How to handle Big Data

Big Data simply means more data. And it is likely that the data you have access to in your company is growing fast and exponentially.

So why is this? Companies don’t have more employees, or more customers than before “Big Data”. Customer demographics are the same as before and customers don’t buy that many more products either.

Big Data is driven by the growing digitalization, and that´s because in digital channels you get detailed tracking of peoples behavior.

So it is mainly two things driving the development of Big Data:

1. Behavioral tracking in digital channels.

2. More digital channels/medias = More sources of data.

If you don´t limit what you track, and if you don´t get your key metrics from a single data source, then you get Big Data built in to your everyday operations in a way that is so challenging it could be called To Big Data.

So where does this development leave companies? In a mess actually.

If you only know a customers name, address and purchase history, you can build a focused, consequent and efficient marketing/crm strategy based on that information. But adding behavioral data from multiple channels means that setting up and executing a marketing or crm strategy is getting far more complicated. And the risk is that you don´t manage to be focused, consequent and efficient at all, which leaves you with a worse result despite the higher potential.

So what is the solution to the mess created by Big Data? My recommendation is quite simple: Scale down before you scale up!

Limit your scope by defining a certain area/process you need to measure, such as customer satisfaction, or marketing.

Then set up KPI’s and goals. Let´s stick to marketing and say that your KPI is ROI by marketing channel and that your goal is a certain level of ROI.

Then define what you mean by channel and how you calculate ROI (how you attribute the revenue and which costs you take into the calculation).

After doing this, you know exactly what data you are looking for, and that is (in this example): Purchase value and variable costs per marketing channel.

Now you are no longer dealing with Big Data at all, you only deal with a fraction of all available data and by doing so you have increased the likelihood of getting some real value out of your data from “no chance” to “it can actually happen if we stick to the plan and don´t get confused by all the noise”.

Then what you also need to do is to secure that data and make sure it is correct (I will write more about the necessity of ensuring high data quality in later blog posts).

The importance of setting up goals is a well used cliché for a reason. Having a goal works like a searchlight in finding the data you need. Not having a goal will inevitably make you look for data where it is already lit, just because that´s the only place you can see anything (i.e. you only look at data you already happen have access to, regardless of its usefulness).

If you want to be a bit crude (which I do) you can say that Big Data only affect those who don´t know what they are looking for, and if you don’t know what you are looking for, you really don´t know what your are doing.

A more friendly approach would be this: The challenge with Big Data is that you need to opt out of certain things to be able to deal with it. And our natural instinct is to do the opposite; to make sure we collect as much data as possible so we know we have it if we need it. Don´t fall into that trap; don´t go for everything, go for what you know that you need.

See also my recent blog post: Pinterest Analytics – More fragmentation added.

Andreas Franson, andreas[@], +46 733 56 41 51

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

× 8 = fifty six

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>