Overstock.com, Inc.

Good morning, everyone. My name is Erica and I will be your conference moderator today.

I would like to welcome everyone to Overstock.com's second quarter 2005 financial results conference call. At this time all lines are in a listen-only mode.

Later,we'll announce the opportunity for questions and instructions will be given at that time. A web based slide presentation will be used during this call and is available for viewing over the Internet on the Company web site, www.shareholder.com/overstock.

[OPERATOR INSTRUCTIONS]. This call is being recorded and will be available for replay beginning today at 3:00 p.m.

eastern time through 11:59 eastern time August 10th. The replay can be accessed by dialing 888-203-1112 and entering the code of 4443183.

At this time, I would like to turn the conference over to Mr. David Chidester, Overstock.com's Senior Vice President of Finance.

Mr. Chidester, you may begin.

David Chidester, Overstock.com - SVP, Finance Thank you. Good morning and welcome to Overstock.com's second quarter 2005 conference call.

Participating with me on the call is Dr. Patrick Byrne, Chairman and President.

Before I turn to the financial results, please keep in mind that the following discussion and responses to your questions reflect management's views as of today, August 3, 2005 only. As you listen to the call, I encourage you to have our press release in front of you, since our financial results, detailed commentary, and the President's letter to shareholders are included and will correspond to much of the discussion that follows.

As we share information today to help you better understand our business, it is important to keep in mind that we'll make statements in the course of this conference call that state our intentions, hopes, beliefs, expectations or predictions of the future. These constitute forward-looking statements for the purpose of the Safe Harbor Provisions under the Private Securities Litigation Reform within the meaning of Section 27A of the Securities Exchange Act of 1933, and Section 21E of the Securities Exchange Act of 1934.

These forward-looking statements involve certain risks and uncertainties that could cause Overstock.com's actual results to different materially from those projected in these forward-looking statements. Overstock.com disclaims any intention or obligation to revise any forward-looking statements.

Additional information concerning important factors that could cause actual results to differ in the forward-looking statements is contained from time to time in documents filing with the SEC, included but not limited to its most recent reports on forms 10-K, 10-Q, 8-K, and S-1. With that said, I'd like to review some of the financial highlights from this past quarter.

Please note that all comparisons will be against our results for the comparable period of 2004, unless otherwise stated. Total revenue in the quarter was $151 million, up 72%.

Gross margins were 14.7% in Q2 up from 11.3% gross profit dollars increased 124% to 22 million. Operating expenses grew 129% to 28 million.

The overall increase in operating expenses was driven by a 233% increase in technology costs and 119% increase in sales and marketing. Other G&A increased just over 100%.

We remain comfortable with our overall G&A and technology estimate for full year 2005, of 60 million, plus or minus 5%. Our operating loss was 6.1 million, or 4% of sales, versus 2.4%.

However, we had 3.5 million of net nonoperating income as a result of a $4.1 million gain on the retirement of 33 million of our convertible debt. So our net loss was 2.5 million, compared to a net loss of 2.3 million, both $0.13 per share.

We ended the quarter with 120 million in cash and marketable securities. We had total cash outflow during the quarter of 65 million.

10 million of outflow from operations, 11 million for capital expenditures, and an additional 44 million related to stock and bond buybacks during the quarter. In total for the buyback program, including the 48 million of outflows for the structured buybacks that we had in Q1, we used approximately 64 million to buy back 1.7 million shares and used and additional 28 million to retire 33 million of debt.

One last point to keep in mind regarding cash is that on July 1, 2005, we completed our acquisition of SkiWest for 25 million of cash. That being offset by approximately 8 million of cash returned to us in the first week of July as part of the closing out of the the structured buybacks, leaving us with just over 100 million the beginning the quarter.

Finally the average invoice increased to $97 from $86 last year, and up from $91 in Q1. This is partially due to a change in the sales mix, as BMV fell to 10% of our total gross bookings, versus 12% last year.

That covers my financial overview. I would like to turn the call now over to Patrick.

