People who are interested in startups are always in need of feedback because the ideas they come up with on the first, second, or 100th iteration will probably still have lots of flaws. Our need to constantly evolve is even codified into concepts like MVP (minimum viable product strategy). The result is that entrepreneurs are in a constant loop of asking for and giving feedback from everyone around them.
Most people agree such feedback loops are a good thing, but the trickier question is when should you change what you're doing based on the opinion of others and when should you ignore what you hear? The two schools of thought I see most often argue to embrace or reject advice based on a person's resume.
Resumes win
This approach gives weight to a person's opinion based on prior experience or successes. Many would say trust your friend's opinion more than your mother's because your friend has worked at a startup and your mother has not. Trust Steve Job's opinion more than your friend's because he is currently the embodiment of entrepreneurial success and your friend is not. It's a simple set of equalities:
Your friend at a startup > Your mother
Steve Jobs > Your friend at a startup
The side benefit here is being very easy to sell to investors, partners, and employees. You have to spend very little time convincing people of your decision with this technique.
If you ignore resumes you could lose big - remember these are (many times) extremely smart and experienced people.
Resumes lose
The alternative recognizes that successful people can be horribly wrong and if they "just don't get the vision" can be utterly blind to a great idea.
Some prominent venture capitalists famously got up and left during EBay's pitch because they thought "selling antiques online" was a bad idea. Fortune 500 execs are constantly making strategic mistakes that look obvious to others. Many think Research In Motion is the only one who doesn't know their strategy is flawed, even though the company has been wildly successful in the past.
The corollary to this is equally dangerous - not only do big resume people easily make mistakes, people with modest credentials are constantly delivering break through ideas. EBay's relatively unknown founder was right and a talented VC was 100% wrong. That was literally a billion dollar decision point.
Going against the grain or against resumes in some ways is like playing the Megabucks jackpot in Las Vegas. It's risky, many people will think you are wasting your time, but if you are actually right about something no else even understands yet, the reward could be more wealth than you could imagine for rescuing a rebel princess.
Scavenger hunts
I think the best approach frames a successful startup more like a scavenger hunt rather than what types of people are right or wrong. To have a successful startup you must collect the right "list" of "items". The list is the group of items necessary and sufficient for one type of startup to be successful. The items are ideas, strategies, and information. Therefore the goal of listening to the feedback of others should be to collect items on the list.
The items we seek are held both by people of the highest socioeconomic status and the lowest, and we ignore either at our own peril.
Only you know what's on your list so people will try to convince you their items are valuable, but that may or may not be true. They are usually not lying, they just have their own lists which may be different from yours.
In this analogy maybe a person's previous experience and/or "success", means that they are good at scavenger hunts. They are good at finding and collecting items. We should listen to these people to learn their skills and techniques. We should realize the more different our list is from the person we're listening to the easier we can be led astray. We should realize that in this scavenger hunt it's possible my mom is holding one of the items on the list.
Yesterday I gave advice to some friends on their startup. I hope they can pick out which of my items are on their list and which are not.