Lies, damned lies and statistics…
Those of you with a penchant for English political history will know that it was (supposedly) British Prime Minister Benjamin Disraeli who described three kinds of lie in the 1800s: “Lies, Damned Lies and Statistics.”
In some ways, it seems as if not much has changed in the intervening 200 years. We are still won over by big figures, often if they even contradict our own experiences. Unfortunately, software management is no exception to this rule. Here, statistics often relate to the size of a tools vendor’s software dictionary. But all is not perhaps as it first appears.
Undoubtedly, the software dictionary is a critical element of the license management solution. In License Dashboard’s License Manager product, the software dictionary is what enables both software audit information and license information to be imported into the License Manager repository in a consistent and accurate fashion.
When importing software audit information, for example, the software dictionary in License Manager is able to take feeds from over 30 different inventory solutions and process the data against a dictionary of over 9,800 vendors and more than 350,000 EXE and MSI signatures. This enables License Manager to automatically (and transparently) do things like normalize software vendor names – you don’t want a repository where Microsoft, Microsoft Inc and Mircosoft [sic] – are all listed as separate vendors, rationalize version numbers – do you really want to know how many versions of Adobe Acrobat 184.108.40.206 you have versus 220.127.116.11?, and correctly identify product names – Microsoft Office 11 anyone?
The dictionary plays a central role in ensuring that what you see on screen in your license management solution makes sense and is directly relevant to the task at hand – software licensing (which is not the same thing as software auditing – that’s why inventory tools and license management tools are integrated but separate).
But in order to win the statistics battle, some License Management tools vendors have become guilty of justifying Disraeli’s scorn; they’ve started to search for the biggest numbers they can find. So instead of simply stating how many unique products are recognized by the software dictionary, some vendors are now talking about ‘records processed’ or ‘transactions’. Suddenly, a vendor with 100 customers can claim 3 million records processed per day (do the math; 100 customers each with 600 PCs on their network and an average of 50 applications on each PC = 3 million records – assuming they really are doing a daily audit!).
Three million sounds a lot more impressive than 350,000 doesn’t it? But does it actually mean the customer gets a better quality of software recognition? No. First, it doesn’t take into account how many of the 50 applications are actually licensable (think about how many apps you have on your PC that are part of the OS package, or a free add-on to another software title). Second, it doesn’t indicate whether any intelligence is being applied to the recognition – has the solution, for example realized that Word, Excel and Outlook are in fact an Office Suite, or that because a PC has both Office 2003 and Office 2007 installed on it, it only needs to consume one license?
Software recognition is about quality just as much as quantity.
The more important questions to ask are: Is the software dictionary included in the license management tool, or is it an added-cost service? Is the software recognition done locally, or does my data have to be sent off-site? Is the recognition instant, or is there a delay in receiving the results? How easy is it to modify the dictionary to recognize bespoke applications? How often is the dictionary updated and how many updates am I entitled to?
By asking these challenging questions, rather than simply believing the figures touted by some tools vendors, you are more likely to find the solution that best suits your actual needs. After all, there’s much more to license management than lies, damned lies and statistics…Tags: Audits, Microsoft, SAM