Dr. Patrick Byrne, Overstock.com - Chairman and President Thanks for that nice summary.

We have got a great slide deck today. I'll keep this preamble short.

One -- first, I'm sure I sound like a broken record quarter after quarter, but I'll reiterate the Overstock mantra: Our internal stretch goal for the Company is 100% growth, break even bottom line, with “break even” defined as, “plus or minus 1% GAAP net income.” Growth at this point is 86%, not too far away from what I want.

I expect our Q3 operating results to be similar to Q2 operating results, but for the full year, we should be able to break even, if not above. Our model that we are developing (and I'll be getting into a little bit how we've gotten quite a bit more sophisticated in our modeling), shows something funny: that we are learning that the more we push out there in Q3 - it seems to be for every extra dollar we push out in Q3 we make back a dollar or maybe more than a dollar in Q4.

I'm often asked how long it is we're going to chant this Overstock mantra. I thought it was time I defined for folks where I'm really shooting for as a place I'd like to be.

I want Overstock.com to be a company with revenues north of $2 billion, growing at 25% a year, with operating income of 4% or 5%. And which would bring us to the $6 billion range -- actually GAAP income of 3-5%, I'm sorry.

That's something I think we can achieve the year after next. We have to keep growing at these kind of paces and get to 900 plus or minus 50 or 70 this year, and 1.5 billion or something next year.

That would position us to get us north of $2 billion the year after next, slow down our growth and have a company that can make a lot of money. Like any -- we call this O Town.

That's where we're trying to settle. O Town must complete certain infrastructure before the infrastructure is needed.

In a letter to owners, I described some stairstep improvements in our technology infrastructure. These have not been cheap.

But to continue the metaphor, I would say during the last nine months really we've gone from being a country hamlet with dirt roads to a modern town with asphalt streets. We're shooting to be like Singapore.

We're making progress in online advertising. We're just at the point where we can really focus on our site design, propeller, CRM initiatives,and improved IT infrastructure that supports the projects mentioned in my letter.

A couple of quick announcements: Katherine Huang is gone. She was our head of IR and I know many of you knew her.

She served wonderfully for us for two years. I believe I actually said this a conference call or two ago.

She's actually married to a doctor who finished his residency here in Utah, Dr. Hadley, and they moved off, he took a position in Michigan and they moved off together.

I think people knew about that six months to a year ago. I'm sure that won't keep there from being any stories about, “The Mystery of Where Kathryn Huang Went.”

But she -- I know a lot of people were given prior notice. She's been replaced by Kevin Moon who is terrific.

Kevin comes from your side of the table. He's an investor.

He worked at a local investment fund called CrossRoads. MBA.

Terrific. We're happy to have him aboard.

Secondly, thought I'd give people advance notice that I've asked my father to become Chairman, and we're going to do that I think at the next board meeting. I'm doing that just because I like -- I think that the Chairman-CEO relationship is an analog to the father-son relationship, and it's healthy corporate governance to split CEO and Chairman.

He was Chairman the first couple few years of our existence. Because of Sarbanes-Oxley he had to get off the board.

So I'm letting you know in advance, I'm going to propose him for that. I'm not sure if nominated he'll run and if wins he'll serve, but we're trying to get him to be chairman.

Let me turn to some slides that will give you a sense of how your business is running. As I recall, you folks are about 15 seconds behind me on this, so I'm going to try to adjust my voice to respect that fact.

On slide 3, we -- we've gone from 87.8 to 150 million. 72% growth there.

124% growth in the gross profits line. 119% in sales and marketing.

G&A, other than technology, 101. Technology and depreciation is what is short here, and that's really because we hit -- we've taken this step over a stair.

Most of the quote, unquote increase in operating loss came in this increase in technology spending. I'm very glad we've done it and I'll be going extensively into what you got for your money.

Slide 4, again, 72% growth on GAAP revenue. Gross margins did slide a bit.

And I described in my letter what's going on. (Now, every time I try to explain what's going on, I know I have some saying out there, you're making excuses.

