The difference between “fail fast” and just failing

Peter Kim is a smart guy and I usually agree with him. But a friend of mine sent me a post he did a couple of weeks back where he explains “Why fail fast is bad advice.” I could characterize his argument, but it’s better that you go read it yourself, so I don’t prejudice you. (I’ll wait right here.) Peter’s point is that he’d rather help brands succeed, and it’s hard for that to be controversial. But his examples of failing fast are not ones of failing fast–they are ones of just plain failing. As much as I usually agree with Peter, I don’t think he is making sense on this one.
Now, I have a vested interest here, having written a book called “Do It Wrong Quickly,” so take what I have to say with a grain of salt, too. But whether you use the title of my book or you call it fail fast, Peter is off-base when he says that you can’t apply that principle to marketing.

He agrees that you can apply it to R&D. And that makes sense. It’s low-risk to try things in R&D to see what might become the next product idea. But he says that it doesn’t work in marketing, citing many mistakes that lead to failure, not just fast failure that you recover from. Peter’s examples do not show how to fail fast–it just shows how to fail. These are not experiments to learn from the results–the way an R&D project does–these are just screw-ups.

Fail fast is not a rationalization for every mistake that we make as marketers. Instead, it is a conscious strategy to run experiments, see how they work, and expand, change, or terminate them based on what customers need. I’ve been writing about the fail fast marketing philosophy for years, but when I say you should “Do It Wrong Quickly,” I am not actually encouraging people to do things wrong–if you are anything like me, you need no one exhorting you to screw things up. Rather, I am trying to get us to all admit that most of what we try is wrong. It’s not the best–sometimes it’s not even close.

And most of us sit paralyzed to try unless we are sure it will work. Marketing has traditionally been such a high-risk, high-cost operation that being cautious was an extremely rational idea. And it is still a rational idea when what you are doing is high risk. But not every kind of marketing is high risk. The copy in your product catalog can be changed every minute of every day, so as long as you aren’t doing something embarrassingly dumb, you can feel free to experiment and see which version of the copy gets more clickthroughs. That is what direct marketers have always done, and it works even better in digital marketing than in direct mail.

So, rather than saying that fail fast doesn’t work in marketing, we might need to think a bit more deeply about the nuances of each situation. Something high risk (a sexy video targeted at our accounting clients) is not the best experiment. Playing with the headline in your paid search ad is. Ask yourself, if this is wrong–really wrong–how much damage can it do to our results, to our image, to our relationships? How much does it cost to try? How easy is it to undo? Can we tell whether it is working or not? How quickly? The less damage it can do, the cheaper it is, the faster it can be undone, and the faster you can tell that it is wrong, the more likely it is that you can use the fail fast approach to marketing.

Don’t look at a few dopey mistakes and think that “fail fast” condones them. It doesn’t.