There can be many parts that make up an operator’s provisioning factory. There can also be multiple different software packages that contribute towards getting an order through the factory. In addition to that, there are usually many different activities that need to take place, some automatic, some manual, some as external dependencies, conditional gates, etc. That’s why operators design process maps and why most OSS have a workflow component to manage all the parts of the factory. It’s also why TM Forum’s eTOM model has been widely used.
I’ve seen examples where all the groups that contribute to the factory claim great performance (and can produce the metrics to prove it), but barely a trickle of orders are getting activated and the order to activation (O2A) cycle can be measured in months rather than days.
Part of the problem for these organisations has been that they have been unable to track flows through the factory. If orders get caught up in eddy-flows, metrics are met but they go around in circles and nothing gets out the end.
Sankey diagrams offer the potential to visualise end-to-end flows, through multiple systems, perhaps even in real-time.
They’re relatively easy to prepare data for – just create data that has source and target identifier fields and a value / metric for tracking between them (eg number of orders per minute/hour). Hint: you may only want to show your level2/3 process flows, not deeper flows as the diagram might get too messy.
The result might look something like this (courtesy of sankey-diagrams.com):
Note: We’ve now developed a technique that uses Sankey Diagrams to document, benchmark and optimise operational processes directly from activity log files.