How much can you trust website worth value calculators?

It is obvious from your examples above that those web site calculators vary by almost two orders of magnitude in their answers. Clearly there is something wrong in their assessments.

Web sites have value for three basic reasons. Those separate reasons will help you with your question.

1. A domain name itself, regardless of the web site may or may not have value. Clear commercial words have an intrinsic commercial value. Made up words that have been well promoted also have an intrinsic value (take facebook for example). The value of this domain name is based on the amount of traffic it can generate on its own.

2. The amount of traffic to a web site draws will influence how much ad revenue that site can generate. But not all web site traffic is created equally.

3. The quality of the traffic a web site generates is the final piece of the puzzle that you seem to be missing. Not all web traffic is the same. You need a way to quantify just how valuable a given visitor to a web site is.

My guess is that your web site value calculators do not take 1) or 3) above into account. They only work with 2). Why is that a problem?

Let’s say we have two web sites. One sells widgets for $1 a piece to a primarily female audience between the ages of 16 and 29. The other site sells wadgets for $100 a piece to a gender neutral audience ranging in age from 16 to 55. Each web site attracts 1,000 unique visitors a day. Which web site is more valuable? The wadget site is more valuable because of the quality of their traffic.

I really believe that those valuation sites exist to sell subscriptions and advertising and not to give accurate valuations. After all, what else produces revenue streams for them?

All website web-based valuators use only technical parameters that they can access on the sites that actually track those values, like Alexa.com, for example. They are not able to evaluate the content of the site.

Let's say you sell a website of a market research firm that has completed multiple projects with the top largest companies. The fact that the firm was able to secure those contracts and has been able to get the foot in the door is a potential lucrative opportunity for anyone trying to enter that market space. What if this is a very narrow niche market and the traffic to your site is not that great, although highly targeted.

Would you sell the site for what the poorly written algorithm is telling you or for what is a fair market value to the potential customer?

I have worked before to develop value calculator for domain names (domain appraisal). Generally most algorithms work this way: You collect some historical data for sales prices of domain names and as much as possible parameters (properties) associated with domains (number of words, length of domain name, Google search count, Alexa rank etc). When you have a domain name to apprise you train a parametric model on collected data (machine learning algorithms) and then use trained model and new domain name (with parameters for this domain) to estimate expected price and price range distribution. Some well known domain (site) name estimators work in similar way (estibot for example). But most work in more simplistic way. You can't decide the quality of estimation without knowing exact methodology.