How Fred Smith Saved Fedex at the Blackjack Table
warstory.co
warstory.co
That week the company asked employees to delay cashing their paychecks if they could.
At the time, I was doing operations research working on fleet scheduling, etc. Likely there was a lot of money to be saved with better scheduling of the fleet.
I'd written the first software for scheduling the fleet, and one evening Roger Frock (mentioned in the article) and I used my software to do a schedule for all planned 33 airplanes and all planned 90 cities. Our two representatives from Board Member General Dynamics checked the 'feasibility' of the schedule and announced "It's a little tight in a few places but it's flyable". Some members of the board had been concerned or even convinced that a schedule would not be possible. So, the schedule Roger and I did alleviated the concerns and, as I was told, enabled $55 million in funding, i.e., loans, on the airplanes.
At a senior staff meeting, Fred's remark on the schedule was "An amazing document. Solved the most important problem facing the start of Federal Express.". Yup, it had been six weeks in my living room in Maryland connected to a time sharing computer for 80 hours a week writing PL/I code while finishing teaching two computer science courses at Georgetown U.
That FedEx was close to going under didn't bother me much, but the stock I'd been promised "within two weeks" with my offer to join still was not there 18 months later and bothered me a lot.
If I couldn't get the promised stock when the company was close to folding, then staying around and helping to save the company would get me what? With no stock, I was going to graduate school to have a better career in operations research.
Then Fred called me to his office and said: "You know, if you stay, then you are in line for $500,000 in Federal Express stock." But he wasn't giving me that statement on paper with a signature. My manager, Mike Basch, SVP Planning, was there with me with Fred. My office was next to Fred's with Mike's across the hall. No, given that the promised stock was already 18 months late, the company was close to going under, and Fred was not putting his statement in writing, I didn't know I was in line for any stock at all.
Apparently none of Fred's talk about stock meant anything: With the original copy of the offer letter, all the promises, all my records from then, when I contacted Fred with all the information, he and his FedEx lawyer refused to pay me anything, and a lawyer told me that legally FedEx owed me nothing.
The work I was doing would likely have saved FedEx a LOT of money soon. E.g., a good first shot would have been to modify the scheduling software to include some accurate costing and then just use it to develop schedules with the least cost by just intuitive methods. For a second cut, generate and cost out many possible individual airplane trips from Memphis to the cities and back. Then use that data for integer linear programming (ILP) set covering with one column for each possible trip and one row for each city to be served, say, a few dozen cities growing to the planned 90. Then use some column generation (Gilmore and Gomory) and some branch and bound. Easily should have saved millions a year in fuel and other operational costs quickly. I had world expert in integer linear programming George Nemhauser lined up as a consultant, and I was writing software in PL/I on an IBM CP67/CMS system.
Actually, well into the start of FedEx, Fred had planned to do the first schedules himself. When he tried, for just a few cities, just by hand, from his office, when FedEx was still in Little Rock, at the end of the afternoon he came out of his office saying "We need a computer". A guy I had known in college heard that and gave me a call.
Yes, ILP set covering is in NP-complete. So, no one has a guaranteed polynomial algorithm for worst case examples. But that doesn't mean that can't save a lot of money fairly easily in cases of interest. Sure, might have trouble guaranteeing to save the last 10 cents, but if save $5 million a month over the best solution otherwise, then just saved, let me see here, right, $5 million a month.
I didn't get the stock, and Fred didn't get the savings.
Fred has been very successful, but he didn't do it alone, and some promises about stock were not kept.
One lesson is, from what my wife told me as I joined FedEx, "Get it in writing". Well, my offer letter said I'd be in the stock plan, but apparently that was meaningless legally. Better still, get a lawyer. Better still, start and own your own company. Back to it.
Joining as a low-numbered employee is a really big act of trust, bigger in some ways than being a founder. You ought to find out everything you can about whom you'll be working with, and what they think of you, before you consider it. It helps to have the ability to read people. Getting things in writing definitely is a must-do, but it's not a silver bullet.
The reason it's not a "bigger in some ways than being a founder" is precisely because you CAN follow the suggestions you then list
>You ought to find out everything you can about whom you'll be working with, and what they think of you, before you consider it. Getting things in writing definitely is a must-do, but it's not a silver bullet.
There is simply NO way to do that as a cofounder! The paperwork SIMPLY DOESNT EXIST. As an early employee, you can view our patents, trademarks, etch stack, talk with other employees, see our level of funding.
As a cofounder, you get to see an idea.
There is no guarantee whatsoever of ANYTHING as a cofounder, EVEN if you get it in writing. The only thing you can be guaranteed is 50% of nothing.
So in some ways you could say that joining as a low-numbered employee is for people who want to make it rich but don't have the balls not take any salary, not to have any paperwork, anything in place whatsoever that they can do due dilligence on except the person of the their cofounder and the ideas and work that they personally bring to the table with their own two hands.
as a cofounder, you're not owed shit. (except 50% of nothing).
the same thing applies to options contracts or anything else.
What I mean is as an early employee you can SEE all this in writing. As a cofounder, you also get to have your equity agreement on paper. If it's 50% it even means something. As a minority cofounder (e.g. if you're the 10% guy with the other two having 45-45) that would mean considerably less.
I just don't see how you can anywhere NEAR approach the level of trust as an early employee, that you need as a cofounder. It's not even within three orders of magnitude of each other.
