by Mateusz Gordon, e-commerce expert & Tomasz Lechowicz, business advisor at Gemius 

IDC analysts estimate that the amount of created and stored information globally doubles every eighteen months. For companies operating on the internet, easy access to a wide scope of data sources is an obvious need. What is important, however, is the skill of archiving and performing real time analysis, which in turn provides a dynamic picture of the market situation, including the business environment, as well as forecasting and strategy building. In such industries as e-commerce, we do not only deal with companies that are naturally set in the virtual realm, or with vast amounts of stored data. We can quickly react to changes in customer (internet users') preferences, adjust marketing budgets and the offer in response to the competitors' undertakings and general trends.

Wheat or chaff?

The 'vast amounts of data' (Big Data) referred to above are both a chance and challenge to business. Innovative technological solutions are necessary to separate wheat from chaff, providing the right tools for analysis and ordering up information. How can you benefit from advanced analytics and how is Big Data used in e-commerce?

Segmentation of audience

The most basic analytical tool is segmentation of users. This provides an overview of those net users who are either customers or visitors to an e-commerce platform. It comes down to description of customers with a set of features that define their similarity to others in the same group and at the same time, differ them from others. The data used to perform such segmentation may be obtained from:

• CRM systems (customer base management; contact details, shopping history, frequency of shopping, etc.);
• external web analytics systems (following internet users' actions on a website, their behaviour on a website, sources from which the users came, what they dropped in the cart before finally abandoning it, etc. );
• external systems within the framework of a chain or brand (e.g. loyalty cards and data about offline customer behaviour);

• ad hoc research (surveys clarifying some of the user's actions, e.g. why they abandoned the cart, whether they have seen the ad, if they have been to competitors' websites, what price comparison sites they use, etc.).

Lifecycle

Every internet user, including those visiting the e-commerce platform, is at some point of a customer lifecycle. The cycle is most often divided into three stages: gaining, conversion and retention. At each of these stages the user's 'life' can be analysed with consideration of:

• the costs that should be incurred per an internet user to change his profile (e.g. from a potential to an obtained customer, from an obtained to a converting one, from a converting to a loyal or retained one);
• the chances for such profile change (the probability that, for example, a potential customer becomes an actual customer of an e-shop);
• the profits that can be obtained as a result of profile change or of keeping a user at the current stage (converting or retained ones);

• the average time for which an internet user is to stay within a particular profile (i.e. at one of the stages).

By analysis of the above indicators, you can assess the potential of individual website users, optimize the indicators (with the available marketing tools) and improve sales figures at the online platform.

Customer potential and loyalty

 Accurate analysis of data is a way to assess the loyalty of customers (and their groups) as well as their purchase power. 'Share of wallet' is a kind of comparative analysis (based on data on the entire segment of market, including competition results) concerning soc-demo profiles of internet users (their age, education, place of residence, earnings, number of persons in a household, etc.). This way it is possible to fish out the data describing your group of interest from an ocean of information, to check what amounts are spent on shopping, what share of that money lands in your budget (by confronting market data with your own data generated with CRM).

Effectiveness

Big Data also offers you the possibility to conduct a number of analyses carrying effectiveness value (ROI) – providing information on how profitable an e-business investment is. A most interesting example, which may serve as an illustration of the new trends in the online world, is 'post view' – an analysis of a campaign's branding effect. Gemius data (AdMonitor reports) allow to conclude that internet users are increasingly inspired by marketers' messages. They interact with a brand or make purchases not as a result of a conversion from an ad to a website. Instead, they take up action based on their interest in a brand, in response to nothing more than a campaign (searching for products online, liking a Facebook fan page, etc.).

***

All possibilities offered to e-commerce by Big Data are strongly related to automated marketing processes and implementation of advanced IT tools. Behavioural targeting, or marketing communication adapted to a user profile (compiled on the basis of historical data on such user's online behaviour) may be one example of that, so can recommendations, personalized messages, as well as sales forecasting (management of stock based on Big Data). Huge data resources, when appropriately exploited, economise time and costs, add up to effectiveness of a business, with immediate impact on profitability.