It may seem obvious, but a central aspect of these technologies is to try to attach appropriate statistics to something you see by eye in the microscope image. As much as this statement essentially brings all of cell biology under the umbrella of HCS, the reality in the field can be a quite a bit different. A big pharma drug discovery group may be interested in a toxicity profile of 200,000 potential drug candidates whereas an academic researcher may want to accurately quantify the cell-by-cell behavior of lipid droplets under varying conditions & time points, 5 images at a time. The speed, flexibility, accuracy and costs needed to set up these two scenarios would vary wildly.
Along the same lines, it’s important to consider how open, flexible or dedicated you need the systems to be over time. As we are all intimately aware, microscopy images for biology tend to be highly variable. All the vendors have analysis packages and “validated” assays. These are great if you perform that exact assay. The challenge comes as you want to do something 10% different than the standard or want to compare results from different systems. It takes a certain amount of knowledge and the right kind of flexible tools to make the required adjustments.