And that's part of why there are so many fewer cofounders than employees in the world.
I mean think about it!! What can I say over IRC and Skype that would convince you to be my cofounder without meeeting? What kind of paperwork can I produce? What kind of documentation?
Now compare that with being a remote early employee.
The level of trust required is literally 3 orders of magnitude less. Not risk: trust. But risk too.
There are a lot of ways to lose out. If you're not a lawyer, there is almost certainly something in the contract you overlooked. The contract might not even be valid if two non-lawyers came up with it after a few beers. You could be fired right before your options vest. (This happened to a lot of people at Skype, for example.) Or you could be diluted down to nothing after a few rounds of financing.
Sure, you can sue, but is that really going to help? If the company goes out of business you're trying to get blood from a stone. If the company succeeds, they'll have a legal budget 1000x yours.
Ultimately, the founders are going to set the direction of the company. You have to trust them, and also trust that you will be useful to them over the long term. Any other strategy is just a losing bet.
I'm specifically comparing to the trust required to be a cofounder. I've been both an early employee and a cofounder. The level of trust I needed in my cofounder to enter the business on the terms we did is simply 1000x bigger than the level of trust I needed to start working as an early employee.
I mean think about it: you're talking about losing out on the upside as an early-employee, but the salary still comes (or you just leave after 2 weeks) and so does the experience.
As a cofounder, nothing "comes". There IS no business.
Your assertion that a salary agreement isn't worth the paper it's written on is ridiculous. Even paper isn't necessary: if an early employee didn't get paid at all they would leave almost immediately.
it just doesn't require anywhere near the same amt of trust.
If you read the article itself, it talks about pilots using their own credit cards to pay for landing fees in a pinch, and couriers selling their watches to pay for fuel. So basically, I would urge you to read the article and the posts you're replying to.
I admit that co-founders usually put more money into the business than early employees. But time is also worth money, and early employees put a lot of that in, for not much money up front.
I agree with you that the experienced gained as an early employee is valuable. Some of that experience might be getting a better idea for whom to trust in the future :) And with that, I think we've come full circle...
most people who can claim the title of "early employee but not cofounder" here at HN would not say they have gone 18 months without pay in that title or would do so. so yes it's a data point but it's a pathological one.
On the other hand, 18 months without pay doesn't scratch the surface of what a cofounder can expect: it's one of the best-case scenarios that anyone could possibly expect!
A cofounder would jump at a chance to see liquidity within 18 months of founding a new startup. More typical is more like 24 months absolute minimum, 36-72 months average, and often longer is the norm.
And during that time you don't just hock a watch or pay for fuel: most cofounders pay tens of thousands of dollars of their own money, often every month and putting every source of liquidity they can into the business as it ramps up.
and more often than not, as sad as all this is, most cofounders do lose 100% of their investment and have nobody to blame, nobody to even ask for a break. they have cliffs, and CANNOT jump ship. WHen a cofounder jumps ship they lose 100% of the entire equity they have in the company.
But at the same time, most startups simply don't go anywhere.
I would be hard-pressed to find a metric by which the average early employee gets absolutely nothing for his time and has to pay for all the infrastructure used to employ him.
it's just not the same planet.
of course there are pathological cases that are the worst of both worlds: early employees who should be called cofounders, are paid nothing, and do not get meaningful equity.
you don't need much trust to be an unpaid intern though.
I'm not sure where the line between reckless gamble and calculated risk is, but this feels like it's over the line. Yes, the company would have likely folded either way had the founder not gambled and won, but this strategy together with the "never give up on a particular business" cockroach mentality feels akin to winning a lottery and subsequently advising people that lottery tickets are a viable way to prepare for retirement.
Also, I'm quite sure that the founder would have found himself in a lot of personal trouble for gambling company funds on a blackjack table had he lost.
I get the feeling the Cooper Union situation was, although bad, not as bad as what Fedex was in. Cooper Union made a gamble, failed, and was still able to continue to run (2 years of debate, plus funds to run until the 2014 school year when the fees actually kick in). That tells me it wasn't a do or die situation; they still had rope left. Even now, I'm not convinced they exhausted all available, smarter avenues. I'm also kinda surprised their books aren't open if they've been debating for 2 years.
If this story is true I wouldn't want to have anything to do with Mr. Smith.
I put myself in some pretty tight personal financial spots for the success of my first company and just seeing my employees happy and at work was enough for me to never regret it. After selling the company they all got an even more stable foundation for their lives and I got the "reward" that comes to those too dumb to give up.
Here's to guys like Fred for giving me the inspiration to do whatever it took for my company!
I think somebody who is reckless enough to gamble the remaining funds of the company, is not disciplined enough to walk away from the blackjack table after having won just enough for the fuel payment.
If he actually did play blackjack, it would be more likely that he started out winning, won enough to pay for the fuel, but then continued gambling and gave it all back.
Say I'm an investor and get $10K if he goes bankrupt. If bets and wins, I have, say, a 52% chance of getting nothing, and a 48% chance of getting a million, for an expected value of $480K. The rational decision, in the best interest of investors, is to make the bet.
Even more so if he was counting cards, improving his odds slightly over 50%. Counting cards isn't all that difficult, the hard part is just doing it over time without being detected, which in this situation wasn't necessary. It's interesting that he chose blackjack, a game that punishes you badly if you're not skilled, instead of roulette.