What does Big Data mean for your company?

Something that is widely discussed and often discussed in digital marketing is data and the importance of a "data-driven approach" to how we reach our customers. It is not uncommon for it to appear in these discussions buzzwords like "big data" or "machine learning". However, in the same discussions there is often a great uncertainty about what big data actually means and what effects it can bring the company. We are, therefore, in this post partly to go through the basics of what big data means but also how all the information contained in the data can be used to contribute benefits towards your goals.

What can Big Data mean?

To lay a good foundation for the continued part of the post, we thought briefly go through what big data can mean. Please note here that what we say can be, one of the basic parts of the basic parts of the big data area. Big Data can also include everything from managing to collect large amounts of data to use it to build models. In general, however, the word big data refers to data sets that are so large that traditional ways of managing the data and processing it are no longer sufficient. To sum it up in a clear way, one can simply say that when Excel constantly crashes because the files start to reach sizes of several gigabytes – then you start to approach what is called big data.

Big data is also used to terminate the processing of the large amounts of data that we can collect today. One of the areas of techniques available for processing big data is machine learning. This in turn is based on different techniques that are based on training an algorithm to learn from the data it processes to find patterns that in turn lead to an understanding of the data and enable the use of the data for marketing purposes.

Rating and Clustering of Big Data

To make the above mentioned a little less abstract, we thought to refer to an exercise we did on a course in 2017. The exercise was that we had a table with samples of the contents of different chemicals in about 300 wine bottles. These, in turn, would have three different names based on the type of wine it was based only on the chemical content of the bottles. To help, there were three bottles that had been defined as the three different target groupings of the content. With the help of machine learning, we were able to train an algorithm to recognize what the three different reference bottles contained, and then it could be used to divide the remaining 300 bottles based on their contents. This works simply by the algorithm calculating the probability of the contents of the bottles based on the references. For the computer, these calculations and groupings took less than a second to do while we had had to sit for a full day in Excel to group in the same way – there's simply a lot of time to save.

What we have mentioned here are the basics of what is called "Classification", it is simply based on the computer classifying based on specific goals. Another option would have been to use a technique called "Clustering"then you instead give the computer the freedom to look for built-in patterns in the data that it groups into a specified number of clusters. In all likelihood, similar results should have emerged in both techniques for the example of the wine bottles, but what distinguishes them is the availability of information on final results in advance.

Segment your Big Data

At this point, perhaps it is time to anchor what has been mentioned above in your company's needs. One idea that may have already struck some of you is that the above could be used to group your customers into different segments using a variety of variables or grouping products.

To make it even clearer, we thought we'd inform you of an opportunity where you can take advantage of this directly in your Google Analytics platform by going to Audiences –> Behavior –Conversion probability (however, requires that Ecommerce tracking is active and that there are at least 1,000 conversions per month). This report reports segments with different probability that a user will convert within 30 days. The calculations are done with machine learning by Google where they use algorithms to look for patterns and similarities with users who have already converted – how exactly it is built is, however, secret. The data contained in the report can of course be used for remarketing and can therefore lead to increased efficiency in your advertising but also to distinguish patterns by analyzing the demographics of the segment.

Of course, you can also use the technologies mentioned and others, as well as technologies outside of Google Analytics and other platforms. However, it places relatively high demands on the data collected based on the objectives that want to be achieved, but it should not be forgotten that the possibilities are endless when it comes to the results.

We at Anegy are happy to help you go through the opportunities that are available for you to take advantage of your big data. Please contact us and we will book a sitting and look at your possibilities based on your needs and conditions!

Do you want a free guide in
tracking with GTM?

Fill in your details and we'll send you a guide with our best tips. 

News! Webinars

Read more and book upcoming opportunities!


Blog of the year - Anegy.se