F E A T U R E S    Issue 1.07 - November 1995

The Relentless Pursuit of Connection

By Po Bronson

An inside look at Marquee Venture Partners, one of the most famous venture capital firms of Silicon Valley's Sand Hill Road.

If you hang out for any period of time with the venture capitalists on Sand Hill Road in Menlo Park, California, sooner or later you will see a young man come by on a very odd bicycle, constructed like a rowing scull strung between two small spoked wheels. This man is Rollo Duncan, and he is surprisingly willing to stop and explain to strangers that his precarious contraption was invented by his father in their garage the year after Rollo's mother died of colon cancer, in 1993. Road-testing the "rowbike" was Rollo's only way of sharing grief with his father. After giving you this explanation, though, Rollo is quick to interview you and find out who you are and what you do for a living. And unless you're a kook or a crackpot or a junkie, he will probably pump a business card into your hand.

Rollo Duncan is a rookie venture capitalist, and it's his job to get to know people. I'm a rookie reporter. It's my assignment to get to know some venture capitalists. The "rowbike" has been lucky for Rollo. When he started riding it, he was enrolled at the Stanford University business school and - as is customary for most second-year students - interviewing at those venture capital firms on the crown of the hill. His grades weren't particularly good, and his experience wasn't very broad, but sooner or later in his interviews - during an awkward lull, making a stab at conversation - Rollo would remark that he "bikes past here every day." At the firm of Marquee Venture Partners, it just so happened that the partner interviewing him, Colin Grissom, had just put US$500,000 into a high-end video display for exercycles.

"You ever go up to Skyline on Old Page Mill?" Grissom asked him, referring to the curvy, pebble-covered road nearby which served as one of the courses for the video races.

"No, my bike can't make those turns," Rollo answered.

Grissom asked him what he meant, and Rollo Duncan explained his rowbike. Grissom became animated. "No fucking kidding!" Colin Grissom responded, pointing his finger at Rollo's chest. "You! ... you're that guy! The guy on that thing!" Grissom went to his office door and shouted down the hallway at whoever was within hearing distance. "Hey! Hey - you won't believe it! I got in my office that guy! That guy on the wheel thing!"

And so Rollo Duncan's career as a venture capitalist was born, and Rollo was given his first lesson in how the business operates: connections are everything.

That was nine months ago, and Rollo hasn't yet brought to the firm any serious candidates for funding. He admits that his contribution has been questioned by a couple of the older senior partners, and as he says this - sheepishly, as if confessing - I sense he too is questioning his value. Tonight I tag along with him to the bougainvillea-covered Hacienda Court Hotel in Santa Cruz, where Rollo will be a panellist at a soirée for members of the Loma Prieta Hillbillies & Wafer Boys, a loose affiliation of freelance engineers who live throughout the Santa Cruz Mountains. They are creators of Usenet utility freeware, which they like to call "moonshine" since not all of it is free and clear of others' copyright. (You might remember the Hillbillies & Wafer Boys as the clan feuding for years against the Jihackers, that troop of shareware coders immortalised by Tom Wolfe in his 1994 Rolling Stone essay.) They have been lured out of their hollows by promises of meeting moneymen like Rollo, though so far the engineers seem far more comfortable chatting with the bartender than with the deep pockets in suits. They're like boys at a junior-high dance, trying to work up the nerve. The participants have worn a path in the carpet between the wet bar in the foyer and the buffet table outside near the pool. Beside each dish is a little white card, on which is written, in calligraphy, the dish's sponsor. Nobody seems to be trying Marquee's batter-fried calamari.

"I knew I should have plunked down for the chicken wings," Rollo says in self-reproach. "It's just not fair; Microsoft always gets to sponsor the beer. At just one of these things, I'd like to sponsor the keg."

