Have you ever done your grocery shopping online? If so, you’ve probably noticed that the products the website recommends are remarkably well-suited to you, even if they are new products or products you’ve never heard of.

This is possible because of the way good supermarket operators use the vast amount of transaction data they collect every day. Loyalty cards such as Countdown’s Onecard allow the supermarket to present you with product suggestions based on your individual history, and gather detailed information about the buying habits of demographic groups to a very fine-grained level, so they can target you with products that other people in your demographic have found appealing.

This is an example of the rise of “Big Data”.

Big data is an increasingly common phrase used to describe the analysis of datasets of such enormous size and/or complexity that they cannot be analysed using simple, traditional tools such as spreadsheets or desktop software.

The aims of big data analysis are usually to

  • spot new trends and opportunities that competitors might not be aware of
  • tailor marketing communications to groups (or even individuals) based on a holistic view of their behaviour and preferences
  • provide a better mix of products or services based on what is selling, where, and when
  • adjust prices dynamically to optimise profit
  • optimise stock levels based on predicted demand
  • etc.

If you’re in business, you already have access to a great deal of data, in a number of disparate systems, both online and offline, in the cloud, maybe stored on your server, your PC, your phone, written on pieces of paper and filed in cabinets. The important thing is to integrate the data from these systems in a way that provides you with meaningful information that helps you make better business decisions.

The best sources of data about your audience’s online behaviour  include:

  • Google Analytics
  • Google Adwords
  • Social signals (e.g. Facebook likes, tags, updates, tweets, Youtube comments, blog comments, etc.)
  • Email newsletter subscriptions

Other online data might come from

  • Competition entries
  • Survey participation

There are also many other business systems that you could use to capture behavioural data

  • Accounting system – e.g. who are your most profitable customers?
  • Customer relationship management (CRM) software

It’s a great idea to put together a framework of the big data that you can capture, based on your organisation’s goals, and then start making smart decisions with what you’ve found.

There could be patterns, e.g. you could spend a great deal of time and money marketing to a certain demographic, e.g. 18-25 year old males living in Auckland, see a spike in sales from that demographic, and then see from your CRM software that hardly any of them made a repeat purchase and from your accounting software and digital marketing campaign management tools that you spent more on acquiring them than they brought in to the business.

“Oh well”, you might think, “it was good to get the brand out into the marketplace, even if we didn’t make any money in the short term”. Your social media monitoring software could then tell you whether awareness of your brand has increased across the social networks – and can even tell you the proportion of people saying nice things about you.

So based on all of these insights you might rethink your approach for your next campaign.

From a purely marketing point of view, all of the data makes the most sense when you think about the customer journey – from attracting the prospect’s attention all the way through the conversion, retention, and finally getting them to advocate on your behalf.

These are the metrics that are typically used to measure behaviour at each stage of the lifecycle:

1. Attract

  • Website visits by source (social media, ad campaigns, email, referring sites, etc.)
  • Bounce rate (what is their ability to deliver qualified traffic?)
  • Branded search volume (how many people know about the brand?)
  • Unbranded search volume (how well are we performing in search engines for generic terms?)
  • Cost per click – should be as low as possible

2. Engage

  • Pageviews per visit
  • Video views
  • PDF downloads
  • Social shares
  • Facebook like, Twitter follow

3. Convert

  • Online purchases
  • Demo requests
  • Voucher downloads
  • Contact forms submitted
  • Requests for more information

4. Retain

  • % of repeat online purchases
  • Newsletter subscribers
  • % of returning visitors

5. Advocate

  • Social media shares and mentions
  • Social sentiment
  • N.B. it’s important to have monitoring in place to measure global social signals, not just from your own site(s)

Your next steps

  • Think about the customer journey(s) in your organisation
  • Then think about how to integrate online analysis tools with your business’s own internal systems to get a complete picture of these journeys.
  • Design marketing campaigns so that customer behaviour can be tracked all the way from awareness to advocacy.
  • If necessary, tweak campaigns as they are running.
  • Rinse and repeat…