12

CX segmentation

6 min read

Create communications that actually matter to your customers

Having the right data at the right time is essential to a successful experience. We’ve all been on the receiving end of too many emails that bombard us with information and offers. Conversely, we’ve all felt the frustration of not having enough information, or for receiving it too late. Here’s how to get it right.

Data and segmentation

The quickest way to create a personalised experience is to slice up your data and send contacts relevant information. Segmentation allows you to direct your budget into smaller, more targeted campaigns that should trigger a better response and give you a better return on your investment.

You can start with the ideal customers within your own database and use these as the basis for your personas. Relevancy and personalisation are key to ensuring we understand users behaviour and trigger the relevant message at the right time.

What is ‘personalisation’?

Buyers and customers expect their marketing to be personalised. But what does that really mean? Here are some examples we can all resonate with:

  • Opening up an app and seeing relevant product recommendations
    Visiting a website and being signposted to a relevant product or service
    Status updates such as an email and/or text to confirm that your recent order/appointment/delivery is confirmed or on its way
    Being contacted in the way that suits you best (e.g. post, email, SMS, WhatsApp)
  • Seamless checkout, contactless payment, user-specific discounts

But first – data!

To create personalised customer experiences, you need to create a single data layer. This means unlocking data silos in your business and connecting them. This could involve using an external consultant who possesses the skills to join these moving parts together. Whoever you choose to manage this for you, they should insist on a full data audit to ensure the resulting database is the best it can be. For example, the data should be GDPR compliant, clean and from ethical sources.

Data segmentation – great ideas to get you going

When it comes to segmentation, there’s no one-size-fits-all approach. What suits a B2C client won’t necessarily be a good fit for a B2B. However, most segmentation falls into four camps:

  1. Segmentation by customer type: split your customers into ‘pots’ based on their purchase history, defined by the average number of purchases across a chosen timeline, and the average time between purchases. This is a traditional segmentation approach that works for most sales models. For example: leads, first time customers, active customers, lapsing customers and inactive customers. For service-based models and those with long periods between purchases, you could use renewal dates to set parameters for different segments.
  2. Segmentation by interest: this uses purchase behaviour to determine what customers are interested in. At its simplest, you’d break down your key sales by category (e.g. kitchens, bathrooms and garden products for a homeware store) then review all your purchase data, website tracking and email opens to place those with high ‘scores’ in those categories in those segments. 
  3. Segmentation by location: this is a fantastic way to deliver localised campaign activity. Using a customer marketing platform, you would be able to identify all the contacts living in a specific radius or location and send emails specifically to that segment. Dividing your database into different country locations is also a powerful way to create a personalised experience, even if you operate thousands of miles away from your contacts.
  4. Segmentation by engagement level: the best customer marketing platforms work out how ‘engaged’ customers are with your communications by scoring their interactions with you. Someone who rarely opens an email might be a candidate for a monthly round-up, whereas an engaged subscriber who clicks on everything you send could receive more content.

Segmentation – how much is enough?

A word of warning: only collect just enough data to achieve the outcome you are after. It’s tempting to ask for people’s names, addresses, gender and much more, but asking for too much, too soon, will likely put people off. You will need to test this theory, of course, and in time you might gain enough trust among your subscribers to ask for additional information – for example, their date of birth – in return for an appealing offer or discount.

Drag Read