Won't they be angry?

When I teach people to run experiments, one of the first questions I get after they come to terms with the benefits of the experiments is "Won't my customers be mad?"

The answers are...

1. NO. You would be very surprised. There is evidence. 
2. YES and that's ok 
3. YES and that's great.
4. YES but not as much as your stakeholders if you do not run it. 
5. YES and the end justifies the means. 

The simplest and most common answer is no. They will not be mad. People are generally supportive of entrepreneurs and if you do not waste too much of their time or get their hopes up too high for something incredibly important, most people will not mind. Especially if you make it up to them with a discount to your other products, reciprocal offer to do them a favor or help them on a project, or coffee/beer/gift card. Most of the time though, even those things are unnecessary. Sometimes it is just interesting to be a part of an experiment, meet a new person and learn of a new venture. Another way to think about it is - think of a candid camera prank show. How often are those people extraordinarily mad? Almost never... and the only upside for those pranks is comedy. Start-up experiments can save tremendously valuable resources and solve real problems for the world. 

YES, they might be a little mad. Some people have a low tolerance. They just don't want to be fucked with. They believe (rightly so) that their time is extremely valuable and any interference is simply unacceptable. Life goes on. Test things off brand if necessary. Do not piss of your own existing customers, FOR THE SAKE OF PISSING THEM OFF, but if you think you can learn something valuable, test on a small number until you learn that thing... then bake the results of your learnings into your company. 

YES, and that may be a good sign. Again don't go pissing people off for its own sake, but remember that one of the biggest tenets of lean is proving TRUE DEMAND. There is no stronger sign of true demand than a bit of anger at the fact that your product does not exist (yet). This is a good sign. It is also something that is fixable. If you have enough of these incidents, you are probably going down the right path and solving an important problem. There is a good chance you will be able to deliver on this promise eventually. Use good judgement here. Do not promise something where relying your promise and not receiving it causes more harm than disappointment or modest anger. (Use the golden rule here, think about how you would feel. Avoid promising large sums of cash, and of course cures to illnesses that you cannot deliver). I have also heard of founders who look at these cases as nothing more than data. They believe that customers during an experiment are disposable and exist just for the data. I personally think this is a bit extreme but at least 2 of these companies are quite successful. Use your own judgment.

Also -  Guy Kawasaki says "Don't be afraid to polarize people." It is way better for 2 people to love you, 2 people to hate you and 6 people not to care, than to have 10 people kind of like you. I kind of hate to suggest this, but most entrepreneurs need to get more comfortable with the notion that you can't please everyone. You might as well start with an activity that will add tremendous value to your project - running good experiments. 

YES, but testing is just flat out necessary. Sometimes I will tell a client or mentee, would you rather waste 5 minutes of someones time or 100k of investor money building something no one wants. 

YES, the risk may be worth the reward. Lean is about applying the scientific method to business.  "Real" scientists use this mentality to cure diseases. Along the way though, they run real experiments on real people and some of those people get the placebo. They get the sugar pill. It's not an easy decision. Just like most things in entrepreneurship (and life in general) the Goldilocks principle applies. You need to balance risk reward, cost benefit, respect  cavalier(ness?) to a point that is juuuuuuuuuuuuuust right.

Do you remember Facebook Beacon?

Exactly.