The art of segmentation - that is, splitting a target audience into groups so that each can be targeted in different ways - is evolving fast, as deeper and broader sets of online consumer data become available, according to Ed Weatherall, Business Development Director for VisualDNA. This article is copyright 2015 The Best Customer Guide.

With the availability of completely new types of data about the psychology, behaviour and buying propensity of online consumers, brands and advertisers can now use data to drive their creative messaging, tailoring their communications to appeal on an emotional level to different personality types.

Firing off a weekly email to the usual suspects can still work as a means of driving revenue, but most marketers are aware that sending the same message to all is becoming less and less effective. Today, brands and advertisers can use new data to power fact-based approaches that captivate the audience with relevant content and nurture long-term brand loyalty - producing game-changing results. They can also build detailed profiles showing who their online customers are and why they act the way they do, adapt their segmented communications strategy accordingly, and find new people like their existing best customers online. The result? Larger, more engaged and higher spending audiences that can be reached more easily and cost-effectively.

The challenge of online segmentation
It is time to move on from just using third-party postcode data, including census information, DVLA data, financial data, consumption data, and so on. While this data can be rich and cover multiple personal attributes, it has two major shortcomings. First, a single postcode can represent anything from one person up to 3,215 people, with each postcode covering 43 people on average (source: Office for National Statistics, 2011 Census). Second, it doesn't provide any useful way to link offline profiles with online users. Even where data providers can go down to the level of individual households, each may contain multiple digital users and not all will have the same tastes, preferences and buying propensities.

If advertisers and brands are content with a race to the bottom, selling commoditised products and services via a one-size-fits-all approach, then traditional postcode-based segmentation may be enough. But for businesses looking to build a loyal, high-value customer base, with a brand that transcends products or services to build a lasting emotional connection with the consumer, this kind of segmentation has clearly had its day. Brands that want to compete on attributes other than price, availability and features must find better ways to understand and segment digital users if they are to find their target audiences rapidly and efficiently online.

New times call for new capabilities
Brands and advertisers need not only a larger set of attributes on digital users but also the capability to see how these attributes evolve over time - not least, in response to ad campaigns, and for A/B testing of promotions or onsite content. Even where traditional data providers offer what appear to be sufficiently large numbers of attributes, their segmentation is usually fixed, and does not provide information on whether a particular set of customers is over- or under-indexed for a particular attribute. Essentially, this leaves brands and advertisers unable to see the wood for the trees.

To build lasting relationships with customers, brands and advertisers need to adapt their communications - including elements such as tone of voice and imagery - to suit each individual. The broad-brush approach from conventional data providers makes it all but impossible to understand and speak to individual segments of the audience in the right way.

Even where companies have rich in-house segmentation that adequately captures the attributes of their existing audience, they typically lack the ability to find new users online with the same attributes (and therefore the same propensity to be a loyal customer). For offline customers, segmentation by postcode can make great sense, because proximity to the store will heavily influence propensity to buy. But for online customers, it is unlikely to be the defining factor in choosing one brand over another: much of this will come down to personality and preferences rather than demographics.

Happily, new sets of data are becoming available that permit dynamic segmentation at huge depth and with unique additional attributes based on online consumers' personalities and emotions. These sets of data, typically overlaid on existing in-house segments, explain whoonline customers and prospects are and why they act in the way they do. They also provide a dynamic view that shows how the audience changes over time.

Having a dynamic view of who people are is extremely important: demographics change frequently, but this doesn't necessarily have an impact on brand preferences or buying propensity. For example, changing jobs, moving house or getting married is unlikely to immediately change a person's taste in music or books.

Creating the right experience
In this highly competitive world, consumers have a huge number of options about what they buy, where and when. By using richer, psychographic data - including information on personality and psychology - companies can reach out to different types of consumers in ways that appeal directly to their emotions, thereby building lasting brand affinity. Equally, they can use new sets of data to clearly define the top 10% of their customer base, and reach out to more people like them online.

Using psychographic in addition to demographic data, companies can start to see which online consumers are impulsive by nature, and which are more conservative and measured in their purchasing behaviour. They can also, for example, segment by attitudes to spending money. With this degree of insight, and the tools to extend it to search and email marketing, online businesses can target the right people with the right offers, rather than indiscriminately blanketing everyone - and potentially alienating many future prospects - with the same messages.