Customers want personalised experiences.

To ensure personalisation, you need to know where your customer is and how they are interacting with you at any one time. Every interaction provides you with the best insight as to what they care about and what you can do to ensure you continue to delight them.

That’s where your data source comes in. You can use your data to improve your customer experience – as long as you’re using the right data in the right way. 

We’ve seen it in the Meeker Internet Report; customers consistently respond better to experiences that are personalised in the right way, when they’re treated respectfully. Targeted campaigns deliver much better conversion rates than batches and blasts.

Read more: 7 insights for start-ups and small businesses in the Meeker Internet Report

It’s not a malicious lack of respect that plagues marketing. Instead, it comes from poorly designed experiences based on flawed, dysfunctional or disconnected data gathering and management.

Everything starts with data. When your data is trapped in different silos, or improperly labelled or treated, you fail to generate meaningful links and insights. Creating a clean, connected data source is a first step and it’s a hard one. This is particularly so for organisations who have complex systems that have had to be updated with each new iteration of the business – or even just each new iteration of computer upgrade.

But success in data management connection creates a valuable resource for businesses. Once you have free-flowing and connected data, you can start to ask meaningful questions of it, and create stronger customer experiences.

Read more: Social media delivers long-term brand value

So what do you do once you have all your disconnected data in one place?

You start splitting it up again, of course. But look at it like this:

• Disconnected data: You have 4 different data sources. You ask the same question 4 times and get 4 different answers.
• Connected data: You have 1 data source. You ask it a question and get 1 (more meaningful) answer.

Segmentation starts with 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 continuous; continuous review and monitoring ensure continued relevance. You want to do more of what works, and less of what doesn’t, don’t you?

Start small, segment the data you have to test and see what works. Expand and modify when you see the results. The most important thing is to get started!

The inspiration for this story was originally spotted on which-50.