I'm not. I literally take me letter to the Board and the CFO's letter to the Board and I clean it up a little bit and I give it to you guys.)

So we actually did make a little bit of improvement in the expenses within logistics. The projects that we thought were going to deliver 20 to 30 basis points a quarter are two or three months delayed in their implementation.

I thought they would be done in March, April. It really happened in June, July, so they didn't really affect the quarter much.

They did give us about 20 basis points. We had this 55 basis point head wind as a result of the way we expense inbound freight, which is something we might revisit at the end of the year.

What happens is we sort of had a 55% head wind against us and it all comes back to us in the fourth quarter. I explained in my letter what happened and Dave tells me most companies actually account for it, they would capitalize it.

We may revisit it and Dave may have an opinion on that. David Chidester, Overstock.com - SVP, Finance We'll definitely revisit that.

Most people expense it as you sell the inventory. It makes for a nice flatter movement in your margins.

Dr. Patrick Byrne, Overstock.com - Chairman and President So we expense it not as we sell it, but as we take it in.

And that means that we get some noise both directions in our margins. Anyway, still all that said, we are a little bit slower than I thought we would be in the infrastructure, the logistics improvements that should be adding 20, 30, 40 basis points a quarter.

But they are ticking in. They just took eight to ten weeks longer to implement than I’d expected.

Slide 6, what we like is to see the green line diverge from the red line. We can grow the red line as much as we want as long the green line is much higher.

They have diverged. They have now converged.

And that's really -- that's the stair stepping. When you reach the stair step, it's hard to keep them apart.

Slide 7 -- I'm having trouble. Slide 7 is not coming up for me.

There it goes. Gross profit per transaction.

This is nice. Up 44% versus last year at almost $14.

Slide 8, quarterly net margins, minus 1.7. That's the best it's been in any non-Q4 quarter.

Other than -- well at least in the last few years. I think in Q2 it was a little bit better.

Next slide, slide 9. There's a delay.

Unique visitors were up 62%. 62% growth underlying 72% quarterly growth.

Revenue growth. Unique customers, 61% growth.

New customers, 45%. So we're only having 45% more new customers, but again, the 72% growth.

Year-over-year growth in customers up to almost 7 million customers, that's 81%. That's a pretty good analog with the overall growth in the last four quarters.

Customer orders, 55%. Again, so to get 72% growth, we've got to be getting more per order.

Gender, few slides about our visitor demographics. Our visitor demographics where the Internet is 51% female, we're 56% female.

I said in the past that our customer demographics are 66% female. Our visitor demographics are 56% female.

Draw your own conclusions from that. Household income, slide 15, we have a nice, sort of it seems to me a middle class, upper mid class base visitors.

Education. Slide 17, there's a couple of things I want to correct.

I have got to correct two things that I have given you in previous phone calls. One, is better than it was in the past and one is worse than it was.

The one that's worse is this: At the end of the the fourth quarter in the January conference call I said I thought we had done $10 million of GMV, gross merchandise value in our auctions in the fourth quarter. That was a mistake.

I know exactly where I got the information, I know exactly who I asked and I was given the wrong information. I'm not sure if there was cross talk between us or our reporting since the auctions was so fresh.

Our reporting with auctions wasn't very good. Fortunately on a GAAP basis it's a very immaterial difference.

The truth is in our fourth quarter, it turns out we did 7.1 million of auction gmv, dropped in the first quarter, and has come back in the second quarter. I still have hopes -- the loss has contracted from 1.6 million, to 1.2 million.

I want to see this under a million this quarter. By as quickly as possible not having any quarterly loss.

Not costing us anything to keep auctions going. Then some good news.

Things that are a little bit better than I've reported in the past. In the past we've been reporting this CPA, and cost for acquisition, and we showed you a bar that was broken near the top, and said this is for shopping only and this is it if you add auctions.

We were actually adding the cost of auction marketing to the numerator, but we were adding nothing to the denominator. We were not adding any of the acquisitions in auctions.

