Guesstimate is great for modeling things that aren't certain. You can forecast your earnings from a new venture, predict the amount of time completing a big project will take, or experiment with scientific theories. People have used it to optimize video games, understand lottery payoffs, and estimate the costs of childcare.
Guesstimate is most commonly used for cost estimation and forecasting within small companies. If you're not sure how Guesstimate can help you or your business, contact us at [email protected] to set up a free consultation session to analyze your modeling needs.
The most popular model on Guesstimate is How Long it Takes to Get Ready for Preschool. We also recommend checking out Ozzie Gooen's models here, and all our public models, here
Yes, Guesstimate offers paid plans for private individual or organizational models. Check out our pricing page for more information.
An in-depth look at our available functions can be found here.
Additionally, you can find a list of the available functions while modeling by clicking on the documentation widget on the right, then clicking Functions: Available Existing Functions.
The documentation widget is located here:  
Your own models autosave when you edit them. If you edit models created by other people, you will be able to modify them, but your changes will not be saved. This can be great for experimentation or for entering your own assumptions in a different model.
Guesstimate is a web application, while Crystal Ball is a suite of Excel-based applications. This means that with Guesstimate you can get up and running in 30 seconds on any computer and then share your model with anyone. Crystal Ball is better for more rigorous analysis, and gives you access to the Excel environment and ecosystem.
One great combination may be to quickly experiment with models in Guesstimate, then replicate the best model in Crystal Ball when you want to do more detailed analysis.
Ozzie Gooen and Matthew McDermott.
Yes. You can check out the GitHub repo.
We have a blog here.
Please file it in our bug tracker on GitHub, or just open up chat (bottom right) and let us know. If your bug is security related, please alert us immediately at [email protected].
Guesstimate uses Monte Carlo techniques to produce our results. The Monte Carlo method involves repeatedly sampling the underlying probability distributions of a random variable and performing all calculations involving that random variable many times, with those sampled values. The Monte Carlo method is a brute force, random process of approximating the true, resulting distribution.
For example, if you wanted to compute the value of
Monte Carlo simulations are far more general than analytical solutions, so apply to more equations and distributions. In the future, analytical techniques may be used when possible.
Range & proportionality input mechanisms can span normal, lognormal, uniform, and beta distributions. Additionally, many more distributions are available through the function interface.
5000 samples are performed per stochastic expression. 5000 is enough to be useful for most estimates, but not enough to slow the system down. In the future, this amount may be variable depending on the need and circumstances.