A comment left on the previous post made me realize that it was time to discuss a subject I was going to save for latter – ethics.
While the story about the Potereo benchmarks was about relentless execution, its glib description of designing the benchmarks could be read as we cheated. Given we consciously worked hard not to, here’s what we were thinking.
We decided to work with our engineering department to create the Potreo benchmarks because we really wanted to see how our boards performed with the four applications customers told us that they used; Photoshop, Quark, Illustrator and PageMaker. These were applications we had never seen or ran before when the boards were designed. As we ran our tests, our engineering team found ways to improve our graphic boards performance for these applications and they made revisions to the boards firmware (its operating instructions.) The goal was to make our boards run really fast on customer applications – the benchmarks just reflected that.
It would have been easy for marketing to skip all of this and just write a set of benchmarks that made us look good. It would have been possible to have our graphics boards recognize a benchmark and just speed that test up, but not really be faster in the real world. All these shortcuts were available to us. And we decided not to. And here’s why.
Even in the smallest of companies ethics matter. Culture matters. As a private company you can decide that winning at all costs is your culture. You can decide that coming in first at all costs is your culture. Unless your board of directors is looking over shoulder they may never know that’s what you’re doing and no one will tell you to stop.
Don’t confuse or rationalize “relentless and focused” with cheating.
Shortcuts are easy. But besides being morally wrong, in the end they come back to bite you big time. (Think about the baseball Steroid scandal, Tour de France doping scandal, housing bubble, etc.) When your employees see that it’s “an anything goes” culture you’ll find unethical behavior occurring that you will regret. And in a big company most of it is illegal and can have enormous consequences.
If you are a founder of a startup ethics begin with you. Think through if you want to win at any cost. (I avoid these entrepreneurs like the plague.)
A final note. I’m sure at Enron and Madoff there were plaques and posters about ethics. Just remember ethics and values are about what you practice when the going gets tough. It’s the decisions that you make that might cost you an order, a sale or a higher stock price. Do the right thing. It pays off in the end.
Share this:
Filed under: Customer Development, SuperMac, Technology | Tagged: Startups, Steve Blank |
Nice post, Steve. I think this addresses a significant piece of what came up in the brief discussion of “social ventures” at the beginning of last week’s class. A major pillar behind the ideas of “doing well by doing good” is that operating a business in ethical, honest and even transparent fashion creates (marginal) value in the long term.
An example that will probably come up in their presentation/discussion next week is the adoption of “zero waste” policies or goals by decidedly competitive companies like Walmart and Nike (http://en.wikipedia.org/wiki/Zero_waste). It’s not the implementation of environmental policy for the sake of the environment, but rather that there seems to be a growing belief that environmentally-responsible and human-centric business practices lead to less waste, higher productivity, and more competitive businesses.
Or at least that’s how I understand it.
Thanks for that. I would have been surprised if this weren’t your position. But the manner in which you described the tests was to the say the least, ambiguous.
Also, I phrased my question the way I did purposefully, since the ethical line isn’t always clear. One might omit, for example, “less important” tests where one hadn’t performed as well, without crossing an ethical boundary.
Well put.
I teach teams various Agile techniques, and occasionally get told some particular trick is cheating. E.g., we take a big spec, cut the 80% that, for some core group of customers, is nice rather than necessary, and ship the 20% ASAP. Shipping 8 months early, with only half a product! That’s cheating!
Compared to what they were planning to do, or to what a competitor will do, it can look that way. But for the people who buy it and get what they need sooner, it’s not cheating at all. I say that cheating reality or common practice on behalf of the customer is exactly what startups are for.
Your Potrero benchmarks are a great example of that:. You folks may or may not have been able to beat the competition in academic hardware design, but your customers didn’t care about polygons per second, they cared about getting work done. Changing the game your industry is playing to one that better serves the customer is the kind of cheating I’m entirely in favor of.
[…] Here is the original: Startup Ethics: Albatross or Esse&… […]
I’m with you on ethics but with a caveat: people do cheat with predictable regularity. Much of the ad biz is a cheat, although I am fond of the GEICO gekko.
In the early days of my IBM employment there were a couple of internal jokes. Probably, you have heard them.
1) “Ship it; we’ll fix it in the field.”
2) The sad story of the IBM salesman whose marriage never got consummated on his honeymoon. According to his widow, “he just sat on the side of the bed telling me how good it will be.”
For all of you who are supervisors and managers who may be reading this: It is not only important to do things right, but it is important to make sure that the right thing is done right.
I’ve commented elsewhere on the distinction between benchmarking (objective 3rd party), benchcrafting (selective subset – app specific) and benchmarketing (simplistic overeliance on blunt metrics such as MHz speed). Benchcrafting can be interpreted both in good and bad light, It requires some skill and insight … for example in HPCC there are the LinPack kernels which measures raw speed, but there are also real computationlal physics codes from the TriLabs. Selective omission and or overemphasis is starting to blur the line though.
[…] Post by Steve Blank: ”Startup Ethics: Albatross or Essential?“ […]