This phrase shows up periodically, like a comet, in close orbit around the startup world, particularly the Silicon Valley part. Fred Wilson wrote about it a few months ago.¹ It’s in Marc Andreessen’s twitter bio. It originates, as far as I can tell, with Paul Saffo, who says he started using it in the 1980s. What he means by it is something Popperian:

“Since the mid-1980s, my mantra for this process is “strong opinions, weakly held.” Allow your intuition to guide you to a conclusion, no matter how imperfect — this is the “strong opinion” part. Then –and this is the “weakly held” part– prove yourself wrong. Engage in creative doubt. Look for information that doesn’t fit, or indicators that pointing in an entirely different direction. Eventually your intuition will kick in and a new hypothesis will emerge out of the rubble…" ²

Terrific idea; the heart of the Scientific Revolution (or at least, how the SR likes to see itself in retrospect.)

But like many terrific ideas, there’s more value to wring out of it.

When Flashpoint founders “get out of the building” to talk to customers, they don’t go to ask questions. That’s for three reasons.

1. You don’t get answers you can rely on. People don’t know why they do things; they hide problems from themselves and from you. They misremember what they’ve done, and misattribute why they did it. This is a huge problem. Much of the work in Behavioral Economics is focused on sussing out the biases and contradictions around what people think they think.
2. When you ask a question, you’re “privileging the question.” For example, if you ask a customer what color they like, you’re fixing, in their minds and yours, an idea that color matters. This is a huge problem and it goes deep. There’s a great piece on LessWrong.com about it.³
3. Interacting with people in any way is creating a relationship. And if you’re asking questions, you’re creating the wrong relationship. The relationship between an interviewer and an interviewee has all kinds of characteristics. It’s hierarchical in one direction or the other – sometimes questioners are interrogators demanding the truth, sometimes they’re seekers trying to receive wisdom from the oracle. But in either case, the other person can’t help but fall into the role of suspect or oracle. These aren’t the roles that matter to a startup. As a founder, what you’re looking for is a relationship with customers; a reciprocal relationship in which you give something they can’t or won’t provide for themselves, and they give you something you can’t provide for yourself (usually money in your bank account.) That reciprocal relationship is something like a partnership. Let’s call it a “partner-customer” to keep it distinct from other definitions of “customer.” Even if you’re the seller and they’re the buyer, it’s helpful to think of the relationship as reciprocal, and that you’re both partner-customers.

“Strong opinions weakly held” doesn’t just mean create a hypothesis and try to disprove it. It also means “show up.” You, yourself, the founder as a human being, needs to confront another human being; someone who you believe might become a partner-customer. You need to show up not as an interrogator or seeker, you’ve got to apply for a job in their life.

To do that, you (in the form of your startup) have got to identify the job correctly and present yourself as the person for the job. That’s the strong opinion part – if you don’t authentically act the part of knowing the job and knowing that you’re right for it, they’ll never hire you. If you show up in a subjunctive mode and say “If I were to make this, would you buy it?” Then you’re not asking them to hire you, you’re asking for their opinion. No number of positive opinions adds up to a job, so don’t depend on great-sounding opinions. Try to get the job.

But of course you’ll be wrong most of the time. That’s where the “weakly held” part comes in. Most of the time the “customer” doesn’t actually need that job done. For example, maybe they complain constantly about something, but the complaining itself gets the job done and nothing more is required. Or maybe they do have a job, but they’ve immediately sized you up as not fit for it. Luckily, that interaction is not your only shot. You can refine your understanding of the job, or abandon it all together for something else. You can refine how you present to the next possible customer.

At least since Paul Saffo in the 1980s, good founders have known that they need to hold their strong opinions weakly so that they’re open to being proved wrong and changing directions, instead of blindly barging down a failure path.

The new thing here is that it requires the strong opinion, it requires going in as someone to who presents themselves as a customer-partner, in order to get the signals that point you in a better direction. You don’t want to iterate and hone in if you’re not yet even in the ballpark. You don’t want to pivot at each rejection, because you’ll pivot constantly – what some investors call, not pivoting, but pirouetting.

To avoid those to pitfalls, you need need reliable signals. And the only reliable signals come from authentic reactions to the authentic proposal of customer-partnership.