Why VA bought Slashdot

Jeff Covey has responded to the demise of freshmeat/freecode by reminiscing about his past at Freshmeat. He reports a rumor that VA bought Andover just because the CEO thought it would be cool to own Slashdot.

I set the record straight in a comment on the post. No, that’s not why it happened; Larry Augustin never would have made a business decision for a reason that flimsy. In fact, buying Slashdot was my idea, and I had a solid reason for proposing it. The rest of my comment explaining this follows.

I can supply a missing piece of the story. VA Linux’s purchase of Andover was my idea.

Here’s how it happened. It was just after the record-breaking IPO, and the U.S.’s crazy accounting rules more or less forced us to do an acquisition to maintain our valuation. There were four acquisition targets short-listed: Andover, SGI, SuSE, and a Linux service business I forget the name of that cratered spectacularly not long after. The other board members argued back and forth but were unable to reach a resolution

Up to that point I had been pretty quiet in the board meetings, keeping my mouth shut and my ears open. It was with considerable surprise that I realized that I remembered something basic that the other directors had forgotten – most mergers fail through cultural incompatibility between acquirer and acquiree. I’ve never been to business school, but I hear that anyone who does learns this early.

So I stood up and reminded everybody about the compatibility issue and made the case that our first acquisition needed to above all be an easy one. Then I said “At Linux conferences, think about which of these crews our people puppy-pile with on the beanbag chairs.” Light began to dawn on several faces. “The Slashdot guys. It has to be Andover, ” I said.

Silence. The king-shark VC on the board, Doug Leone, thought for a moment and said “There’s a lot of wisdom in that.” And so it was decided.

That is the only time I recall driving a major decision at VA, but it was enough to earn my stock options. Media businesses like Andover don’t deliver huge growth, but they’re reliable cash cows. Which turned out to be exactly what VA needed to survive the dot.com bust and eventually morph into GeekNet.

23 thoughts on “Why VA bought Slashdot

  1. >link to Jeff’s page is broke (looking @ this page’s source, URL is good, HTML tag is not)

    Thanks, fixed.

  2. Which accounting rules drove the decision to make an acquisition?

  3. >Which accounting rules drove the decision to make an acquisition?

    IPOs leave you with a pile of cash. Some things you can buy with the cash are considered assets for purposes of net-present-value accounting, some are not. If you use it to hire N people into your company (because you expect them to enable you to make more money) they aren’t counted as assets. If you buy another company employing N people, the book value of the company is counted as an asset.

  4. > If you use it to hire N people into your company (because you expect them to enable you to make more money) they aren’t counted as assets. If you buy another company employing N people, the book value of the company is counted as an asset.

    Well, duh. The 13th Amendment pretty much makes it illegal to count a person as an asset, so their wages/salaries, benefits, and associated costs (e.g. unemployment/worker’s comp insurance) are all considered “expenses” under GAAP. How would you go about changing the accounting rules to be less “crazy”?

  5. >How would you go about changing the accounting rules to be less “crazy”?

    Probably by estimating their revenue per employee and treating the time-discounted value as an asset. It wouldn’t be any crazier than “goodwill”.

  6. @esr: “most mergers fail through cultural incompatibility between acquirer and acquiree.”

    I don’t know if I’d say “most”, though it is a major failure mode.

    One other notable one revolves around failures to understand what businesses the two sides are actually in. This tends to bite financial guys who look solely at the numbers, and sometimes fail to realize that different businesses are *different*, with different business models, sales cycles, and everything else, and management that works for one of the merging outfits *won’t* work for the other. (Look up Bill Agee and Morrison-Knudsen for a grimly amusing example of what happens when a financial guy takes over a business unlike what he had been in before. The tl;dr version is that he damn near killed the company before the board had enough and kicked him out. Stories like that are why I read the Wall Street Journal for fun.)

    Another failure mode revolves around perceived synergies that turn out to be elusive. Publishing got savaged by it as traditional publishers got acquired by media conglomerates that saw synergies in having all forms of content under one roof. The fly in that ointment is that book publishers *cannot* generate the kind of revenue and profits movie studios can, and were under pressure from corporate parents to meet impossible targets. Some of those combinations have since come undone because senior management had a fiduciary duty to invest funds where they could generate the highest returns, and publishing wasn’t such an investment.

    Both of the above include cultural incompatibilities as part of the package, but cultural incompatibilities are normally cited as a primary factor when mergers fail between companies that are in the same business, where different cultures are the main sticking point.