Rollo grins politely as an engineer strolls past. "Hi, I'm Rollo Duncan with Marquee Ventures," he says. The guy takes Rollo's card and moves on. Rollo isn't offended. He's only 28 years old; most of the engineers have 10 years on him. "Besides," Rollo tells me, "I'll let you in on a secret: They don't have to like me. As long as they've met me. Two years from now, when they've suddenly got a bankable idea and nowhere to turn for financing, they'll remember me fondly." Rollo doesn't expect anything to come from tonight's efforts. But he's got to start somewhere. While driving down Highway 17 earlier tonight, several times he repeated the mantra, "You gotta lay pipe." Trade associations, political clubs, boards of directors, softball teams. When a senior partner turned down an invitation to speak tonight, Rollo wrangled himself in as a substitute.

During his 15-minute address to the audience, Rollo Duncan explains that Marquee Ventures is currently raising a new investment pool of $50 million earmarked for products that will sell for less than $50 retail. Marquee believes that in two years the consumer market will be so hot that shoppers will, figuratively, "be buying the shelves and the light fixtures, too." A senior executive from Kmart just joined Marquee to help sign distribution deals.

It's hard to concentrate on what Rollo says, because he's being heckled by a guy in a wool hat who's sitting in the front row with a ThinkPad on his lap. These are no ordinary jibes; they're in Jimmy Stewart's unmistakable voice, and they're coming from the ThinkPad speaker. The guy types a few words, and out comes, "Now wait just a gosh darn minute!" or, "Say, fella - you don't know where a fella could get a tie like that, do ya?" Occasionally, his wisecracks are interspaced with a "Is that right?" created by hitting the F2 key. Rollo tries to ignore him, but the guy starts transcribing exactly what Rollo says, and it gets repeated aloud a second later by Jimmy Stewart. The crowd starts laughing. With the F5 button, Stewart laughs along. Rollo gives up and abandons the podium.

Later, I'm in the bathroom trying to get some guacamole off my tie when Rollo comes in. He's being chased by the heckler, who's got the ThinkPad on a strap over his shoulder, speaking for him like a ventriloquist's dummy. His name is Jim Humes, and his voice-imitator technology is called, creatively, the Talkman. Wizard technology inserts Jimmy Stewart-isms, such as stuttering, into formal written English. The result is audibly enunciated "in real time," Humes assures us.

Humes asks for Rollo's card. Rollo is hesitant, and mumbles something about Marquee not normally investing in "stuttering devices." This is a big mistake, because Humes goes ballistic, cursing at Rollo and repeating Rollo's prediction that consumers will be buying the shelves. "Now hold on for one second there, Mister - Can't you see? Talkman software can sell for $39.95! That's well under $50!" Then he accuses Rollo of being insecure about his masculinity and, as long as I'm standing there, accuses me of this, too. Rollo shoots back with the old venture capital maxim that they invest in companies, not products. We'd like to leave, but Humes is standing in the doorway blocking our exit. Finally, Rollo gives him a card, and this seems to calm the guy down instantly.

"Say now, that'll be fine. I, I ... I'll give you a call tomorrow then, Mister," his ThinkPad says, suddenly all chummy.

We head out to the parking lot. Rollo looks beat. I don't ask him questions and just let him drive, but in 10 minutes he launches into a tirade. "There's always going to be kooks. And it's new guys like me who get 'em. People ask me if I use the Internet to make connections - and, I mean, sure, I read what people post. I snoop, sure. But if I were to post a notice requesting business plans, I'd get bombarded by the kooks. I'd have to close my account."

He cracks a window to let in some air. I had been waiting for the right moment to tell Rollo about my grandmother's formula for Mud Shampoo - three parts peat moss to one part pottery clay, softened with cider vinegar - but I realise this is probably not the time.

Rollo goes on. "In a couple more years, all those people I met in business school and all those people who knew me when I was in tech support at Oracle - all those people will start being bigwigs. They'll join start-ups and they'll head divisions. Some of the partners - I wish some of the partners around the firm would realise this. I mean, it's not like I don't know people. It's just that the people I know haven't matured yet." One of the first things the Marquee partners ordered him to do was make lists of everyone he'd ever worked with or gone to school with, and find out where they are today. Rollo discovered that he couldn't remember any of the names of people from his first year of college at Cornell. It was a traumatic year: he was stoned half the time, which led to viral pneumonia, which led to an F in Western Civilisation. Eventually he switched schools. "I kind of pushed that year out of my memory. I could barely picture my old dorm room. I couldn't come up with any names." Marquee Ventures sent him to a psychiatrist, who contrived an immersion therapy that involved smoking hashish and reciting Rousseau. With time, she helped him recover about 30 names and faces of Cornell classmates.