So the whole thing got -- the whole thing was kind of meaningless. What we decided to do is break it out between shopping and auctions.

Shopping has seen this 39% increase in the CPA. Something else is going on.

If you look at the red line here, you see that the marketing expense as a percentage of revenue fell. It was down to the 6% to 8%.

Came up to 10% last quarter. Then it fell half a point this quarter, or more.

That's really the key number for us, is the percentage of revenue that we spend on marketing. Now what’s happening that may seem paradoxical is the cost of acquisitions is going up, although that percentage spent on marketing came down a little bit.

The reason is there's less new customers within our traffic -- within the traffic stream that you get from online advertising. That may have to do with things going on on the Internet, it may have to do with where we're advertising.

I know it has to do with that. That's kind of an interesting paradox.

If you would asked us two years ago we would have said you would never have seen that happen, but now we see it happen. And we understand it.

We're going to continue splitting up auctions and shopping from now on. If you go to the auction CPA, you see this is the CPA for auctions.

Actually if we had combined them appropriately and added the costs together and added the acquisitions, we would have gotten a lower number than the 22 or whatever. Not a higher number.

Okay. Repeat revenue analysis.

My first comment is, let's stop the madness. I've given you different ways of calculating it, and I keep on trying to explain it, and each time I try to explain it, I probably made matters worse and people get confused and they write about it as if I mean one thing when I mean another.

What I've decided to do is take a couple of minutes and really carefully explain this. And then I won't have to do it again.

I'll just give you the results from now on. The first way we reported this ran like this.

Calculate the repeat revenue for any quarter as being equal to the revenue in that quarter generated by non-first time invoices divided by the total revenue for the quarter. Well, the result of that is, each customer in at that calculation, each customer from inception is included, including customers that purchase twice in the current quarter.

The problem with that is that if a new customer comes in in the quarter and purchases once and then purchases again, that second purchase counts as a non-first time invoice, and introduces noise into what you're trying to measure, which is really how the customer base that pre-existed, that predated the quarter, how they behaved. So that was version one.

We gave that to you for a year, year and a half. Then I think we did version two for a quarter, which is to say, only count the existing customers before a quarter starts in the calculation.

The calculation would be the same as the version one, except it would exclude all customers from the current quarter, all new customers from the current quarter. Well, that's -- that has problems, too.

One is it's affected by growth. Your growth -- your growth rate introduces noise in the math, as well as the problem is as time goes on and the tail gets longer it creates a built in inflation.

There's a longer tail from which you can draw your customers to buy in the current quarter, so you get a built inflation. However, look at those two ways, version one and version two, version one is the number we reported for quite a while.

It dipped a little bit. It's come back up to 59.3.

Version two is now 50.5. They've both done well.

But as I say, I don't think those are the right ways to report it. Now I tried to report it the right way the last time and I think I caused more confusion than I should have.

Again, I'm walking through this so carefully this time in hope that in the future people can refer to this slide deck if they need an explanation. Okay.

The right way is to take a bracket, a period of 12 months and look at all the active people in that period and say, everybody who was active in that 12 month period, how did they behave over the following 12 months? So for every $100 they spent within the blue bracket period, how much did they spend within the green bracket period?

As time goes on, these two brackets can slide to the right. That's a good measure of your churn, loyalty, et cetera.

On that number, this is what's happened over time. And in turns out that this is the -- now that we're not just trying to figure all this out for ourselves, but we actually have hired a bunch of pros, this turns out to be one of the major ways people measure churn and loyalty.

That shows nice progress in the high 20s up to 54, over the three years that we didn't even know -- we weren't even measuring this. We weren't even -- we weren't even looking at it.

To me, the gain that we've seen so far is mostly a function of, it's a reflection of the fact that we were answering the phones right and shipping on time and satisfying people and they were just getting more and more loyal. We weren't doing anything to manage to this number.

