With traditional eDiscovery processing services, you essentially have to cross your fingers and hope to get the smallest, most relevant set of culled data from raw collections. The defac to standard for doing this has long been keyword and basic attribute culling.

The problem is that the primary purpose of behind-the-scenes keyword culling is to find what is possibly relevant ... not what is truly relevant. Big difference. And as such, false positive results will leave you with large volumes of irrelevant/non-responsive data. And the more useless data you have, the more expensive the overall discovery and review processes are.

• Behind-the-scenes culling by vendor means little control of data
• Large volumes of irrelevant data filters into review
• More irrelevant data means more time ... and more money