Data Encryption with business intelligence and reports has always been a problem. Users are constantly requesting reports, and once data is in someone’s hands, it’s almost impossible to control data dissemination and what I’ll call data diaspora. One must admit, especially in large organizations, that trying to put controls from a procedural perspective is not particularly realistic. With hundreds or thousands of managers out there, controlling the actions of each person is particularly difficult.
In the good old days of ad hoc report files, excel spreadsheets, and powerpoints, any person who got their hands on data could easily forward it to someone else. The fact is that technology was sufficiently difficult to use that most organizations, even the very large ones, have used Excel as the easiest way to aggregate and analyze data from multiple sources. Even for single source reports, excel has long been the easiest way to communicate a data set. Managers didn’t really have robust capabilities to tap into reports on their own, and even then, one of the selling points from software vendors has been the ability to export data into excel where managers or practitioners could continue analysis.
HR technologists have been talking about dashboards and business intelligence for years, but it does seem that the lately emergent technologies are finding some adoption in larger organizations. Perhaps this is just maturity of the technology, perhaps the prices have started coming down from the fully customized ERP BI software to more vanilla and off the shelf analytics tools, or perhaps it’s possible that spending was just down so far in the last 2 years that nobody was buying the stuff. Whatever the reasons, the technology and the market seems to be ready now.
Certainly, increased controls are now much more prevalent with each manager going to their own dashboard to view data, and with the large number of analytics available in the HR and talent realms, ad hoc requests are hopefully going down. All this just means that if you can deliver a set of analytics to the manager desktop as opposed to frequent ad hoc requests, your data is controlled by the application security layer upon delivery. Since you have never sent an email with an excel spreadsheet, there is no data to be fowarded.
You’ll argue with me that this technology has been around for years upon years – at least a decade. I’ll absolutely agree that this is true, but I’m pretty sure that every single vendor out there (whether publicly or not) will agree with me that until recently the delivered reports were not sufficiently robust or comprehensive. ERP vendors are now also delivering robust prebuilt analytics with sufficient drill downs and drill throughs. The goal of the whole thing is to have enough data presented in a simple but detailed enough manner to eliminate most ad hoc needs. If you can create an environment that does this, you utilize your application’s security as opposed to releasing your data to the winds of fate.
3 responses to “Business Intelligence and Data Dispersion”
I agree that the new technology and self service does provide a new degree of control of who gets to see data though I’d argue that good policies with effective communication are a better way of providing control whilst opening up the data to be used in effective ways as needed. It’s rather like rules on not using social technology at work – it’s better to have it open but communicate what is allowed than to block it forcing people to find work-arounds. What the technology does offer is a way of avoiding the version-control issue.
There is an overlying issue with most implementations of information self-service and that is whilst it can provide data too often this is used instead of insightful information that can be acted on presented in a manner that highlights where focus is needed (Stephen Few’s book ‘Information Dashboard Design’ covers this well).
Furthermore I don’t believe that the provision of reporting tools can totally replace good analysts. There are two issues I believe. The first is that whilst many of the tools can handle reporting well, they are often poor at enabling meaning to be inferred. There are a bunch of data exploration tools out there – things such as Tableau – that are a step forward than the traditional BI toolsets.
Secondly whilst it is important to have good analytical skills and the appropriate tools it takes an analyst who truly understands the business to add meaning, especially when the data is used in more of a Bayesian manner. I therefore feel the optimal way of providing information for action is twofold:
1) some self-service dashboards* but designed to facilitate clear understanding of actionable information
2) a central analytics team who can master the available data, see and commision what is missing and provide management with meaning / stories based on the available information. This team needs to easily available for all managers, not just reserved for senior HR and available to consult on what the data can show.
It is clear that enabling extraction and excel analysis is only suitable for a tiny subset of the population who do have resilient statistical understanding and the ability to interpret what the data means given their deep understanding of the business.
* Given that the BI tools shouldn’t replace good analysts but cover the ongoing basic reporting the business case for their adoption isn’t necessarily always there. HR data doesn’t change that much (unlike sales transaction data for example) with most employees having a handful of changes per year. Therefore the benefit of using live data is diminished for many applications. In many instances automated report production on a regular basis programmed by a central team may cover the information need more efficiently.
[…] Systematic HR – “Dubs” […]
[…] HR JUNE 14, 2010 Business Intelligence and Data Dispersion HR technologists have been talking about dashboards and business intelligence for years, but it […]