Now I'm going to dive into that in more detail, going through this quickly, breaking it up not by quarters, but by weeks. I've been asked these questions so many times, over the last three years, I'm happy to finally give you information.

Our new customer repeat rate; someone who buys in week one, how did they do in the weeks 2 to 53? We're up to about 40% repeat.

Average repeat orders per customer is a little over two -- a little over one. The average repeat orders per customer who repeats is 2.5.

Again, I'm sort of glossing over all the improvements that have been made in this, and I think that those improvements mostly reflect not that we've become -- that we became over the last two years terrific in CRM. We didn't even know what it stood for.

We became better and better at shipping on time and things like. Average repeat revenue per customer, and you see it went up and then it dropped and came back up.

I think that dip came from when we introduced BMV and starting getting into BMV because those orders are about a third of the size. Those customers turned out to be pretty loyal, but the order quantities are smaller.

Now it's climbing back up to where it was pre-BMV. Average repeat revenue per -- and this is key -- average repeat revenue per -- sorry, my slide's not loading.

Per repeat customer. Of the people who repeat, they come back and buy at this point $250 to $300 after their first purchase within the first year.

So this will present - to people who are familiar with CRM, you see the opportunities here. On the one hand I think these graphs show something good.

We've gotten better and better with loyalty, even without focus on loyalty. Now that we're able to look at this data, it really suggests some powerful techniques we can apply to improve.

Okay. What did we spend $35 million of your money on in IT buildout?

I'm going to be a little bit technical here. I think there are tech investors on the phone that might be interested in what's going on.

One is, the key players in the service space are Dell and IBM. You can think of this as a totem pole.

And Dell, of course, started off in desktops, and home computers and laptops, and got into servers, and has gotten better and better with servers with one, two, 4 CPUs. IBM has done something very smart.

By the way, these guys are both partners of ours, so I don't want to say anything that offends either side, and they have been unbelievable partners. We get their scientists and engineers helping us.

I'm trying to call this as we see it and explain what we have built using their technology. IBM has done something very smart, and that's pushing back up the totem pole by taking technology that's been around on their main frames for many years and putting it on their servers.

That's this whole P5 series you see advertised. It's fantastic.

And there's -- I think actually boxes that take 4 to 16 CPUs and 16 to 64. I think there's actually boxes that take 1 to 4.

What we decided was the natural break was between -- was to break at the 4 CPU per box level, giving those to Dell and bigger boxes taking -- like the data base servers going with IBM. So the web and application servers would be Dell and the data servers IBM.

So using those, that technology, we've created this kind of a module. This module is a bunch of Dell blade servers that are typically in our case customer-facing, and they're clustered.

They're wired to IBM P 570s which are the boxes in the 4 to 16 proc range and storage solutions from EMC. That -- we did extensive testing, on one side's technology against the other, and this looked like the natural break.

And what I find so attractive about this is it gives us a very natural path to scalability. First of all, you can scale across the Dells, - which are very inexpensive as time goes on -, and the IBM P570s can be scaled in three directions.

They can be scaled horizontally You can cluster more and more of them next to each other. You can put more procs in the box.

We have four in ours. You can take them up to 16.

And ultimately, if we were many times our current size, you can replace these with the P590s, 16 to 64. Now, why that's so -- and this front line, as we see it, this is sort of -- this place where we've divided between Dell and IBM is kind of the front line of technology, the technology battle, and Dell is doing something clever.

I don't mean to slag their -- they've come up with a technology called Infiniband, that I think is going to be introduced in December, that will help them push down from the four proc box into the 16 proc box. But IBM's got -- I had never worked with IBM.

I had no idea how good they are as engineers. So this is the basic module that we're now built around.

How that plays out is one of these modules, one of these clusters of clusters, really, under our shopping, our BMV tab, our auctions, our travel, maybe something new in the future, and that's our web store tier. Then you have a back office tier, which is Oracle financials, Oracle customer hub and a click tracking system, for clickstream analysis of tracking.