Those names went into a master database, called The Index, which contains all of the contacts of all of the partners, with brief bios of each contact. It's reprinted each month, and each December one copy is leather-bound and locked under glass in the firm library. The Index is the closest thing Marquee Venture Partners has to intellectual property. For this story, I was given open access to Marquee, but I wasn't even let near The Index. Rollo has now contributed 538 names to The Index, but that's not enough - one of the other gung-ho rookies has already contributed more than 1,100 names.

I ask Rollo Duncan if I have been entered in The Index.

"Oh, sure," he laughs. Then, "Of course!"

So I ask him if Jim Humes, the rabid inventor of the Talkman, will be entered in The Index.

Rollo smirks. "We'd have to establish a special code for people like him."

When I arrive at 10 o'clock the next morning, the Marquee parking lot is half empty. Partly this is because partners are out "laying pipe," but also because venture capital is not, on the whole, a very intense business. Most of these guys burned out on a previous career; the rat race holds little appeal anymore. If they weren't venture capitalists, they'd be at home reading a Bloomberg screen, trying to get 20 per cent return on their retirement savings. The older guys talk about how lucky they are to still be "part of the process," that process being the movement of capital from those who have it to those who can put it to good use. One partner compares being a venture capitalist to being a node on the Internet: the cash can get from A to B a hundred ways, travelling connection to connection. Investment banking, on the other hand, is like America Online: money has to go through this big bottleneck dominated by a few New York syndicates.

The Marquee offices reflect this egalitarianism. Every office is exactly the same size. There's one receptionist for every six partners, but nobody gets a personal secretary. Similarly, there's exactly one digital videophone for every three partners, one toilet stall for every four partners, one Fran Tarkenton football (signed) for every five partners, and one small Rodin sculpture for every ten partners. This measurable equality is written into the firm-partnership agreement, and as such it means that the quantity of partners in the firm is limited to a number evenly divisible by 6, 3, 4, 5, and 10. You do the math.

Rollo spends this morning digging through the slush pile. Each week, Marquee receives 300 business plans. These are logged in by a receptionist, slipped into a manuscript box, and passed on to a rookie like Rollo, who stacks them up in his office until they become a huge cardboard bunker. Some people have an "In" box, others have a "To Do" pile. Rollo Duncan has a "To Do" wall.

"Guess who called this morning?" Rollo says, flashing a message pad.

"Jimmy Stewart?"

"Three times."

I waste a few hours with Rollo's wall. Among the proposals is backup software that downloads your hard-disk data to a network server every night. There's an infrared headlamp-cursor to replace trackballs - you put on this miner's helmet and move your cursor by moving your head. There's slender computer cable, and a new design for a compact-disc jewel case. A book publisher has put text-only, home-reference titles on recycled 8-bit Nintendo cartridges for showing on your TV - you navigate the menus with a joystick.

If it weren't for the business plans, I'd never have realised how much I needed some of these things. For instance, did you know that the average mum microwaves hot dogs to an internal temperature of 250 degrees - which can cause scalding of mouth tissue in her children? Thank god for this new handy microwave thermostat. The house is safe from disaster once again.

One of my favourites, though, is the insulated burrito-holder. Similar to the design of a plastic coffee mug, it also has a strap that runs from one lip, down under the burrito and over the other lip, where it snaps in place. By adjusting the strap, you raise the burrito as you eat. I am tempted to take this burrito-holder home for "focus-group analysis."

If you're out there writing a business plan, I can give you some tips that will quickly make you look as if you are a seasoned professional. One of the phrases you should include is "a 40 per cent compound return on investment, with cash-out in three to five years." It usually goes in the executive summary and on the pro forma financials, but squeeze it in wherever it fits. In domestic markets alone, my Mud Shampoo offers investors a 40 per cent compound return, and your hair will grow back in three to five years.

