How to make personalisation work for you


By Renate van der Vaart, Customer Experience Specialist & Sitecore MVP

As a customer experience specialist I am always looking to find better ways to communicate with consumers and understand them better. But how do you grab the attention of a customer when everybody else is competing for seconds of their time? And even more importantly when you have their attention, how do you keep them and ensure they come back for more?

If you read the marketing press, you'll see everybody is talking about how getting the attention of your audience is about making your stories personal and relevant. But how do you actually make your communication more personal? We can all give examples of when we've visited a website to buy shoes, clothing or holidays and couldn't find anything to buy, and for the next couple of days, you see adverts and promotions for these products all the time. Unless you clear your cookies regularly, these brands will follow you everywhere - it's personal, but also very irritating.

So how can you make your communication personal and relevant, without being intrusive? That’s the question we asked ourselves. Together with our partner Sitecore, the leader in web centric customer experience management, we explored this question and set up a testing program of personalisation and A/B tests to see what was working and what not.

What did we learn in this period?

1. Choose your testpage carefully

The page you use for personalisation testing needs to have enough visitors to give a clear indication of each variant. So do some data analysis on how many customers come to the page and try to find out what they are doing next. We call this defining the content-path. To find out high entry pages you can use your Google Analytics account. After finding a high entry page I used the path analyser.

2. Make the test-variations not too small or to big

You need to find a balance between the variations. If they are too big you can’t compare them in your results because it's hard to define which elements the customers are attracted to. But on the other hand, if you make the variations too small, you can’t conclude why the customer prefers the default or variation version. So it’s very important to find a balance.

3. Define a clear goal

This sounds logical, but with only defining what you want to achieve is not enough. It's insufficient to describe “what you want to achieve”. Ask yourself what the “background” of the test is and if it is in line with the goal. Then you can test the outcome against the hypothesis. In this way you defined the full context of your personalisation test.

4. Big brother

Watch out you don’t become big brother, as it is a very thin line in being personal and intrusive, especially with all you know about your customer. For example welcoming your customer with “How nice that you visit our webpage” if you know your customer is coming from Facebook. It is better to adjust your text in line with the copy of the Facebook post and make it feel personal in a subtle way. If you scare your customer with a too personal message you will lose them and they will never come back. So use personalisation wisely.

As we learn more, we see that we improve the quality of the customer experiences using personalisation and A/B testing to drive engagement with our brand. Along the way we develop an understanding about the different personalisation tactics and using them to optimise the contact we have with our target group.

The results we are seeing are showing an increase in conversion uplift across different goals like membership sign-ups, engagements and sales. Also we delivered a positive ROI of costs to set-up and execute a single test. We will continue with testing, to go on improving the quality of our personalised (in the right way) and relevant (the correct format) communication experiences.