`Cross-posted to ebizQ
Bill Gassman presented on adding business intelligence about and within your business processes. Bill’s emphasis is on the value of bring Business Intelligence, Business Activity Monitoring and Business Process Management together as well as the challenges of doing so. Bill is using the Large Hadron Collider as an example and talking about the risk of smashing things together and creating a blackhole. So, why are these technologies converging, how can BI technologies improve processes and what are the best tactics?
BI adds value, Bill says, in two ways.
- Firstly BI within a process adds value by improving decisions through what I call decision management, using business rules and scoring/analytics, as well as through operational BI to give those executing a process context-specific analysis. BI (or more accurately analytics) can also detect the need for additional steps (fraud review) or the opportunity to remove steps to fastrack claims for example and reduce costs/time.
- Secondly BI adds value when applied about the process. One can link the results of a process to business goals, increase awareness through status and alerts and ultimately help optimize the overall process.
Bill’s examples include Netflix using analytics to predict what you might like, multivariate testing to see what works (adaptive control in my vernacular or a/b testing), real-time risk analysis using predictive scoring, supply chain optimization and the famous Harrah’s example to use analytics to improve customer experience.
Business Intelligence (certainly a definition of BI that includes business analytics) and Business Process Management are complementary. Using Business Activity Monitoring to seek new patterns, BI to do analysis that drives a model that can be implemented using business rules management or decision management to adapt business process management that creates new results that are considered by BAM and around we go. This view is, frankly, a much more practical view of the pattern based strategy stuff than anyone else has presented!
Gartner sees a BPM maturity continuum as well as a BI one. Bill presented them overlaid and clearly the increasingly sophisticated used of BI drives / complements BPM maturity. The BI model goes from measure to analyze, decide, align, optimize, discover, innovate. BPM goes from acknowledge inefficiencies (measure), to process aware (analyze), automation and control (decide) and so on.
The framework for adopting BI and BPM support a pattern-based strategy, indeed they turn the vagueness of pattern-based strategies into something much more concrete. The framework involves an overarching strategy and metrics framework wrapped around people, process and technology for managing data and doing BI. Bill’s framework, sadly, lacked the kind of high-end business analytics that business processes need to execute analytic decisions and was focused much more on reporting and dashboards. But the general point that a formalized approach to thinking through metadata, analytic applications, business processes and more is needed is a good one.
Need awareness, flexibility, adaptability and productivity. A combination of BI (helping with awareness and adaptability) with BPM (flexibility and productivity) delivers agility – the ability to sense environmental change and respond efficiently and effectively. This is true, assuming you include analytics inside BI and rules/decision management inside BPM.
To this Bill argues you need new skills. Skills around information delivery (reports, dashboards, document integration etc), analysis (statistics, forecasting, data mining, visualization) and process-centricity (alerts, simulation, rules, event processing, process mining, SOA).
Bill’s formula for doing this:
Analytics services + business software services + BPM orchestration + SOA = Composite processes with embedded BI
Of course I would talk about decision services not analytic services and I would disagree that this is not ready yet – wrap the analytics with rules and those can be embedded in processes. The BI-like analytic applications and OLAP are not ready for this version, for sure, but the operational side is doing better. Repository and metadata integration is also weak, as he says. Of course the drive to vendor consolidation is also increasing the process focus of BI vendors and vice-versa. IBM, Oracle, SAP, TIBCO etc have process-centric and BI-centric elements of their stack.
From a tactics perspective:
- Ensure data quality within and at the edges of processes
Yes, though I would add that the quality needs to reflect the planned usage of the data – begin with the decision in mind
- Establish a metrics framework
Absolutely – have to be able to link process and decision improvements to the metrics that people are tracking and being held to
- Align your competency centers
True but I wonder if you would be better having competency centers focused on business outcomes – decision competency, compliance competency – not on technology
- Define a convergence architecture
Focus on a stack vendor, do best of breed with your own integration, pick a process-centric BI platform etc. Just make sure it includes business rules and real analytics, not just BI.
- Address the challenges of managing process improvement
Especially make sure that people’s compensation/rewards reflect what the new approach will drive.
Your plan should include identifying differentiating processes where BI/analytics might make a difference. You need to find some quick wins and define IT’s role in making it happen. And I would add, think about decisions so you can find the leverage points for all this.