Another thing you can do is make up new catch phrases for your industry, which convey your boundless enthusiasm and the immense power of your thesaurus. A good word to mix in is "entertainment," since media content is so hot right now. For instance, the plastic burrito-holder can be silkscreened with Disney characters, firmly planting the manufacturer in the "legume-tainment" business. Everyone wants to be an artist these days. Developers are tired of coding in C++. Cooking hot dogs to exactly 210 degrees can be described as "working in the frankfurter medium."

Thinking of food has made me hungry, so I tag along with Colin Grissom on a "milk run." Grissom has been with the firm since the '80s as a software specialist. Previously he had been a lawyer. He has contributed 8,454 names to The Index. He speaks in blunt sentences of one or two words, composed entirely of nouns. Listening to him is about like reading Basic. As soon as we get in his car, he cuts me short to deliver his judgment of what I've learned so far in my research. "Number One. Rollo Duncan. Still got the jaggies. Lacks ... what's the word I'm looking for? It was '93. Silicon fever! Partners were worried we were stodgy, we were out of touch with the young hotshots of tomorrow. Rollo was the answer. So far, Rollo's just been the question."

I ask him if Rollo is wasting his time reading business plans.

"They're obsolete as soon as the ink is dry. You want to understand venture capital? All you need to know: sperm."


Grissom rolls his tongue through his mouth like he's playing with an ice cube. He refuses to use the brakes on his car, preferring to downshift. "Two styles of people: guys and gals. Females, what? They caretake. They nurture. Men, what? They squirt and move on. So, business start-ups - same thing. The entrepreneurs who run businesses? They're like women. Caretakers. Venture capitalists, though? Gigolos. Roosters. Seed capital. Get it?"

I'm checking the batteries in my tape recorder. "You're never tempted to cross over, to join the start-up?" "I sit on their board. That's more than enough. I got my visitation rights, but I don't have to exercise them. Some guys, though. Pussy whipped. Not cut out for VC."

I kind of like Colin Grissom's manner of speech. It's not very refined, but it conveys authority and experience. I can imagine him marching into some programmer's garage, ripping off his sunglasses, and pronouncing, "Me. Tarzan. You. Jane."

We pull into a bicycle shop in Mountain View. I guess milk run doesn't mean lunch; we're here to check on that exercycle video-racing system, SpinCycle, which is being developed by some of the shop mechanics. We trot into the back room where the system is running on a few microcomputers. The long-promised alpha version has suffered recent delays. Grissom introduces me as "an interested second-round investor," then accuses the master mechanic of fabricating bugs in order to hold out for more stock options. Grissom believes that if the options are granted, the bugs will suddenly vanish. The master mechanic doesn't deny it. The mood is very intense, but I'm ready to eat a napkin. I kind of space out. I get the idea - I'm not exactly sure how this comes to me - that Marquee invested in the stand-alone units first, and when that wasn't performing at 40 per cent compound return, they ordered this guy to create an online racing system linking exercycles at health clubs around the country. On top of all this, the master mechanic pledged his bicycle shop against the software deal, and now, as if he already owns the shop, Colin starts questioning the inventory levels of the bike shop, arguing that the flywheels which have sat on the shelf for over a year can be returned to the manufacturer and special ordered when needed. And then I realise what's really going on: Colin Grissom is a thug! He's the guy the firm sends out to do the dirty work! That's why he used to be a lawyer - he can attach liens, impound assets!

Colin, why don't you and Mister Mud Shampoo take a little ride down to the County Assessor's office?

On the way back to the office, Colin Grissom just about confesses to me that his role around the firm has evolved into being the heavy. He accuses the partnership of being too gentlemanly and not being able to cut its losses. According to Grissom, in the last two years the mood around the office has grown fearful. "They're thinking, 'obsolescence.' Example: We funded development of EGA screen technology. But EGA is what? It's obsolete. So, is it possible venture capital is obsolete, too? That's one of their phrases - 'Is it possible?'"

