tl;dr: Tech has a habit of peddling dreck. Journalist and editor Owen Thomas returns to the Mattermark blog to remind us that time is a flat circle.
Alex keeps asking me, “Is the Era of Bullshit over yet?” And I have to break his innocent heart and tell it to him straight: No.
Bullshit has always been with us in the tech world. When things are changing quickly and growing fast, there will be people who take advantage of those circumstances to tell you things that aren’t exactly true.
Note that bullshit is not the same thing as a lie, as Harry Frankfurt ably argued in his 1986 essay “On Bullshit.” It is not that bullshit is strictly untrue. Rather, as Frankfurt put it, “It is just this lack of connection to a concern with truth—this indifference to how things really are—that I regard as the essence of bullshit.”
How Can You Detect Bullshit?
If you can’t measure it, you can’t manage it, right? But just because you’re measuring it—generating metrics—that doesn’t mean the metrics have anything to do with what you’re trying to manage.
So here’s how to detect bullshit in the business world, the Owen Thomas way: Is a company showing you metrics, figures, or numbers of any kind? Basically, are you looking at Roman numerals next to some kind of plural noun?
Start by assuming the numbers are bullshit. Prove to yourself that they are not.
Therein, of course, lies a death spiral of paralyzing skepticism verging on paranoia. There must be a less psychologically taxing way to do this, right? Here are a few specific scenarios when you should flip the bullshit detector all the way to the Owen setting.
The rest of the time you can relax. Maybe.
Why Gross Revenues Deserve The Name
Charles E. “Junior” Johnson is, as far as I can determine, still serving a nine-year sentence for securities fraud. He went to the slammer in 2008 after being found guilty of deceiving shareholders in his business-to-business e-commerce company, PurchasePro.
Eight years before that, a magazine I worked at put him on the cover. We exhorted readers to follow people like Johnson “into the pool”—online, in the magical waters of the Internet, where endless business opportunities awaited them.
A colleague of mine, Josh McHugh, asked Johnson how much he’d made on the Web. “$500 million to $1 billion,” Johnson told him. “It depends on the day.”
That was bullshit.
In fact, Johnson’s company, which went public in late September 1999, was making about $30 million a quarter—if you didn’t stretch, as he did, to include the total value of transactions performed using his software.
At the time, I wasn’t running factchecking at the magazine. I only got that job later. If I’d been in charge, I would have redlined the heck out of that statement. Johnson was making about $300,000 a day. It was right there in his public financial statements!
Companies do this all the time—particularly marketplace businesses. They state the total value of the transactions running through their pipes as “revenues” when they only lay claim to a tiny portion of it. Enron famously did this when they swapped big energy contracts, counting the total value as revenues. Enron was the most famous financial scandal of the early 21st century. Don’t be like Enron.
The proper way to do this is to talk about your gross merchandise value, or GMV. But then you’d better be able to explain your rake, or cut, or interchange fee, or take rate—all technical terms for the percentage you claim from each transaction as your own revenue.
It turns out Johnson, in reality, wasn’t even making as much as his financial statements claimed. We later learned that PurchasePro was doing shady deals with companies like AOL to inflate its revenues. That’s why Johnson ended up in jail.
I think he should have been put away just for making this statement in a press release whining about a story in Barron’s that peskily questioned his business model: “The addition of license revenue is a revenue stream the company is able to charge as a result of the instant liquidity in its interoperable platform.”
Feel The Burn Rate
Flip Filipowski is a prolific entrepreneur (still at it, now he’s all up in the blockchain, baby!) who had a Chicago-based company called Divine InterVentures back in Ye Olde ’90s.
Divine was not named after the drag queen, though that would have been fun. Instead, it was an incubator of business-to-business companies, something which was faddish in the decadent market conditions of the fin de millénaire. (Be alarmed: It’s faddish again.)
In September 1999, Divine filed to go public. In May 2000, Filipowski gave an interview to the Chicago Sun-Times where he said the company had $100 million in the bank and could run for 20 months on that cash.
That was bullshit.
In fact, the company revealed in a revision to its S-1 that “[t]he statements made by Mr. Filipowski do not accurately reflect our financial condition.” By June it was running through nearly $15 million a month—meaning it would be out of cash by the end of the year. (Filipowski later claimed that the Sun-Times used “old material taken out of context.” A source familiar with the matter tells us otherwise—that Filipowski went around his PR team to give a fresh interview to the newspaper.)
Divine muddled through its IPO, though it had to slash the offering price in half. By February 2003, it was bankrupt.
Even if you’re not radically understating your burn rate, a burn rate is a forecast. Your revenue might go down, or not grow as quickly as you hope. Your expenses might go up. When people talk about Twitter lasting for four centuries on its cash on hand, they’re assuming that nothing will ever change in Twitter’s business. That’s not a safe assumption. So when someone talks about how long it will take them to burn through their cash, put on your kiln suit.
EBITDA, Or Earnings Before Insanity, Twaddle, Drivel, And Absurdity
In April 2009, following the abrupt departure of its chief financial officer, Facebook CEO Mark Zuckerberg told employees that the company had just seen five quarters of positive EBITDA, or earning before interest, tax, depreciation, and amortization. COO Sheryl Sandberg echoed Zuckerberg’s comments and even claimed that the company was “profitable” on that basis.
At the time, I pointed out that using EBITDA was a questionable metric. During the dotcom bubble, executives, investors, and analysts all loved using EBITDA. At its best, it made fast-growing companies look better. At its worst, it allowed accountants to shift real costs into categories that somehow didn’t “count” as expenses.