  7. IPOs leave you with a pile of cash. Some things you can buy with the cash are considered assets for purposes of net-present-value accounting, some are not. If you use it to hire N people into your company (because you expect them to enable you to make more money) they aren’t counted as assets. If you buy another company employing N people, the book value of the company is counted as an asset.

    The only reason this hack worked for you was because you were living in the era of crazy bubblenomics when hiring lots of employees was by itself sufficient to convince stock market investors that your company was worth more. When you acquired Andover, their market value got added to your book value, with the difference between Andover’s market value and its book value being added to your books as goodwill. In a less exuberant market, this goodwill would have been much smaller and you’d have been just as well off hiring talent rather than acquiring it.

    Employees aren’t assets. Depending on their role, the things they produce might be assets. In particular, if you’re paying employees to produce software for production use (as opposed to researchware), then you can treat their salaries as a capital (rather than operational) expenditure and add that money to your books as an intangible asset. See FAS 086.

  8. >See FAS 086.

    Your explanation is internally consistent, and I have no reason to doubt it describes current reality, but it does not quite square with what I remember being told in ’99 – which is that hiring talent would do zip for the valuation, not merely less. I could have misunderstood, or FASB standards might have changed; dunno which.

  9. It was a little confusing because I’m not familiar with freshmeat but it looks like “reshmeat/freecode” should be “freshmeat/freecode”.

  10. Difference between adding personnel by acquiring a company, and by hiring.

    With the company, one gets the work the people have already done, and the revenue stream that work is now generating, if any. Or the work done could be applicable to the acquirer’s future development, and the staff is valuable as the people who can best complete that development, such as Apple’s acquisition of NeXT.

    There can be situations where previous work is not of value to the acquirer. For instance acquisition of a contract programming operation with a talented stable. An extreme case might be acquisition of a company whose previous work is of no further commercial value to anyone (the target went away) but demonstrates the talent of the staff. Buy them and put them to work on other things, not even as a team. That I think would a pure example of “hiring by acquisitiom”, but then the acquiree wouldn’t have much book value.

    Side note: there is or was a situation where personnel were treated as assets for accounting purposes: pro team sports. The right to a player’s services would be assigned a value that would depreciate over the term of the contract. Browns/Indians/White Sox owner Bill Veeck wrote about this. But there the asset wasn’t having the player on the team, it was having the player bound to the team.

  11. >Probably by estimating their revenue per employee and treating the time-discounted value as an asset. It wouldn’t be any crazier than “goodwill”.

    That begs the question, assuming that “revenue per employee” is actually a valid statistic. Does doubling the number of employees then double the revenue? Great! Let’s hire everyone on the face of the planet. Imagine how much revenue we’ll have then!

    The money spent on those employees’ wages/salaries/etc. purchased their labor for a specific period of time, which was thereby consumed. It must be accounted as an expense for that period. To blindly assume that all of the cost of those employees is captured in some asset is straight out of Marxist Labor Theory of Value or some other cargo-cult economic theory. To do that for a software company, wouldn’t you pretty much have to buy into the very manufacturing delusion against which you argued in CatB?

    You can argue that some of that labor went into building something of enduring value as opposed to being an expense, (such as when you hire masons, carpenters, plumbers, and electicians to build you a house, at the end of which you can point to the house as an asset) but then you have to figure out how much that enduring value is. And that requires some third party appraisal.

    “Goodwill” is essentially the “binding energy” of a firm. It is no crazier than the fact that a proton’s rest mass is two orders of magnitude above the sum of the rest masses of two up quarks and a down quark. It too is determined via an independent appraisal.

  12. @esr: “I could have misunderstood, or FASB standards might have changed; dunno which.”

    I think there’s a confusion here between book value and market value.

    FASB98 says that under certain circumstances, salaries of developers can be capitalized and treated as assets. The asset values become part of the book value of the firm. The market value is rather different, and is reflected in the stock price. Will adding said assets increase the value of the stock? Most likely not. Since it’s all about the stock price, what you heard in ’99 was valid.

  13. > With the company, one gets the work the people have already done, and the revenue stream that work is now generating, if any.

    This is the important distinction. The wages/salaries/benefits/taxes/etc paid for the privilege of having someone work for you are treated as expenses. What they do can be sold for immediate profit, or it might build value in an asset, but neither of those things necessarily happens simply because someone is employed.