Grissom starts talking about angels, the new source of money in the Valley, the "alternative" - he says the word with scorn, adding, "as if money were rock music." He goes on, "So many engineers made a bundle in the '70s in hardware or in the '80s in software. Plain ol' joes, now sitting on a wad of cash. Turning 50 - a half century. The desire to be young again occurs at the same time they wonder, 'What am I going to do with all this money?' Two problems, one solution: invest in start-ups. Do you think they care about the ROI or exit strategy? They're thinking one thing: endless youth. Here, take my wallet, take my bank account, build me a CD-ROM that gives me a woody."

When he says this, I'm thinking that if they ever make a movie about a venture capitalist, they'll get Al Pacino to play Colin Grissom.

"Entertainment. My god, where did all that come from? Who saw that coming?" He takes a deep breath through his nose, as if resigning himself to something he'd rather ignore. "Coders wonder - angels wonder: Is my love for this box of wires sensible? Or am I a rhesus monkey in a cage clinging to a piece of shag carpet? Entertainment is a way to get normal society to love their box of wires. But can it be done? Turn code into tears?" He shrugs. "Macbeth is just one letter after another. Straight ASCII. Input: bits. Output: transcendence."

I ask Colin if venture capitalists aren't similarly tempted to invest in entertainment. He explains that VC stays away from entertainment because it's not a two-stage investment. In a two-stage investment, you get a million dollars to invent the product. If it works and there's a market, you get another five million to "ramp up" production. The two-stage process allows you to cut your losses after the first million, when you've spent only one-sixth of the total. It gives investors a sense of having "saved five million." With entertainment products, the entire investment is spent getting to gold master. It may still only be one million invested, but they never get to save that five million.

Colin Grissom believes his partners' fear that angels will replace VC is "complete horsepucky." He predicts the angels will get burned by a deal or two and go back to Treasury bonds within two years. "With their testicles packed in ice," he adds, a moment later.

Over the next few days I spend time with different role players from around the firm. For some, their role is knowledge expertise in biochemistry or materials engineering. For others, it's hard-core business experi-ence in wholesale distribution or overseas manufacturing. They each bring something to the table. The table is in the conference room, and you've never seen anything like it. A perfect circle, it seats all 60 partners. The slice of 2-inch Italian marble was flown in by helicopter.

My favourite role player is an old guy they fondly call The Wart. In the '70s, The Wart crossed over. He bought a business that refurbished the plastic shell of standard AT&T rotary-dial telephones. The 30-employee assembly line used a lot of paint and wax. One night, the warehouse burned. He tried to save it, but was burned horrifically, leading to his nickname. Three times a day he applies a jelly salve to his eyebrows.

Around the firm, The Wart is legend. His tale serves to caution partners from crossing over and leaving the firm - his presence reminds them of what can happen if you get too involved in a business. Since his return to venture capital, The Wart has prospered. In 1984, he became mildly famous for his Manweight Theory of Analysis. Until that time, venture capitalists always asked themselves, "How do I know when I've studied the investment enough?" The Wart's elegant little method was that for every million dollars invested, you had to stack up paper equal to your own weight. It was more of a rallying cry than an actual theorem, equivalent to patriot troops of the American Revolution being told, "Don't shoot until you see the whites of their eyes." In other words, "You've never done enough analysis." The Manweight Theory became a popular mass-market paperback.

One day Rollo and I caught The Wart in the middle of performing a very tricky and unstable procedure - an equity valuation of SpinCycle, the exercycle-videorace company. He had the company's financials in front of him, and he was running various ratio tests when we interrupted his concentration. SpinCycle will be out of cash by the end of the month and needs another round of financing - but Marquee will fork over more cash only if their original investment has gone up in value. The Wart believes he can prove that it has. His plan involves spinning off some of SpinCycle's code as a shrink-wrappable authoring tool for road-racing games. He explained that any investment banker off the street can value a company by multiplying earnings, but only seasoned venture capitalists can value companies with negative cash flows operating in unestablished markets.

