Monday, January 9, 2012

Metrics, Analytics, And Common Sense


A tenet of Leadership is if you don't measure you won't change.

There's plenty of software and tools available to measure just about anything and provide statistics as well.

Some items can be counted – like responses to a survey questions. Other items can be measured by the responses – click-through sales. Sometimes an indirect approach is required – effect of a sales campaign by measuring gross sales before, during , and after the campaign.

There are some situations which are not quantifiable and effectiveness must be inferred – like the effect of print advertisements on sales.

A key to useful metrics and analytics is identifying what you want to learn and how best to measure it - by direct or relative measurement.

Trends are direct measurement candidates – here's my baseline of x-units over y-timeframe ...the measure is the change in volume of units for each time block.

Effectiveness can be measured on a relative basis – here's the volume and outcome of the target compared to the volume and outcome of other comparator producers or the norms of the group – like traffic on a website related to on-line orders.

Since the computer can be tasked with the slicing and dicing unlimited amounts of data, we can get hundreds of statistics from even simple measurements. When we had to pay for to compile data, the amount of statistics, ratios and comparative information requested was more focused than than a click of a key availability of today – a practical limitation of time and cost.

We can be overrun by statistics and the analytics to interpret them. The time investment now is in the use of the output. And we can be mislead by results from indirect and relative measurements if the comparator group is not appropriate to the target.

For example, traffic on a website can be mistakenly shown as aggressively above the norm or horribly behind it, if a static information site and a promoted sales-based site are compared. Apples and oranges.

Keep metrics and analytics simple to get meaningful results which can translate into changes that affect revenue.

I have run an organization using about eight key indicators available on a real-time basis. The system had capabilities of generating over 200 elements of real data and ratios – the other 192 were of no value to me – so I did not request them.

Best story about meaningful analytics was a comment from Dick Davies after we had attended a meet & greet event – he said he had spoken with 35 individuals. I asked how he knew – he said 'I counted' (resource sheets at the beginning of the event minus sheets at the end). Simple. Effective. Practical.

What are the best (and most effective) metrics and analytics you have seen?


No comments:

Post a Comment