> For the complete documentation index, see [llms.txt](https://isubasinghe.gitbook.io/isithas-wiki/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://isubasinghe.gitbook.io/isithas-wiki/software_engineering/management/cost_estimation/challenges.md).

# Challenges in cost estimation.

## Basic idea

There is no exact science for cost estimation — estimates are wrong by construction. The best you can do is improve confidence with historical data, decomposition, and empirical models.

## Key facts

* Estimates are guesses; calibrate with past projects.
* Decompose: small parts are easier to estimate than the whole.
* Defer estimation if you can — accuracy ⇧ as the project progresses.
* There is no exact science for cost estimation - it will never be considered as all accurate.
* No person can reasonably predict what can go wrong in the project.
* Most estimation methods assume things will proceed as expected and simply adds some slack to account for what can go wrong.

## Possible solutions

* Delay estimation - 100% accuracy but less usefull
* Base estimation on data from previous projects that have been completed.
* Break the system to smaller parts and generate the estimates for smaller parts, which is easier.
* Use emperically-based estimation methods.