"Go ahead, give it a try," he says confidently, pushing his stack of paper towards Rollo. "See what you come up with."

Rollo takes up the green pen and studies The Wart's scribbles. Looking over Rollo's shoulder, I can't make any sense of it. It's been 10 years since I sat in an economics class. A bead of Rollo's sweat drops on the page, blurring one of his notes. The Wart chuckles at his agony, then gets up and leaves the room. Rollo works through the math, but slowly - and his lips are moving. It's not a good sign. Then I remember that the original investment was $500,000, and a 40 per cent return on that for two years would be, let's see, eight tens plus two hundred to the seven hundred, add three zeroes - "Seems to me the company's worth about $980,000," I whisper.

Rollo looks at me. I show him my crude methodology.

"You can't just do it that way," Rollo says, with some scorn. He bends back over the paper. I've broken his concentration, and he has to back up several steps.

The Wart leans up against the doorjamb of his office, a paper cup of water in his hand. "Got something yet?" Rollo puts the pen down. His eyes drift back down to my methodology. "Just shy of a million," he says, guessing wildly.

The Wart's jaw drops. He studies Rollo for a moment, his eyes narrowed. Then he relaxes. "That's about what I got, too," he says, then takes a sip of water.

At all the venture capital firms along Sand Hill Road, partners' meetings are held on Mondays, in keeping with scientific experiments regarding decision making and mental attentiveness. Study after study (commissioned by the VCs) has concluded with statistical significance that Mondays are the worst day of the week to play golf. Club selection is overconfident and putts are rushed. In 1982, a coalition of venture capitalists took their irrefutable test results to the rules committee of the nearby Stanford University Golf Course and argued for a "Monday adjustment" to the handicapping system. To their complete shock, they were turned down. The Wart recalls, "I don't even think they read the studies. Their response was, 'Well, if your score's worse on Mondays, then don't play Mondays!' - as if I could just skip bunker holes because I'm bad with a sand wedge."

Today's meeting begins with an approval of the previous week's minutes, then attention to a few internal matters. One partner gives an update on the efforts to raise the $50 million for the new fund. Most of the $50 million, it seems, will come from publicly held companies that Marquee invested in years ago, when those firms were start-ups. It's a form of tithe, or colonisation, and all transactions are recorded in The Index. What goes around comes around. Without a doubt, the best way to get $3 million from Marquee to pursue your idea is to have invested $5 million in Marquee before you quit your job as CEO of a Fortune 500 disk-drive manufacturer. If you were never the CEO of a Fortune 500 disk-drive manufacturer, then any other sort of Fortune 500 company will suffice. The second-best way to get $3 million from Marquee is to already have gotten $3 million from Kleiner, Perkins. This is the "Let them eat cake!" foolproof system, and it safeguards Marquee from investing in people who really desperately need the money.

Fifteen minutes of today's meeting is allocated to a discussion of what to do about Colin Grissom's latest manoeuvre. Apparently, the master mechanic at SpinCycle wasn't too happy about adapting his proprietary platform into an authoring tool. Grissom, in return, stole the master mechanic's toupee in order to keep him from going out until the program is finished. The dark-haired toupee lies in the middle of the table. Opinion around the table is divided, and the debate exceeds its time allocation. By charter, all partnership decisions at Marquee must be unanimous. Someone moves to table the resolution until the next week, but this motion fails by three votes. Another partner moves to buy the mechanic a hat. Grissom is in favour of plucking one hair from the toupee for every day the beta is late. Finally, unanimity is reached when The Wart moves to chop the toupee into thirds, then mail back each third as milestones are reached.

Eventually we get around to the presentations from the entrepreneurs. They usually last 20 minutes each and are followed by questions from the partners. There is no schedule, so the trick to these presentations is ending them concisely. At some point shortly after each presentation, all of the partners suddenly run out of questions to ask. It's uncanny. I suspect Grissom is giving a secret hand signal, and during a break I ask Rollo Duncan if that is the case. "Watch The Wart," he answers. "When he takes his dentures out and drops them in his water glass, we're supposed to cut it short."

