Many companies today are interested in using data to improve their business and customer experience.
Most of these initiatives ultimately leave a human in the loop, making decisions and implementing tactical or strategic business changes as a result of the reports, dashboards and analytics they are armed with.
Having this human in the loop brings the usual problems associated with manual processes, such as delays, errors, subjective judgements, and cost.
There is huge potential to take a subset of these business intelligence and analytics use cases, identify the situations of interest automatically, and automate the interventions as they happen in real time.
This involves:
- Listening to or ingesting events as they happen;
- Using an automated “decision maker” algorithm, based on either simple business rules of something more complex such as a statistical or machine learning model to decide on the steps to take;
- Connecting back to a system to implement the recommended change via an API or other interface;
Each time we do this, we remove a human from the loop, avoid the problems above and deliver benefits such as a more efficient business and improved customer experience.
Three examples of this in action might be:
- A product is experiencing a reduction in sales, so we decrease the price by 1% and raise the response in search rankings;
- A manufacturing machine is showing increased heat and vibration, so close it down and queue it for maintenance;
- A call center agent is taking longer than over agents to resolve calls of a certain category, so re-route those particular calls away from him temporarily;
We refer to this as “closed loop” analytics. It is by no means as easy as dumping the results onto a report or dashboard and hoping the findings are actioned, but it is much more powerful and much more likely to have a compelling business case.