it seems pretty obvious that the cost of maintaining (or reacquiring) a customer is likely to be a lot lower than attracting a new customer. This is a pretty enormous issue when it comes to Overstock, as I'm sure everyone here knows, since marketing expenses in 2005 were about 10%, dwarfing the operating loss of ~3% -- while the stock is priced as if 5% operating margins will never happen.

I thought Overstock's comments in the Q4 Conference Call about the planned easing of growth were potentially helpful in thinking about this issue. If I understand correctly, they guided toward a ~2% decline in marketing expense over the next two or three quarters as they slow growth to industry rates of 10-15%, meaning that marketing as a percentage of revenue will fall by 200bp. If we assume 13% revenue growth, then that implies an early 06 run rate of around $908m in sales, with a hypothetical $72m in full year marketing expense, 200bp below 2005 (ignoring for the moment they expect to ramp growth back up in Q4). So you spent $80m in 05 for an extra $310m in sales (including some possible overspending in response to system problems) and you spend $72m in hypothetical 06 for an extra $104 million in sales, but with a much larger starting base of sales to "maintain," at $803m versus only $495m.

One partially back-fitted solution to these equations could start with estimated CPA. They said CPA was $21 for the year, and the average transaction was up to $103 by year end, implying about $1 in marketing need for $5 in customer sales per order times the average number of annual orders per customer (anyone know -- I am using Overstock's customer data presentation and comparing customer orders with unique customers and adding a slight bump). Let's say based on the prior parenthetical that number is 1.5. That would imply $7.50 in new revenue for every dollar in marketing expenses.** Thus, for 2005 it would mean that $41m of the $80m marketing budget went for "growth," leaving $39m (or $12.70 in sales per dollar in marketing ) for maintenance. In hypothetical 2006, it would imply $14 million for growth and $58 million for maintenance ($13.86 in sales per dollar in marketing). Obviously these are grossly oversimplified ways to look at a very dynamic phenomenon -- for example, Amazon uses shipping promotions to drive revenue as much as incremental marketing, nothing is actually this formulaic -- but it's still probably worth thinking about. If it generally costs Overstock one dollar to maintain a $13-$14 in customer sales, then maintenance marketing would be around 7.5%. This is dramatically higher than Amazon's 2.3% spent on marketing (even less if you gross up sales for marketplace bookings), which seems unfair to OSTK, and it doesn't factor in the maintenance benefits of scale (Overstock spent, I believe, around $12 million in offline marketing last year, for example). It is worth mentioning, though, that Amazon only spent 10.7% of marketing on sales in 1999, a year where it grew sales by 168% and had only $1.6b in sales. Moreover, in 2000, Amazon only spent 6.5% of sales, despite sales growth of 68% and $2.76b in total revenue, albeit in a strong period for ecommerce in general.

2. Are comparisons to Amazon incurably muddled by its substantial but ill-disclosed third party business? What kind of estimates and assumptions does Overstock make about this portion of Amazon's business when you use it as a benchmark? Does the prevalence and growth of Amazon Marketplace allow them to create scale benefits and non-marketplace pricing pressure (such as shipping promotions) that will continually threaten to compress Overstock's gross margins over time?

For obvious reasons, Overstock has in the past often made comparisons and benchmarked some of its progress by comparison to elements of Amazon's North American business -- particularly when it comes to the expected benefits from scale. Most recently, in the Q4 CC, you rightly adjust Amazon's gross profit for fulfillment expense, and said that Overstock had essentially caught up in comparable gross profit this year (15% vs. 15.2%). But isn't this calculation using Amazon's worldwide numbers as opposed to merely its North American numbers, which is what seems more relevant and what Overstock in most cases appropriately cites as the right comparison?

TTM gross margins for Amazon North America was 26.9%. Amazon doesn't break out fulfillment by segment. If the 8.8% TTM fulfillment can be apportioned ratably (say greater US efficiency offset by credit card fees and bad debt related to Marketplace), then Amazon's TTM North American gross margins are more like 18.1% -- a pretty substantial difference. Clearly, a big portion of this comes from third party and ancillary service gross margins, which probably range between 80% or 100%. Unfortunately, Amazon doesn't provide nearly enough disclosure on this critical issue (at least relative to what I would want if I were an investor). We do know that over 30% of worldwide units sold are third party units, and that number is almost certainly much higher in North America. I don't know how much higher, but I wouldn't be surprised if 1 of every 2 North American units sold on Amazon were third party, albeit at a lower ASP. It's easy to do the math (and throw in very high gross margin revenue) and guess that this business probably accounts for all or more than all of the disparity between Amazon North America's gross margins and Overstock's gross margins.

You can either look at this as a gross margin disparity, or as a scale disparity (by grossing up third party volumes), but it seems to have important implications for the comparisons, both theoretical and the actual comparisons Overstock says it makes internally. For example, is there a risk that Amazon decides that it can no long reap rewards from its current supposed rash of "investment" in R&D (all lopped into the North American segment) and instead pushes the pedal harder on a more threatening levers such as price or marketing? Does the third party component of Amazon's business makes its apparent strategy to use price (particularly shipping price) to continually drive absolute gross profits more threatening to a company like Overstock without the same third party mix? Or is this all made less relevant in the end due the Overstock's differentiated closeout model (or any other reason), despite the ostensible comparisons? And, finally, does this render Amazon's apparent scale over partially line items like technology and G&A less comparable to Overstock's ambition because Amazon's gross volume is much higher than it seems?

3. What percentage of Ovestock's units and sales in 2004 and 2005 were from "overstocked" goods versus more traditional stuff? What were the average gross margins on overstocked goods compared to non-overstocked goods?

I think most people would agree that Overstock doesn't have as much allure as a smaller Amazon than it does as a closeout category killer. It's pretty obvious from the site that the majority of its merchandise is still overstocked goods, but that there seems to be some amount of non-overstocked inventory, not necessarily limited to the loss-leading (or priced to break even) book category. While any mix shifts would show up in gross margins fairly quickly, the underlying information seems helpful.

4.Given the IT hiccups of late, what gives you confidence that Overstock's hyper-efficient R&D and capital spending (per dollar of sales generated) relative to Amazon's profligate early years is a continuing advantage as opposed to a deferred liability caused by under investment? That may be too general to be a good question. How you believe that Overstock is or will be in the next few quarters at a similar level of technological capability, logistical efficiency and overall structure capacity as Amazon was at a similar stage in their North American business development?

Amazon obviously did a lot of crap that wasn't optimal in hindsight, and I'm just talking about the investment. I'm hardly tech savvy, but I can understand the potential first mover disadvantage when it comes to developing complex systems using depreciating hardware and layered code, and also that building one and a half warehouses for every one you use is something south of six sigma. So I can see pretty plainly why Overstock might have some serious cost advantages. But its cost disparity has been so large, and its systems flare ups prevalent enough to at least wonder how the company can be confident that can operating a growing, near-billion dollar business with only $25-$30m million of capex (still dwarfed by Amazon at all levels of development) without any permanent efficiency disadvantages. Does Overstock believe that Amazon has made a fundamental mistake in opting for largely self-distribution as opposed to the split insource/outsource model used by Overstock? Does Overstock's use of third party fulfillment present greater risks -- in terms of efficiency and productivity -- as Overstock grows toward Amazon's size (or is it the opposite)?

*Why is this post different from all (most) other posts?

**And no, you can't just take 15% * $7.50 per Jeff Matthews and declare this per se uneconomic because it implies $1 in present marketing dollars for $1.125 in future gross profit less incremental customer service expense; that would only be fair if customers completely disappeared after one year.

Become a Complete Fool
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.