Digital River: Bridge Between Two Revolutions
The middle of the nineties will be marked as a transition between the two great eras in computer technology: the era of software and the era of the internet. Digital River (DRIV) is an interesting company that has one foot in each of these trends.
Business Model
Digital River sells downloadable products directly over the Internet. They currently "carry" over 30,000 software applications and 70,000 images and fonts.
Digital River does not sell software directly to the end-user. There is no "digitalriver.com" site for you to purchase software. In this sense, it is not a direct competitor to something like Beyond.com. (BYND).
Instead, Digital River's model is that of an outsourcing service. They host all of the transaction and inventory portion of a software sale over the internet, for the benefit of the software publisher. The client company builds its own web site and provides links for purchase and download of their software. Digital River provides the transaction processing and download capability to complete the sale.
An example of this relationship is JASC Software, makers of Paint Shop Pro. JASC hosts their own web site, and offers Paint Shop Pro for sale directly. If you choose to have a shrink wrap box sent to you, JASC handles the transaction. But if you choose to download the software directly over the net, Digital River handles the transaction. For this, they take a percentage of the total sale, typically 16%.
Financials
Digital River's revenue growth, at first glance, looks unbelievable. Revenues in 1996 were just $110,000, rising to $2.5 million in 1997 and nearly $21 million in 1998. However, it is extremely important to note that Digital River counts the entire revenue received in a transaction and then uses the amount paid to the software publisher as the cost of goods sold. (Revenue shown below, in thousands.)
1996 1997 1998
111 2,472 20,911
This is actually a misleading approach to viewing Digital River's business because it is really a business model with no cost of goods sold. They do not buy inventory in advance. Although there is nothing wrong, from an accounting perspective, with counting all of the sale as revenue, an investor should look beyond that in this case.
If you count only the commission payments to Digital River as revenue the picture looks very different. (Revenue shown below, in thousands.)
1996 1997 1998
16 420 3,242
The numbers above are Digital River's true revenues as they derive no cash flow benefits from the total amount received, other than any float they may enjoy before they pay publishers.
Nevertheless, the growth numbers for Digital River are pretty impressive. At the recent Robertson Stephen conference, Digital River stated that their site has 361,000 registered users, relationships with more than 2,500 publishers, and over 2,000 transactions per day. Most of Digital River's business stems from business-to-business sales, although they are the outsourcing vendor for CompuUSA's software downloads.
Losses are considerable, however, as the table below shows. (Earnings shown below, in thousands.)
1996 1997 1998
(689) (3,485) (13,798)
(0.13) per share (0.47) per share (1.01) per share
On the plus side, Digital River has plenty of cash, with $74 million in the bank (as of 12/31/98), as a result of its December secondary offering.
But Software Business Models Are Fading
In the 1980s software business models became the dominant fashionable model on Wall Street. Companies like Borland and Lotus were outstanding stock performers. The essence of the model was the idea that the product value far exceeding the cost of delivery the software. Software was, in fact, one of the first business models ever with gross margins of 95% or more. Lotus 123 originally listed for $495 and Lotus sold more than $53 million worth in its first year of business, and was bigger than MicroSoft (MSFT), which had only MS-DOS in its arsenal at the time.
The overall trend in software, however, is towards lower prices, especially for "shrink wrap" applications. Ever since the great pricing battle started by Borland in the summer of 1993, software prices for all point products have been falling. The "suite" wars also had a lot to do with driving office applications down to sub-hundred dollar levels, and today most computers come with office applications preinstalled. The only company still commanding "high" prices for individual copies of widely used software products is MicroSoft.
A second trend, currently young, but getting stronger, is the trend to web based applications with very little processing on the client end, the "thin client" trend. This trend is making itself most present in the enterprise software world. Enterprise software is often sold on a "per-seat" basis. As you add more users to a PeopleSoft system, you must pay additional license fees. But the trend towards web applications is destroying the opportunity to sell "per-seat" licenses. Some web companies are trying, such as Concur Technologies (CNQR ), but being able to access a web application from anywhere does not fit well with per-seat licensing models.
A "per-seat" model is perfect for Digital River, as the download model makes it easy for corporations to add more users as they grow. But that model is facing a lot of pressure lately.
The other major problem facing Digital River is that eventually, software publishers will likely create their own downloadable services, as tools are created to make it easier to build such a site and prices come down. The major advantage Digital River has is that it gets a publisher into the internet download business immediately. That advantage will not last forever.
Briefing.com Analysis
Digital River is certainly one of the most intriguing Internet companies because it has a foot in the previous revolution, PC software, and the current revolution, the Internet. It is logical, from a technology point-of-view, to sell software over the internet, as it basically reduces the COGS (cost-of-goods-sold) to zero! However, with the overall trend in software being driven to lower point prices, Digital River is headed straight into a business whose prices, long term, will be squeezed. Since Digital River operates on a percentage, the question becomes whether they can grow the breadth of their business before the price pressures depress their revenue growth.
An example of this is yesterday's announcement that Digital River will provide download capabilities for Block Financial's TaxCut product. Although this product must be purchased every year (a good thing), we've seen it in stores for as little as $9.95. If the online price gets driven this low, Digital River may get less than $2 per transaction.
Digital River is obviously clearly aware of this inherent problem in their business model. In their presentation at the Robertson Stephens conference last week, Digital River made a strong point of emphasizing that selling software is not the only component of their long term business strategy. A key component of the future will be creating new vertical applications, outside of the software industry, that will generate new per-transaction fees, and leverage the existing technical infrastructure. No details were given, but CEO Joel Ronning hinted that news would be forthcoming in the next few months.
Working in Digital River's favor however is the growing trend towards downloading software over the internet. Unfortunately, we strongly believe this will also force prices lower, much lower, in the future, for all software that fits Digital River's model. The long term future for point products is free software, or software by subscription, neither of which help Digital River.
But, with a market capitalization of $721 million, Digital River is richly priced at a Price/Sales ratio of 25. But if you count only Digital River's commission as revenue, which we feel is more accurate, the Price/Sales ratio is an absurd 225, putting it off the scale for comparison with any other Internet company. And with details of the strategic future unknown, this makes Digital River even riskier.
In summary, Digital River represents what we think of as "walking on lilypads." By this we mean that Digital River can use its existing business model, software, which is facing a future of squeezed margins, to branch out into new areas. However, it must successfully accomplish this transition before the current model weakens. The question is rather it can make the transition, before weakness in the existing models hurts them. It is incredible that a company growing this fast has to think about this now, but that only demonstrates how quickly the Internet is changing things.
Nevertheless, Briefing.com feels Digital River will be an important company to watch, whether you are invested in it or not. With one company, you watch the maturation of the PC software revolution, and the emergence of the next revolution.
|
