When evaluating data classification tools, you need to ask the following questions:
What file types must the product support and will it deal with structured and unstructured data? Some products are limited in the file types they'll handle, while others claim to work with a large number of them.
What rule sets and automation will the product support? All data classification products use rule sets that drive the classification engine. These can be generated in-house or imported to support industry standards. If you'll be using the latter, make sure the product allows imported rule sets to be modified to your needs.
Does the product support tiered storage and migration? If you haven't classified data and matched its value to the appropriate performance storage media, it might be costing you more than it should to keep it or it's inappropriately protected. If you want to implement tiered storage -- something everyone should consider -- make sure the data classification product can migrate data between tiers so it's matched to its correct service level.
What kind of performance and scalability does it offer? Your organisation may possess millions or billions of files. Data classification products often have limits on the number of files they can support, so be certain the products you evaluate can deal with the volume of files you have -- or will have based on expected growth rates -- at an acceptable performance level. When testing, keep a close eye on performance figures during the classification or search process by measuring files per hour or GB per hour. In addition, you should determine whether the product is better at handling large numbers of small files or smaller numbers of large files.
What degree of heterogeneity will the product support? The data classification tool you select must interface with a number of platforms in your environment and issues may arise during that process. Testing ensures a data classification tool will work with another policy manager or data mover.
Does the product have hold capability? If your organisation needs to locate specific data for litigation purposes, you may need a data classification tool with litigation-hold or file-hold support. This enables data to be frozen to prevent changes or deletion.
Is the tool implemented as hardware or software? Data classification tools may be implemented as software installed on one or more servers in the enterprise. You might need multiple servers to provide the performance required in larger organisations managing billions of files or if attempting high rates of classification. Tools may also be implemented as hardware appliances, which are essentially dedicated servers running data classification software.