After the break, to our complete shock, Jim Humes - the Jimmy Stewart impersonator - walks into the room and takes a place at the table. He's being sponsored by the new partner from Kmart. It looks like the new partner is going to embarrass himself terribly. I can barely watch. But as Humes speaks, he doesn't emphasise Jimmy Stewart or go to his ThinkPad. Instead - get this - in order to properly synthesise Stewart's voice, Humes has stumbled upon a revolutionary "soundfont" technology. "A standardised filter translates written English into vector-based phonetic syllables," he explains. This dramatically reduces the memory requirements versus bitmapped sound. "You apply a famous voice to the phonemes the same way you apply Helvetica Bold to ASCII." A chuckle goes around the room when Humes taps into his ThinkPad and out comes, "Say, fellas, you wouldn't know where a fella could get a drink of water, do you?"

Humes predicts that within two years his English-to-phoneme conversion system will be as prevalent as Type 1 PostScript. Still, though, the partners are nonplussed. They don't have many questions. Humes is sweating. Groping for anything, he says what turn out to be magic words: "I need half a million dollars anda few quality introductions. If I get a distribution deal, then I'll need another three million to build out the voice library. If I don't get the distribution or somebody else beats me to market, then, well - I won't need your three million."

They can save three million! It's not like saving a full five million, but it'll do. Somebody pours Humes a glass of Pellegrino. One of the partners asks if Humes can do Rod Serling of The Twilight Zone. There is a round of bad imitations for a couple minutes. It looks like The Wart is grabbing at his denture, but it turns out he's drying his upper lip to offer his own Rod Serling impersonation - "A seemingly innocent burning building...."

When they finally kick Humes out of the room, Grissom returns the conversation to a serious tone. "I don't give a fuck how good his idea is. We don't know this guy from Adam! Since when do we invest in people who just walk in the front door?"

The Kmart guy admits that Humes called him out of the blue, and that they'd never met before that morning. He wonders if they can't "buddy up" Humes with a businessman they know and trust.

Grissom continues: "Of course he'll use the buddy system. But hell - even if we syndicate shares in this deal to KP or Mayfield, this is against all principles!"

Somebody else agrees. "He's got a great product, but what does that have to do with anything?" One of the partners who did a decent William Shatner stands up. "If we don't invest in this guy, he'll end up with angel money for sure. Heck, if he got in our front door, he could just as easy get a meeting with Adobe or Apple and have his own division by next Monday."

This creates a flurry of responses. Nobody wants to see Humes get angel money. Momentum builds on Humes's behalf, but Grissom's intractable - "He's not getting my vote unless somebody here can testify on his behalf. Has anyone met this guy?"

There's silence. Rollo fidgets in his seat. He looks to me, forlorn, his jaw tight with anguish. With a great sigh of despair, Rollo pushes back his seat and stands up. "I can vouch for him," he says.

Postscript: To do the fact checking for this article, I revisit Marquee Ventures one last time. By now, Rollo has moved his office. "All the offices are the same, but some are more same than others," he says, winking. It turns out that just three weeks after Marquee put their half million in Humes, they flipped the technology to IBM for four million. Rollo's contacts are paying off, and his illuminating suggestions for lower-cost technical support no longer fall on deaf ears. Now Rollo is hot on another company, one founded by two of his fellow students from the business school. His classmates have created a new venture capital firm, but not as a partnership - it's a corporation, and it's looking for money. Marquee has pledged 20 per cent. This seems ridiculous - a venture capital firm investing in another venture capital firm? - but Rollo assures me that it meets all their criteria. "We see venture capital as a real growth industry," Rollo says. "And we want to get a piece of the action."

He puts his head back down to his work, signaling my time is up. As I'm leaving, I look back through his glass door one last time. Rollo Duncan, unsung pipelayer, reaches for the phone.

Po Bronson (pobronson@aol.com) makes things up for a living. He is also the executive director of a nonprofit organisation which has as its mission replacing US currency with yellow fin potatoes. More of his fantasy life can be found in his novel, Bombardiers.