4 ms·
You're talking as if better estimation were an end in itself. It's not. The goal is a successful business. There's no point in spending any more effort on es
by ScottBurson 10y ago
You're talking as if better estimation were an end in itself. It's not. The goal is a successful business. There's no point in spending any more effort on estimation than is needed to support the planning required to make the business a success; any more effort than that is simply wasted.
If the expected ROI on a project is small, then having really good estimates matters a lot, because it can make the difference between profit and loss. But if the expected ROI is large, precise estimates matter less. Tom DeMarco has famously made this point [0].
[0] http://www.systemsguild.com/pdfs/DeMarcoNov2011.pdf http://www.systemsguild.com/pdfs/DeMarcoNov2011.pdf
- jacques_chester 10y ago> There's no point in spending any more effort on estimation than is needed to support the planning required to make the business a success; any more effort than that is simply wasted. This is basically what I argued. Software businesses often have such fantastic margins that you can do any old thing and still rake it in.