Step 3: Normalization
Enterprises often have multiple discovery and inventory solutions. Normalization is the consolidation of discovered inventory datasets to remove duplicated or conflicting information.
Normalization - The questions you need to ask yourself
- How do you normalize your inventory data today? What inventory data is included?
- Do you maintain your own catalog? How is it updated?
- To what level is data normalized (publishers, title, edition, version, release)
- What other tools (i.e. ITSM, CMDB) do you populate with normalized data?
- What processes do you have which are fed by normalized data?
Data can be extracted from many different sources which means it will not be consistent. The process of Normalization presents it in a friendly and easily recognizable format. It removes duplicates to present just one source of truth about each software asset.
The primary benefit of a normalization process is an accurate organized inventory across different datasets. For example: One dataset recorded an executable file name as ‘application.exe’, whereas the other dataset recorded it as ‘application’. Even though it is only one licensed application, it has two unique descriptions, but should not be counted as two separate installations.
Many approaches to normalization classify and categorize inventory automatically using databases of vendors, product and service names to standardize naming conventions of discovered inventory. Another good example of normalization is bundling identification. The inventory process might have discovered multiple applications including Microsoft Word, Microsoft Outlook and Microsoft PowerPoint installed on a Laptop. It determined that these are not three independent licensable applications, but instead a single suite.
Some normalization solutions may also provide default metric information, group applications into application families, add upgrade and downgrade paths, release dates and end of service information.
The Snow Way
Snow Software addresses these needs with its offering of the Data Intelligence Service (DIS), formerly known as the Software Recognition Service (SRS). It is available as a subscription service and provides our customers with up-to-date information for more than 500.000 applications from 80.000 vendors. If you would like to know more about our DIS/SRS, visit this blog post.