1) Is the person selling it unsavvy about business? Just like engineers suck at salary negotiations and end up working for $45k, if you suck at negotiating sales price, that has an outsized impact on value relative to any objective factor.
2) Does it solve a strategic need for the acquiring company? For example, if you're Google, it isn't maximally relevant to you that an acquisition target "only" has 10k users, because you're going to put Google Muscle behind it and distribution is trivial to you. If it solves a problem for, e.g., AdWords adoption in a particular vertical, like say local, valuation is utterly unconnected to current revenue/users/etc. (A foreign telecommunications firm put in a fairly generous offer for Appointment Reminder when it had no customers and no revenue, with the intention of slotting it straight into their product lineups.)
3) Failing the strategic acquisition factor, some web apps (typically ones you won't read about in the newspaper) are sold as turnkey businesses. Valuations are generally based on a multiple of the last year's profits, with the multiple sensitive to the amount of risk, etc. If I sold BCC on Flippa, for example, I could reasonably expect to get ~$25k for it: a wee bit less than 1X last year's profits. $50k would be pushing it. Nobody in their right mind would pay $100k for it if they were just intended on keeping it running as it is.
4) Are there multiple prospects to play against each other? Bidding wars raise prices, sometimes very significantly.
5) Is the app in a "rich" vertical? There are occasionally examples of heavily undermonetized web sites bought for anomalously high sums of money. Some SEO friends of mine love taking people's hobby projects and bolting on effective monetization. If there is an obvious route to this, then the fact that current revenues are terribly might not necessarily predict a 1 * $TERRIBLE sales price.
6) How risky is the web app's current position? Is it highly dependent on a single source of traffic (cough Google)? Is it highly dependent on a single source of traffic and has it gotten that traffic through grey hat tactics which could be discovered at any time via manual or automated review? Is it in a declining field? Is it in a rapidly evolving field where competition could reasonably be expected to eat its lunch?
7) What are the upkeep expenses, including fair market salaries for employees for any skills needed? Just because my time is free to me doesn't mean the Rails programmers and SEOs that the company grabbing any of my sites will need are also free.
Then at the end of your projection, you simulate a sale of your business by 'capping' it at a certain rate. Then you add this terminal value to the sum of all the discounted cash flows calculated above, you have the total value for your business.
Of course in hi-tech sector things fluctuate too much too quickly. But the model should be the same, you just have to adjust your discount rate and terminal capitalization rate to account for those fluctuations.
In many cases, factors listed below may also be taken into account:
* Sources of organic (free) traffic, and how diversified/stable they are (e.g. if 90% of your organic traffic comes via Google, not much good)
* Quantity and quality of (natural) inbound links pointing to your site
* Average conversion rate (even if you're not selling anything, there would probably be a 'most preferred action' that you want from the users - how many take this?)
* Even bookmarking rate can be used a valuation factor
As for your strategy for deciding how much to charge (which is a different problem, again), there are two problems I see:
1. Users don't like when prices are gradually being hiked. There are subconscious effects like number-fixation (where people are attached to the first numbers they see), which causes them to perceive future prices as excessive, and there are conscious effects like the disdain of price hikes.
2. This strategy maximizes short-term revenue. There are other considerations which are more important, like profit (revenues minus costs)--with costs that are not purely fixed, maximizing revenue will not maximize how much money you make ultimately. Another consideration is the long term--charging less may cause faster adoption, which may be better for the long term.
You never want to be completely myopic in pricing unless you don't think there's a future to your app and you want to extract as much money as possible.
Start with $500k.
Application Factors: If it has anything to do with "social", multiply by 5. If it's written in a functional language, multiply by 2. If it ends with -ify or has a .ly TLD, multiply by 3.
Personal Factors: If you live in silicon valley and hang out at hipster bars, multiply by 2. If you're far better at name-dropping than you are at producing actual value, multiply by 4.
If none of the above are applicable and you have an actual product or service that people actually use, figure 12x annual earnings, plus a premium for growth expectations.