I routinely tell clients "I am happy to take your money for X, but I do not believe doing so is in your best interest. Here's why: ..."

This can be anything from "Client believes they want X but X just isn't a good idea." (not too uncommon in SEO) to "Client wants X and X is a great idea but they could get X implemented for a fifth of the price if they weren't using me."

People hire me because I'm trustworthy and give them the warm fuzzy feeling. Those, and my own conscience, are much more important to me than squeezing an extra couple thousand bucks off of a contract.

I'll also tell variants of this to clients when sizing up contracts. For instance, smart people who like me routinely ask me to do pedestrian PHP coding, because I'm reliable and they don't know how to find a reliable PHP code monkey. (Canonical project description: "We have a site built on Wordpress which needs a plugin that interfaces with $EXTERNAL_API. Can you build it?" Yes, I could, but I shouldn't.)

I typically refer them to someone who I believe to be reliable, who charges much less than I do. (Though he should raise his rates, since he's reliable, and reliable is worth a premium -- as evidenced by people willing to pay $X00 an hour for PHP coding to get someone who will answer emails in a fairly timely fashion.)

It depends.

I'm sure the people building YouTube thought it was a pretty dumb and costly idea.

You could tell your client you don't think their project would take off, but you never know unless and until you try. Also, why not give some guidance towards a small pivot in your perceived good project direction?

Crass ideas like when it's about a new OS or something of the like you wouldn't take on anyway. Right?

People dont like being told things. They generally have to come to their own conclusions. Rather than tell them they are wasting money, instead ask them questions to try to better understand the value they think they are getting. It may be that you dont fully understand the situation.

True, in general--but not so much in this case. If a client is paying you to be their expert, they will (usually) understand and expect that you know more than they in this area. Part of what they should be paying you for is your experience, advice and guidance.

Tell: Maybe. Depends on my relationship with the client and the situation. Responsibility: NO.

My general rule is I am not their business advisor or market analyst (unless they hire me for those roles), something that looks crazy to me might be a great business idea/wish. I don't want someone to shy away from their dream because I can't view it from their perspective.

If I was giving advice and if New York times would have come to me for their paywall, I would have said not a good idea, but than I would have lost a $40million project and lost any credibility as a good business person because I said no to a great business opportunity.

Although I have lots of business experience, when I am getting paid to be a Developer or Designer that is where I should limit myself to.

If I am experiencing problems or misgivings with the direction of a project I will first attempt to (gently!) communicate this to my client.

When my client and I are able to engage each other in a process of solving the problem then the relationship strengthens over time and everyone ends up winning.

When we are unable to engage each other, usually because the problem is bigger than both of us, then there is little more we can do for each other but work towards an amicable parting of ways.

Learning to see it like this has been very difficult for me and I still struggle to be better at practicing it!

I always make the client prove the business case before I work with them. I have zero interest in building toiletware.

That said, I had a conversation today with a client who has a fun, cool idea that it is going to be very hard to build revenue on and very hard to market. He's 46. He's been an employee all his life. He's poised for a mid-life crisis. Guys like this either get a mistress or an MBA. What I told him is that this is going to be cheaper than an MBA, so probably we should just build the damn thing.

The devil is in the details: It's a good idea to say something before a projects starts — but if you're doing QA and the project was due a few weeks ago you might not be making any friends (even if they all agree with you).

If your project is going nowhere you can suggest options on how to get it moving — but you have to be careful not to upset other parties that may be involved, and phrase your opinion as a "what if we tried bla?"