fbpx

Marketing analytics are becoming increasingly sophisticated and effective.

To improve the insights and results generated from analytics, many analysts are incorporating “dimensions” into their campaigns.

What “Dimensions” Are and How to Use Them

Though dimensions may sound like a fancy term, it’s really just a new way of looking at the same data.

Think of data as a cross-section of an image. For example, you are examining your customer base in terms of geography, interests, demographics, purchase history, and so forth. Advances in analytics software and data processing make it possible to compile and analyze several dimensions at once, giving a multi-dimensional view of a customer population.

With the advent of digital marketing, analytics have grown in complexity. Instead of collecting and viewing pre-aggregated data that gives a one-sided view of a customer base, dimensions allow marketers to develop highly personalized, targetable marketing models.

And in an omnichannel, non-linear world, dimensions are becoming necessary in order to create strategies that accommodate the technological environment.

Dimensions vs. Hierarchies

Most analytics relies on hierarchies to represent data. By organizing data into a multi-dimensional set – instead of a flat hierarchy – marketers can grab and analyze several dimensions at once.

In the same way that a computer is organized into a hierarchy of directories, data trees are divided into pre-determined paths. First, you choose one category, such as geography, then drill-down into the next layer, such as products.

Dimensions, on the other hand, allow marketers to develop multi-dimensional models of their data. An OLAP cube, for example, has become a popular visual representation of customer data. These cubes provide three simultaneous data dimensions.

While a hierarchy only allows drill-down analysis, multiple dimensions offers more analysis operations:

  • Pivoting – When data is visualized as a cube, then the cube can be rotated to allow different perspectives on the data.
  • Slicing – Slices are like cross-sections of an OLAP cube: you can slice off one portion of the cube in order to examine it more closely.
  • Dicing – This is like slicing, but instead of extracting a cube face, you are extracting a sub-cube. Dicing is useful for examining smaller datasets that still consist of multiple dimensions.
  • Drill-Down – As with hierarchies, you can still focus down from big picture data to detailed data.

This approach to data analysis is clearly more flexible.

Not only will a multi-dimensional approach to analytics improve your marketing, it will also become more necessary in the coming years.

Why Dimensions Will Become Necessary for Marketers

Analytics are already necessary for every breed of marketer. But there are several reasons why this multi-dimensional approach will become necessary in the coming months and years:

Dimensions offer more insight into the root causes of problems. The ability to examine data from several angles or drill down to details from several directions offers more perspective.

Under a hierarchical data organization, you’ll be forced to follow the predetermined pathway defined by your original campaign – say, geography followed by products. But if you use dimension-based analytics, you can start at other points and drill the other direction. If you have a problem with one product across multiple geographies, then, you can start with the product dimension first.

Dimensions allow you to correlate business results with marketing efforts. Marketing’s effect on sales can vary widely based on a range of factors.

This can, in turn, make it difficult to map marketing efforts directly to revenue. But a multi-faceted approach to data makes it easier to discern causes and effects on the bottom line.

Finally, dimensions offer greater growth potential. Since dimensions provide more insight into marketing efforts, there is a greater learning potential. Marketers will have more insight into their audiences, which will offer more opportunities for growth. And as the number of dimensions increases over time, marketers will be able to gain more and more insight into the minds of their customers.

For now, dimensions may give marketers a small edge if their competitors aren’t using them. But we are already seeing dimensions gain more and more traction, so the exception will soon become the standard.

How to Get Started with Dimensions

Most marketers are used to the hierarchical approach, but there are a few ways to get started using dimensions right away.

If your analytics toolbox offers additional dimensions in its reports, start adding these dimensions. Instead of simply using your existing hierarchical report structure, add new dimensions to the reports. This additional layer of information will begin to take on more meaning the more you use it. So when you begin using more robust toolsets, you’ll be ready.

Use tags to create new dimensions. Tags are available in many analytics toolsets, and they can be used to add that extra dimension to your data. This is another good way to get a feel for the possibilities that dimensions offer.

Get used to multivariate testing. Multivariate testing and multi-armed bandit testing take A/B split testing to the next level: you can analyze and refine multiple elements simultaneously.

Expand your multidimensional analysis capabilities. Multidimensional regression models and ANOVA  tests both give you the ability to make more from your multidimensional data.

 

Regardless of how proficient you are with dimensions now, they will become more and more necessary in the near future. The best way to begin is by implementing a few of the aforementioned tips and then investigating some of the more robust analytics toolboxes on the market.