I have long been a fan of Richard Hackathorn’s view of decision latency and use it regularly in both consulting and training. He uses the concept of the value-time curve. This curve is designed to show that the longer you take to respond to new data, business events, the less value there is in your response (assuming you make an equally good decision).
Dr Hackathorn describes response latency as consisting of the sum of :

- Capture latency
How long, essentially, does it take to “notice” the new data or event - Analysis latency
How long does it take to create information and insight from this data - Decision latency
How long to act on this insight – really two parts, deciding and then acting on the decision.
As he says “As quicker actions are taken, we move up the value-time curve, increasing the value gained.” By compressing the time to make the data available (capture), analyze it and decide what to do we increase the value of our decisions. While one can focus on the time to capture data and make it available for analysis we generally focus clients on reducing the time to analyze the data and act on that analysis. By developing data mining and predictive analytic models we can essentially automate the analysis as new data can be scored in real-time as soon as it is available using in-database analytic deployment or scoring embedded in decision engines. By using business rules and Decision Management Systems to automate decisions, we can reduce the decision latency essentially to zero.
One other thing we have found as we do decision discovery for clients is that different decisions have different curves. Some decisions, such as an online cross-sell, have a very steep curve as any delay degrades the value of the decision. Others decline very slowly until some policy or legal limit after which they have essentially no value. Different curves, different shapes, different consequences. When deciding what kind of Decision Management System to build and how to use it an understanding of the decision latency, of that decisions specific decay curve, is a really useful asset.
Comments on this entry are closed.
Great visualization!
You mentioned different curves and different shapes, that makes sense, but isn’t the only _relevant_ point the “action taken” point – since that’s when value is realized?
I mean, it seems like you could have slow-capture, instant-analysis, instant-action, or instant-capture, instant-analysis, slow action (or whatever combination) – creating “different shaped curves” – but they would have the same value, if the delay-to-action is the same.
That’s definitely based on the assumption that value is only realized through action. Is that a valid assumption?
Also – is the exponential decay curve shape realistic (in your experience) or just illustrative?
TIA
Scott
Well sure, but the other points are worth tracking as they help you see where else you could reduce the time taken and thus the overall responsiveness.