Marketers are increasingly being tested against their ability to deliver personalised customer experiences across multiple channels.
There is a good reason for this – it has proven to be more effective. Targeted campaigns deliver much higher conversion rates than batch and blast efforts.
Unfortunately, too often the foundations on which these experiences are built – data, segmentation, and the insights they generate – is poorly designed.
For example, data is often trapped in seperate silos, preventing any linkages and the insights this can produce. The ultimate result is incomplete or poor segmentation and in turn, disconnected customer experiences.
If either the data or the intelligence applied to it are lacking, the result will be substandard. That is because disconnected data combined with disconnected intelligence produces disconnected experiences.
Breaking down organisation data silos is the first step, but it’s not an easy one. This is particularly challenging for organisations dealing with complex and often outdated legacy systems. However, success on this front creates a valuable resource for business, and one that can underpin targeted marketing campaigns.
This is an article in our series on customer experience where we focus on topics relating to connecting data, intelligence and experiences. Further Reading:
- Do Marketers really want to be data scientists?
- Great Customer Experiences Rely On Robust Identity Management
- Silo Busting Essential To Delivering Personalised Experiences
- Why Inconsistent Messaging Is Undermining Customer Experience
- Campaigns Don’t End When You Hit Send: The Importance Of Feedback Loops
- What’s The Key To Improving ROI On Advertising?
Of course, data alone lacks utility. The question for marketers is how to pool the data, and then create useful insights, which can themselves be turned into appropriate messaging for customers and prospects.
Segmentation may sound like marketing 101, but it is a process that still trips up some marketers. Effective segmentation is foundational to any marketing or advertising strategy and spans a whole range of tactics and meanings.
At a simple level it is essentially labelling groups of people based on behaviour, demographics, marketing tactics, and personas.
The temptation is to segment once, establish your customer labels then move on to insights. But effective segmentation is a continuous iterative process that enriches your data over time. That also means segmentation must remain attached to the data and the technology stack.
But the trap many marketers fall into is slicing their customers base into labels without connecting the technology stack. While the labels may be based on sound data insights, failing to tie them back to the data creates problems as it is unclear what characteristics make up the labels.
This is known as aspirational segmentation and it can create problems when it comes time to execute. For instance you might find that when you run a campaign to a certain segment you can’t actually find that group of people because they are not connected to the technology stack.
Let’s say your target group is ‘fun-seekers’. Marketers need to know exactly which attributes or data points qualify a prospect as a ‘fun seeker’. Do they frequent music festival sites? Listen to a particular genre of music? Or perhaps take advantage of travel promotions? However marketers wish to define it, knowing this allows for more effective targeting, but also fast tracks increasingly automated messaging.
If the funseeker segment isn’t attached to data it’s almost impossible to target, and without data it is also subjective.
But getting it right means marketers have a foundational knowledge of their customers which can be leveraged into an effective data driven marketing strategy.
If segmentation is data-driven and data attached, then it can be executed upon and it’s part of the technology stack. So not only can you campaign, but you can also run reports. Then, you can test and tweak it, and ultimately optimise the campaign.
Segmentation means lots of things. It can be as simple as selecting groups of males or females. More data complex segmentation might involve merging multiple data sources and using data science and machine learning to build a customised segmentation system.
The important thing is segmentation remains attached to data and becomes executable.
Marketers should aim to segment all their data, but it doesn’t have to happen all at once. Start small, segment the data you do have access to and see what works. Expand when you see the results. The most important thing to to get started!
Cameron Strachan leads Strategy and Go To Market, APAC, at Oracle, which is a member of the Which-50 Digital Intelligence Unit. Members contribute their expertise and insights for the benefit of our readers. Membership fees apply.