IS IS 788 [Process] Change Management Lecture: Process management – goal setting and BAM Discussion: Western Digital case
IS Measuring and then managing business processes “You can’t manage what you don’t measure.” (anonymous) Management of process includes: Setting goals Assigning tasks Monitoring results Improving processes Solving process problems BAM figures prominently in these two
IS Management is essential In some cases apparent “problems” with processes can be fixed simply by modeling and then beginning to monitor (manage) the activities Left unmanaged, even the best processes deteriorate. Management is key to a successful process implementation.
IS What does it mean to manage (a process) Establish realistic output standards for the process Determine the process activities and its performers (what to do and who does it) Monitor the outputs Take corrective action when necessary
IS Process oriented management
IS In a process organization Managers are responsible for understanding processes even if they inherit them This is actually a shift from 80’s ‘Harvard’ style management practice where all a manger was expected to know were the financial ‘numbers’. Managers are also responsible for understanding the broader value chain into which their process fits
IS In a process organization (2) Goal alignment starts at the top Even though the stress is on process functional or departmental units will be kept in place. And so, managing process organizations is more difficult as our readings have shown.
IS Linking process management to functional centers
IS Goals, Measures and Monitoring Goals – what a company wants to achieve Measures – data gathered and interpreted to gauge achievement of goals Functional measures – traditional; financial; ‘the numbers’ Process measures – general measures of customer satisfaction with process outputs
IS Functional and process measures both overlap and conflict Slavish adherence to functional measures – all some managers know or have been exposed to – can severely damage an organization. Long before process thinking was in vogue, manufacturing (process) engineers were in more or less continual conflict with “bean counters”
IS Common conflicts from ill conceived pursuit of ‘good numbers’ In general any functional measure that disrupts process speed or efficiency to maximize a departmental goal – which is pursued because it is rewarded! Reduce inventory costs but deliveries are late Use of cheaper employees who do shoddy work Redesign the product for less expensive manufacturing, but lower quality Taking orders for non-standard products to boost (temporarily) sales figures The ‘Soviet Nail Factory’ as an example
IS In process thinking Process goals (which are directly related to customer satisfaction) almost always take precedence over functional goals. One (maybe the only) way to assure this is to have process sponsors at a very high organizational level
IS Goal and measurement hierarchies As goals are delegated down the hierarchy (from strategy to tactics) they become narrower and measures become more specific However, Harmon advises against modeling or measuring a process below the level at which it is clear that goals are being achieved. This level is frequently lacking in the detail needed for IT implementation. We’ve seen this before – process modeling and IT systems analysis are different activities with different objectives!
IS A measurement hierarchy example
IS Process measures worksheet
IS Balanced scorecard approach Developed at Harvard in the 1990’s and has a very strong following, especially in the management accounting area Discourages undue emphasis on financial measures. Uses key performance indicators (KPI’s) to measure Financial information Customer satisfaction Internal business processes Innovation and learning (org learning)
IS Business Activity Monitoring (BAM) Given the appropriate infrastructure Successful ERP installation Successful in-house systems integration It becomes possible to implement automated BAM and manage by exception Think ‘dashboards’ on process managers desks, delivering the current status of all key process measures in real-time.
IS BAM (from webMethods BAM whitepaper – URL on syllabus) For workflowed processes such as order processing or insurance approval, BAM can alert managers to deviations from policy at the transaction level. And – BAM can be set up to search large quantities of data for suspicious patterns, i.e. credit card companies
IS Monitor, then alert, preferably in real time Insert webmethods figure 1
IS In order of sophistication 1.Intercept individual ‘out of spec’ transactions (workflow) 1.Discuss the Cisco conversation monitoring appliance 2.Monitor transaction average parameters (rolling or periodic) 3.Data-mining and NLP techniques for analyzing huge masses of numeric or textual data for negative patterns
IS Individual transactions Monitor each transaction for violations of business rules Price of goods Margins Size of repair (may require extra authorization) Incomplete routing
IS Average monitoring Requires (obviously) sophisticated database infrastructure. Functional data stores are useful to allow analysis without slowing transaction processing Determine key measures # of transactions # of changes # of calls Time to event (too long to ship, etc)
IS Large volume pattern matching Requires off-line data storage and large amounts of processing power ‘Fingerprinting’ or ‘coordinated event publishing’ Monitor a key process indicator When it falls below specification, publish all the data that was current at the time of the KPI anomaly for high- level analysis
IS Fingerprinting illustrated