4 Tips to Maximize Your Data

As marketing and demand generation strategies become more and more data-driven, executives and analysts alike need to explore methods to maximize the value from the data they already have. Data gets a bad name in the industry, often times carrying the stigma that developing a data strategy or conducting data analysis are difficult. They don’t have to be. Here are four tips to maximizing data that are applicable to companies of all levels of data sophistication – whether they already have been meticulously tracking their marketing and sales activities for years, or they’re just now establishing their data strategies.

1. When in Doubt: Collect More Data-
We live in a world where one can pay pennies per gigabyte of online storage – hard drive size is simply no longer an issue anymore. Given that development, big data analytics has now become a field focused on taking massive databases and turning them into actionable insights. You can always track something and decide to cut it out of your sample dataset later on. You can’t choose to not track something then decide at a later date you actually needed it.

Recommended for You Webcast, September 7th: 3 Ways to Speed Up Your Marketing Review and Approval Process

2. More Data Doesn’t Always Mean Good Data-
Now this is the major caveat to collecting more data: make sure the data you’re collecting is actually good data. This includes making sure collected data is consistent, the data has integrity, and that the rules for the data collection are clearly defined. If not, then there’s no way to know if the analyses being done on the database are at all accurate. For example, even if one person on a data team knows that a certain data field lacks integrity doesn’t mean somebody else could accidentally use that field for an analysis. In these cases, it is better to recognize which data fields just don’t work and either fix them or remove them from the database.

3. Make an Elementary Deduction, My Dear Watson-
The beautiful thing about big data is that it’s possible to deduce new things from datasets based on existing data. This “proxy” data can be incredibly useful and allows for deeper insights and analyses. For example, if a database contains a lead’s assigned sales person, but not their geographical location or their company size segment, one could deduce these two pieces of information based on the rules for which leads get routed to which salesperson. For example, if The Worldwide Widget Company has John Smith as their lead owner and John Smith only handles EMEA Leads with revenue greater than $250M, then one can (usually) safely assume that the Worldwide Widget Company is based in EMEA with a revenue greater than $250M annually.

4. Correlate Disparate Data-
Many companies have heaps and heaps of data sitting in many different databases, spreadsheets, documents, and perhaps even on the backs of cocktail napkins. Problems arise when behavioral data is kept in a SQL database, but revenue outcomes are tracked in a separate Excel spreadsheet. It’s difficult to establish what behaviors caused what revenue outcomes until the data is correlated. This is done using a “key” field, or a piece of data you can use to match data from one database to another. Commonly, this is a unique identifier value such as an e-mail address or Salesforce Account ID.

While this sounds relatively simple, one of the value judgments that needs to be made when joining datasets is how to handle one record correlating to multiple records in another database. In the example above, if there is one behavioral record matching multiple revenue outcomes, should the master database have a separate record for each revenue outcome with identical behavior information? Or should the master database contain one record containing the behavioral information and the sum of all revenue outcomes? Often, this will come down to the types of questions the data is trying to answer, so it is important for data storage to be flexible to allow for multiple kinds of joins.

These tips should get any organization started down the path of maximizing current data in order to get the most out of an existing data strategy. For those companies hungry for more and really want to take their data and analytics to the next level, check out my previous post on the advantages of adopting the statistical programming language R into one’s technology stack. It sounds much more difficult than it actually is.

Let's block ads! (Why?)

4 Tips to Maximize Your Data