Media release
4 December 2008
Accurate time predictions for website projects
University of Auckland computer scientist Associate Professor Emilia Mendes is working with web companies to help
improve their ability to accurately estimate the time required for new projects, allowing more accurate quotes to be
provided to clients.
Funded through a Marsden Fast Start grant, Dr Mendes has worked with six companies, and is currently working with
another three, to build models incorporating the important factors involved when estimating effort. The end result is a
graphical model that can be used not only to estimate people hours required on a project, but also to provide effective
decision support to project managers for trade-off analysis and risk assessment. The technique used, Bayesian Networks,
allows models to be built from a company’s existing expertise, from data on past finished Web projects (if available),
or a combination of both.
“Web companies struggle in providing accurate quotes to clients, simply because they don’t have the resources to take
time planning rather than working,” says Dr Mendes. “By developing a model that allows them to manipulate factors which
will influence a Web project, they can accurately quote in an efficient, timely way.”
A number of factors are considered vital by most Web companies, such as complexity and number of web pages. The end
results of this research project will be to build a large model aggregating all fundamental factors, relationships and
uncertainties identified by the participating companies, which can be used as a base for all Web companies, and then
adjusted to meet their specific requirements and expertise.
ends