I'll get to where we are actually in all these implementations in just a moment. After the back office tier there's the distribution, the logistics.

Salt Lake, and Indiana warehouses, and the hundreds of warehouses of our partners are all tied into Oracle. This is all tied together with Golden Gate, which is the ETL I described in my letter.

There's just a massive amount of data that has to be shifted around among these systems. It's all tied together with Golden Gate which then in fact ties it into Teradata, which is really the best in class data warehouse, feeds through Business Objects to Overstock users.

This is what we've spent 35 million on, this and the Oracle site license that underlines that we can use as many Oracle -- We have infinite Oracle licenses for databasing. Where we are, is the web store tier is done, the back office tier, Oracle finance is being turned on next weekend.

It's shifted now 14 days from where we hoped. The click -- but they're banging on it in the final stages of testing.

In fact, they're telling me they may consider turning it on this weekend, but that's a major -- that's a heart and lung transplant we've undergone. The click tracking system is live.

Came on two or three weeks ago. Teradata came live, also in mid July, along with Business Objects.

Customer Hub is the only piece that we're putting out in the fourth quarter, or into the new year. This may -- I don't know how this sounds to you, but I can say compared to what we had, what we basically had was under shopping an HP N-class box that was doing all of this, and it literally meant that at times we could do no analytics.

Ever. We had very limited reporting.

Our click tracking, we would actual turn off when traffic got high in the day. We had this one two or three hundred thousand dollar box.

Then we built up things around it and this has been a nine-month process. But this is the architecture and we're putting the finishing touches on it now, and it's cost us $35 million.

But I think it has been brilliant, I think our CIO, Shawn Schwegman, has done - is the best CIO in America. I think he's done a fantastic job over two years taking us from really a rinky dink arrangement into this architecture.

And what's beautiful about it is not only heavy duty, it's so scalable. It's scalable at each layer.

Okay, well that may be more than you wanted to know about technology, but I will mention this: The kinds of numbers, for three years folks have been asking me things about customer repeat rate, and all I could give was the version one, which was very superficial. Not even a good way to measure it.

And even calculating that number was weeks of someone's time, dumping things into Excel and twisting and sorting and all this kind of stuff. To give you the kind of numbers we're showing you now, like version three, and those five weekly scatter graphs that I showed you, that's all practically push button.

For the first time, we're able to get this- an unbelievable look into our customer base and all this data. But it took a massive investment to get there.

Okay. Next slide.

SkiWest acquisition. $25 million in cash.

$5 million to $6 million what we think the first year's operating income will be. I'm sure people have questions about this.

It will have some effect on our business this year. But people will have questions I'm sure.

Stock buy back program. We spent just short of $92 million, got a 1.7 million shares back, 8% of the shares outstanding.

So we had 19.9. We bought back 1.7 million.

Dropped back to 18.2. The reason that didn't -- We also had I think 300,000 of warrants of mine from the early days the company came due.

I think we're back up to 18.7 in shares outstanding. We retired 33 million in debt, paying 28 million for it.

We have 8 million remaining on the stock buyback program. Now I'm going to go into one last issue, and that is, I want to talk about the short position.

Before anybody freaks out, I'm not doing this, I've got no grudge against the shorts, contrary to popular belief, no grudge against the shorts. I'm not trying to do this -- I feel like, as odd as it may sound, I feel like I have a fiduciary duty to the shorts just like I do to the longs.

I think it's important to put this information out there from a fiduciary duty and people make up their own minds about what they want to do about it. So here is slide 47.

Here's the short interest in our company over the last year. Got up to 7 million by June 15th.

No complaints. Not saying anything wrong with that.

More power to them. Next slide.

How much volume was there on any given day. The system is slow.

The shareholder.com web cast. I hope it's not as slow for all of you as it is for me.

What the volume chart shows is that the volume has come down -- well, mine is hung up. I don't know about anyone else's.

Well, I guess we're just going to break off there. And if this ever comes back to life I'll bring it back up.

Operator, why don't we go ahead and turn to questions?