    > But there the asset wasn’t having the player on the team, it was having the player bound to the team.

    Right, the Reserve Clause is an exception to the Thirteenth Amendment. I forgot.

  14. I don’t know the particulars of US legislation, but what Daniel Franke says rhymes well with general accounting principles: wages and salaries paid for constructing a building are not expenses, they are part of the value of the building as an asset.

    It is exactly this principle – that assets are valued based on how much they costed in the past, and not based on how much they will be value in the future at the point of selling it – that make ESR’s idea about discounting back estimated future revenues contradict the basic philosophy of accounting.

    It also demonstrates nicely why accounting is not a proper tool for business decision making – as business decisions would of course be based on estimated future revenues not on sunken costs in the past. Accounting is more of an elaborate way to document history than a tool for decision making.

    BTW part of my life as an ERP guy is about trying to find compromises between accounting rules and producing information that is more or less sane for business usage. It is quite bizarre, to be honest, how big the gap is.

  15. You know, I have seen this exact story written from someone else’s point of view, but I can’t remember where or who. It was probably somewhere on Slashdot.

  16. No relation to Doug. :) I’m a CPA. *drum roll please* I use Linux. Since I am only a very low intermediate shell scripter, I don’t know that I could call myself both a CPA and a hacker, but that would be an interesting mix. (I’ve read the dictionary you maintain…)

    Daniel Franke is accurately describing the accounting, not necessarily contradicting what you said – pretty impressive to have retained the fact that a purchase of another company produced an asset.

    One thing that is easy to forget – a company’s stock price is not hard tied to a company’s net worth.

    In theory, stock price is supposed to be dictated by the present value of future dividends, but the market often violates theory…. and in fact, from what you say, it seems that other things seems to have driven stock price at the time.

    Personally I would have thought buying SuSE would have been a good move, along with a hardware vendor.

    I just happened to recently look at some of the SEC financial statements (unaudited) – and basically they said that the cost to make the machines was mostly equal to the sale. That was a big disconnect between what Larry Augustine said in RevolutionOS…he was able to build cheaper Sun Microsystems machines with Linux.

    The market might have predicted Linux displacing Unix in server space, but the price per share seems to have have been too high to reflect that displacement, even if unrealistically it was assumed completely.

  17. >Personally I would have thought buying SuSE would have been a good move, along with a hardware vendor.

    I’ve often thought in retrospect that SuSE would probably have been the second-best choice. There was the run-by-Germans thing, though, which is a culture divide often found to be a source of problems when American and German engineering companies merge.

    The Germans think the Americans are sloppy, undisciplined, too willing to cut corners on quality, and prone to not doing their technical homework; the Americans think the Germans are ramrod-up-the-ass control freaks who cope with novelty poorly and so prone to delay and fuss about details that they chronically lose the bigger picture. Because there is truth in both indictments, hilarity (and a failed merger) often ensues.

    Daimler/Chrysler was probably exhibit A for how badly this can go.

  18. Side bar – I appreciate all you have done, starting Fetchmail and other stuff. That’s my unapologetic plug.

  19. @James Leone: “One thing that is easy to forget – a company’s stock price is not hard tied to a company’s net worth.

    In theory, stock price is supposed to be dictated by the present value of future dividends, but the market often violates theory…. and in fact, from what you say, it seems that other things seems to have driven stock price at the time.”

    In reality, the market seems unmoved by such things.

    Remember the famous dot com boom, where the assumption was that the Internet was a whole new paradigm, and that stock values of Internet firms would simply continue to rise endlessly, regardless of the absence of little things like revenues and profits? Netscape’s Marc Andreesen whined back then “They don’t understand our business model!” when analysts pressed him on when he expected to make actual money. Gee, Marc. Maybe you didn’t actually *have* a business model.

    Warren Buffet commented back then that if he ever gave a graduate finance course, his final exam would consist of one question: provide a value for an Internet firm. Anyone who actually *provided* a value would *flunk*.

  20. The ‘Linux service business I forget the name of that cratered spectacularly not long after’ would have been Linvanta or Levancare, or something like that. ;->

    Gosh, I would have absolutely no idea about the felonies^Wcircumstances involved.

    Rick Moen
    rick@linuxmafia.com

  21. >Gosh, I would have absolutely no idea about the felonies^Wcircumstances involved.

    Oh, I didn’t forget that part, Rick.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>