Article

Improve marketing strategy by adding free or alternative data to your dataset

As a marketing professional who uses data and analytics in your planning and decision-making, an obvious place to start would be a review and analysis of your own data. But, what if you want to add more perspective to those internal analytic sources?

One particular area for this would be free data, often called “alternative data,” by analytics experts, which are datasets that are publicly available for use. By integrating data from outside sources, you can start to understand how external factors are impacting the business, what sorts of investment decisions make sense and so on. More data leads to more context and by extension, more insights.

Examples of free or alternative data

  1. The Census Bureau – Of course, they conduct the census, but they collect a lot of other data, too, about communities, who lives where, what their interests are and how the population is changing. All of that information is publicly available in various cross sections.
  2. Bureau of Labor Statistics – They have a consumer expenditure survey that is loaded with data about how consumers are spending money year over year and how that changes over time. It’s very specific and geographically aligned.
  3. NOAA – Weather data is freely available from weather stations around the country and world.

So, with sources like these, you can capture population counts, characteristics of that population –demographic information, income, debt and more. These pieces can provide a lot of variables that you can then start to overlay with your own proprietary information and sales data, and start to derive insights and correlations between factors.

Where can your next marketing dollars best be spent?

Take population information, for example. That’s a metric that’s really easy to get but, rarely used. But think, when you’re looking at sales data geographically, you could normalize for population–if you look at flat sales data, your biggest states would be California, Texas and New York. But if you take into account how many people live in the vicinity of a store, that can help predict sales frequency to be more accurate and actionable. You can start to ask questions–do you have room to grow and add stores in one area of the state over another? Or, maybe you identify an area of the country where stores are underperforming and not doing as well as you’d hoped.

Where to find alternative data

While it’s simple, technically, to access and get free data, such as from the Census Bureau, there’s more to it than just downloading one of several files. In addition to locating the data and downloading it, oftentimes large sets of data are not clean, and are not well structured. This is not like a pristine Excel document in many cases. It’s free, yes, but you have to do some work to maximize its value in your own system.

What to watch out for

  1. It can take time to decode datasets. Often, data statistics from government sources are coded, so there is an added step of finding the coding document that’s relevant to the data you want, then using one to inform the other.
  2. What do data field descriptions really mean? Take, for example, a census data set–these often have long, confusing descriptions. It takes time to decipher just what some of these fields are and what they really mean.
  3. Data that’s separate, but should be together. Again, using Census data as an example, this is often provided as multiple downloads that work better together than apart. There’s time required to merge these fields and make sure they all remain accurate against one another.

Alternative data is not a new concept, and has been used for years by financial firms and other industries. But, it’s something that’s slightly newer to marketing and business analytics, and it’s been proving its value alongside more traditional sales data and other proprietary sources for clients and companies. With that added data-informed perspective can come insights and options that may not have otherwise been revealed.

However, be careful. It’s tempting to go out and do a general search for whatever data you’re looking for, but they’re not always the most accurate or reliable sets of data. Be sure to validate your sources and verify the accuracy.