Here’s what you really need to know about EBITDA: Warren Buffett does not like it.
Three years later, Facebook revealed some of its actual financials for 2008 in its S-1 filing. It lost $56 million on revenues of $272 million. However, the 2008 figures for depreciation and amortization are missing. Mathematically, though, it seems nearly impossible for the company to have had that much D and A. (It reported no tax expense in 2008.)
Facebook’s depreciation and amortization for 2009 was $78 million. At the end of 2007, it had $82 million in property and equipment—the stuff that typically gets depreciated, like servers, buildings, and real-estate leases. It typically depreciated that stuff over two to five years. The math doesn’t seem to work for Facebook to have been EBITDA positive in 2008, even with the most generous possible estimate for depreciation of the property and equipment on its books.
The only way you might get to a positive EBITDA figure for the period in question is if you discount stock-based compensation, which was $30 million in 2008.
“But Owen, the EBITDA formula doesn’t mention stock-based compensation!” I hear you cry. It turns out some people do lump in stock-based compensation with depreciation and amortization among the noncash items they take out when calculating EBITDA. Garrick Saito, a former corporate controller, wrote on Quora that “there doesn’t seem to be any consensus” on whether to include it.
This is the problem with EBITDA. It’s not an agreed-upon standard like GAAP net income. It can mean almost anything, depending on what someone chooses to leave in or leave out. Hence bullshit.
It represents a fantasy world where the five-year-old Facebook was profitable if it didn’t have to replace the servers in its data centers or fully compensate its employees or maintain the buildings they worked in. We asked a Facebook representative to share the numbers behind Zuckerberg’s 2009 claim. She declined to comment.
This may seem incredibly picky of me, given Facebook’s subsequent success. (Zuckerberg also told employees the company was on track to grow revenue by 70% in 2009. In fact, it tripled.)
It may not have mattered. But it was still bullshit.
Groupon: Double Or Nothing
In August 2011, Andrew Mason, then Groupon’s CEO, sent a memo to employees which promptly leaked to the press. In the memo, Mason used the word “revenues” when he really meant “gross billings.”
This was a problem, because Groupon was trying to go public, and it was wrestling with regulators on the very issue of whether it could count the entire amount it charged for a daily deal—“gross billings”—as revenues. In the end, regulators forced Groupon to report as revenues only the money it had left after distributing what it owed to merchants who were actually providing the discounted goods.
This change cut its stated income in half.
Then there was Groupon’s freshly made up accounting measure called Adjusted Consolidated Segment Operating Income. Think EBITDA, but 10 times worse.
Groupon cleaned up its books and hired a chief accounting officer but never fully recovered from the air of bullshit that permeated its IPO filings. Alibaba might buy it, but Alibaba is buying just about anything these days.
The lesson: Don’t do that.
Billions And Trillions And Layoffs
Heard of Porch, the home-improvement listings service? Yeah, me neither, though apparently its CEO, Matt Ehrlichman, was USA Today’s Entrepreneur of the Year for 2014.
“To date, nearly 3 million contractors have showcased some 120 million home remodeling projects on Porch, totaling $1.5 trillion spent,” wrote USA Today’s Marco della Cava.
USA Today just published something of a mea culpa in which della Cava followed up to see how Ehrlichman was doing.
In January 2015, the company raised $65 million. In September, the company published a press release trumpeting “over $1.8 BILLION in new business” for the home-improvement professionals using its service (caps are Porch’s).
The very next month, Porch laid off 90 employees.
What about that $1.5 trillion? Or the $1.8 billion? Surely with those -illions flying around, Porch was making money somewhere, right?
Wrong. Porch has never revealed its sales, but it takes just a moment to look at its site to see what’s going on. Porch charges $100 or more a month for privileged listings. But the service professionals it’s courting can get most of what it offers for free. And many of the people Ehrlichman laid off were salespeople who likely had trouble persuading professionals to pay more for the leads they were getting.
Oh, and that $1.8 billion? Later on in the same press release, Porch added the word “potential.” That’s pretty different. When asked about the number, Porch spokesperson Jessica Piha admitted there was “no guarantee” that those leads would turn into business.
“We are a subscription model which means the professionals do not pay per lead but rather are sent a certain number of leads per month based upon the number of ZIP codes they purchase,” Piha wrote in an email in response to questions about Porch’s business model.
There’s nothing inherently wrong with a freemium model, of course. But there is something wrong with all-caps bragging about how much business you’re driving to customers when (a) you’re not actually sure if you’re driving that much business; (b) you don’t charge based on a percentage of sales; and (c) most of your customers likely don’t pay you anything.
These are good things to think about before naming someone entrepreneur of the year.
Porch recently bought a promising startup called Fountain which has video-chat technology. Fountain also served home-improvement professionals, but it planned to take a cut of the fee charged for chats in which the pros talked customers through fixing simple problems. Porch is now offering that service for free to consumers while paying professionals a fee for their time.
Go calculate the profit margin on that.
I know what you’re thinking: If Porch muddles through its current problems, do you think it will report revenues based on gross billings or net revenues?
Lies, Damn Lies, And Metrics
These examples of bullshit reinforce Frankfurt’s point: It’s not so much about saying something untrue as being unconcerned with the truth. When you’re trying to make fetch happen for your startup, you have to believe that fetch will happen. So when entrepreneurs look at all available metrics, they’re going to pick ones that most closely reflect the version of the world they hope will happen.
There’s a fine line between that necessary optimism and bullshit.