It depends a lot on the data you have available in SSAS (and other accessible sources) but some things that come to mind:
- classification/decision tresss-> using something like total project revenue, what are the key influencers to this being larger (or smaller)
- classification/decision tress -> using a subbie gross profit of subcontractor billed rate less subcontractor cost, what attributes (e.g subbie attributes, project attributes etc) will contribute to a profitable project outcome
- Market basket analysis (association) -> what services (or products) we we often sell together - is there an opportunity to 'package' services in a way that we upsell or increase profit because we know that the client is very likely to purchase this 'basket'?
- segmentation -> what are the 'groups' that are our clients? do we market to the identified segments, or is this an opportunity to market differently?
If you want to get a 'toe in the water' start with DM in SSAS, i would definitely download the Excel addin for Data Mining (office 2007). When you create/import a table of data into Excel this becomes active and provides a nice wizard interface to perform all of the approaches above as well as few more.