One of the more interesting topics of PCB design, especially on the small-scale or enthusiast level, is the price at which you should sell products at.
Looking at prices on eBay for products in Hong Kong, it seems impossible to compete - it can actually be downright discouraging. A cheap voltage regulator board may go low as $5 USD. However the hidden indirect costs are shipping wait time, lack of documentation, poor finish quality, and counterfeit parts.
Some of the items on Tindie sell very cheap as well. I took three items that were relatively useful and computed the component and board cost, and they barely cleared 10% margin. This is assuming the designer was paying himself nothing. That is just bad business, and hurts the industry as a whole as it drives the market down to unsustainable rates. Luckily these people quickly vanish to obscurity as they soon discover they are not profitable; but there is always a new individual soon to follow and the issue remains perpetual.
It is very important to know the difference between revenue and cost; and that your time for designing schematics, footprints, and PCB routing must be taken into consideration, and be multiplied through at minimum wage (which isn't particularly proper either)! It would also not make much sense to sell at less than 12% margin (revenue-cost)/(cost) as one could easily get a mutual fund that will out-perform with low risk. That is just smart, sensible business.
The inherit problem is volume however. This is where the wildest, most unpredictable, number is factored into price. You may sell one board a month of a particular design, but if you divide material and labor into a low volume like 10 boards, the cost is going to be noncompetitive. For that reason I usually go on a baseline non-aggressive assumption of 1 board/month for 8 years; or about 100 boards volume. It's up to me to market hard enough and push a product to surpass that target. The risk of not hitting it is on me, not the consumer, and thus if I don't hit it I only have myself to blame. You can't pass poor volume costs onto the consumer who is only buying a single board, otherwise you won't sell anything!
That said, the best approach seems to be itemizing your component costs, marking them up 12-15%, adding in your time at somewhere above minimum wage or your local market rate in an electronics occupation, adding in the price per fabricated board for a 100 board order, and adding in the shipping cost for importing components spread across about 20 orders. (i.e. don't be overconfident and order 100 qty of each component immediately). I've done this numerous times and the sweet selling spot always falls around 3-5 times the component cost. The typical single-function board with no exotic ASIC style chips generally pings around $45-$55 USD. Anything less I am often skeptical that there is any profit to be had; and that is a hobby, not a business.
There have been articles on Kickstarter that show $50 USD is also the sweet spot for pledges as well. Offer something lucrative at $50 USD and your Kickstarter will have more success than a lucrative item at $10 or $100. Coincidence? Unsure, but fifty dollars is by no means excessive in a world where we will pay organization nearly $1000 for a phone that we know has little component cost, but huge corporate overhead cost. There's bit of a communist appetite to be had in this field as well; $50 is a small fee to encourage a budding entrepreneur or smalls scale designer to develop more projects and grow. Moreover you get in return a direct line of communication to offer ideas, revisions and make improvements. The best defense for those who snicker at a board price of $45-$55 USD as too high, is to open source your project and allow them to source components, fabricate the board, and assemble it themselves - suddenly they will realize the true value of your time.