In Snow’s development team, we put a lot of effort into creating the best reporting engine, making it easy for stakeholders from across the organization to generate management reports and examine detailed information relevant to their role, whether that be compliance, procurement, network management or business operations.
But we’re also acutely aware that even the best SAM reporting engine in the world is hampered if the source data isn’t accurate and complete. Just as with other areas of IT, it’s impossible to escape the maxim that, ‘if you put garbage in, you risk getting garbage out’. Anyone who has tried to manually check a large data set for accuracy knows how difficult it is (ever tried looking at the software report in Microsoft SCCM, for example?).
The brain easily becomes bored and wanders, and the eye jaded and tired, skips rows and misses the critical information – or errors – that you’re hunting for. IT Asset Managers are not the first to have suffered from this problem! In fact, accountants have been grappling with it for centuries, well before the age of IT, and invented double entry bookkeeping to minimise accounting errors.
It does this by creating two sets of data which should be equal – and if, after working through transactions and balancing the books the two sets of data don’t end up being equal, you know there is an error somewhere and can start working to identify and correct it. Unfortunately, unlike accountants, IT Asset Managers don’t work with a logical and self-contained system that a set of well-defined rules established and tested over centuries.
IT Asset Managers work in a messy, unpredictable environment where data is frequently not maintained, is out of date, and includes lots of errors and mistakes. However the basic principle holds – that comparing two data sources that “should” be the same (or at least, have some common properties) allows you to easily spot discrepancies – some or all of which will be caused by data errors.
Rather than spending your time going through large amounts of data hunting for a needle in a haystack, you can spend your time understanding why the discrepancies have occurred and using them to clean up the data. And of course, clean data is the holy grail for IT Asset Managers – knowing that your reports and dashboards are based on clean data solves the ‘Garbage In, Garbage Out’ problem once and for all.
Of course, finding two data sources that should be the same is easier said than done, and even when you identify them, it may be difficult to get hold of the data and it may be riddled with errors. A good Software Asset Management (SAM) solution, that combines the ability to clean and normalize inventory data, compare audit reports with other source data and manage software licenses in a central console will allow you to drill down into the detail of data sets, and perform quality checks much more easily.
It is worth persevering – for instance, automatically comparing a list of machines delivered by your audit solution with your Active Directory information can highlight where there are perhaps holes in the network discovery, or where the estate has undergone change (e.g. machines have been retired, but software has not been re-harvested). Below is a list of Dos and Don’ts to make the process of data cleansing as easy as possible.
DO sell the benefits of a clean data set – for instance, most Active Directory structures are a mess, and no one wants to either clean them up or maintain them. However if the business is planning to upgrade Exchange or move to O365 a clean AD is a must – and comparing YOUR discovered data with THEIR AD data helps the infrastructure teams keep AD up to date as much as it helps you identify machines with a broken agent.
DO work within existing processes – most support teams are used to having their workload managed through the IT Service Management system. Log service requests if you need the support teams to do something for you rather than sending an email which will get lost or forgotten. This also lets them prioritize your requests appropriately within their broader workload.
DON’T treat data cleansing as a one-off exercise – schedule cleansing activities on a regular basis throughout the month or quarter – that way you know your data is no more than a month or a quarter out of date, and you’re also not trying to do everything at once, overloading both you and the IT teams with data related service requests.
DO treat data cleansing as an iterative process – take baby steps, and don’t worry if you can’t resolve all the discrepancies in one go – allow yourself to say ‘enough is enough’, leave it for this month or quarter and move onto the next scheduled cleansing task.
You can pick it up again the following month with a fresh set of reports… and you never know, the problem may magically have disappeared in the intervening period! And finally… Process, process, process… as you clean your data, try and understand why it became inaccurate in the first place. As you understand the underlying cause, work with the different teams involved to improve their processes and gradually reduce the number of errors that creep in in the first place.
Gradually, over time you will see the number of data errors decreasing, and you will have the satisfaction of knowing that ‘Garbage In… Garbage Out…‘ is no longer a problem.
To learn more about how Snow can help you avoid ‘Garbage in, Garbage out’, speak to one of our SAM experts today.