Every company has a vault full of data waiting to be mined. This vault is getting bigger every day; fuelled by the expanding technology capability to capture more and the capacity to store more. The massive data doesn’t provide value unless you know how to distil the right information presented in a meaningful manner. Otherwise, the voluminous data doesn’t offer any value.
In order to mine intelligence from data, here are 5 tips you could use.
1. Identify the question to address
Many might find this challenging because most of the time, we have our mind set on ‘what’ to do rather than ‘why’ we need to do something. By identifying why we need the data, it becomes a lot easier to determine what data to extract. For example, a utility company wants to convert as many customers as possible to view and pay their monthly bills online. It has run several campaigns in the past. The question the company seeks to answer is what enticed the customers to convert.
2. Determine what data would be helpful
Once the question is defined, you determine what data you need to mine from your vault. For the utility company, it could start with past campaigns and assess their effectiveness. The data to extract would include the types of campaign ran and their corresponding conversion rate. It would compile data on the campaigns including the type, duration, media, cost, and the target customer groups, if any. On the conversion data, the characteristics of the customers converted by campaign would be useful.
3. Consider how the data would be used
With the data elements identified, you need to figure out how to interpret the data. Would looking at the raw data be adequate? Or do you need to compare it with a reference point for an assessment of relative significance? For each of the campaigns the utility company ran, you might want to look at the results for each campaign individually, or compare them on the speed of conversion and relative return on investment. The analyses you want to do determine how the data would be used.
4. Present information in an intuitive format
There are many ways to present the information: table, line, bar, or spider chart, etc. Each option could be more meaningful than others. For example, a simple table that summarizes the types of campaign and the respective conversion rate provides a quick snapshot of the campaign results. For more a more detailed assessment, you might want to plot the conversion rate by campaign and by duration. The conversion trend provides useful insight on the performance velocity of the campaign.
5. Enrich information with peripheral data
When the information is pertinent for decision making, you likely would want to tweak it so that you could mine additional intelligence to expand the scope of your undertaking or initiate new campaigns. The utility company might want to leverage the converted customer base to test market the use of smart metering program for energy conservation. The peripheral data, in combination with the core data extracted, helps to establish a reliable source for other decisions.
Data is a piece of fact captured at a certain time. There is no value associated with it unless you are able to turn it into useful information. However, you stand to yield big return when you are good at it.