Why personalization isn’t always needed to provide valuable experiences
Exploring an alternative approach to data-driven personalization
As an aspiring UX researcher and strategist, I have received plenty of advice from other fellow professionals on how to expand my knowledge in the UX field.
Among these valuable conversations, one thing that kept popping up was to attend UX meetups. Dipping my toes into these meetups has been incredibly useful in helping me gain a real-world understanding of UX. But most importantly, these meetups have often led me to question the way certain things are approached in the field and whether they necessarily need to be that way.
A recent meetup on the topic of data-driven personalization got me thinking about whether collecting as much data as you can is always needed to provide customers with valuable experiences.
In this article I’d like to elaborate on why detailed personalization doesn’t always have to be the go-to approach.
Why does it matter?
We now live in a digital era with enough online activity going on to paint an adequate picture of our behaviours and interests. This trail of data crumbs has certainly been picked up companies to come up with personalized campaigns and tailor individual experiences to fit customers’ rapidly changing needs. Amazon, for example, tracks the majority of its customers activity including online purchases, activity across the web, voice commands, grocery shopping and even reading habits such as which book parts are highlighted.
Of course, this has not gone unnoticed by governmental bodies or the customers themselves leading to extra regulations. As a result of privacy concerns, the General Data Protection Regulation (GDPR) came into force on 25th May, 2018 to protect people’ personal information. In line with these regulations, companies have had to adopt several approaches to protect customer data including anonymizing it. So what does this mean for the detailed personalization approach? In short, lots of extra work to ensure that all the data being used is collected by the right parties and in a transparent way.
What this boils down to in the end is a cost-benefit analysis in which many companies have taken some steps to ensure that they can reap the benefits of personalization without comprising data security. Still, I’d like to illustrate that collecting extensive personal data isn’t always needed based on a case made during the meetup.
Let me tell you a story
The story takes place at the airport where a passenger is waiting to board her flight. While waiting for her flight, she checks out the airport’s website to get information on the shops and restaurants located at the airport. Because she wants to have an early dinner, she navigates to the ‘Food & Drinks’ page of the website where she is first presented with a picture of an Italian restaurant located at the airport.
Are the restaurants placed in a random order on the web page? Not at all. The airport has been collecting aggregated data from all its passengers on which restaurants are the most popular during that time-slot. Aggregated data is summarized information to gather statistical insights. Unlike personal data, the data can only be viewed per group, not per individual. Based on the analysis performed by the airport’s data scientist, the Italian restaurant was found to be most popular and hence placed at the top the page. In this case, time of the day was used as a data point to increase relevancy and provide customers with a better experience.
What’s the main takeaway here?
The airport example illustrates that detailed personal data is not always necessary to improve customer experience and provide business value. The approach used here centers more around segmentation instead of personalisation.
Segmentation involves grouping data together according to identifiable characteristics which can include age, gender, geography and time. The focus here is on correlations between groups at a more general level instead of personalisation which is broken down further.
According to research done by Campaign Monitor, segmentation can lead to increased revenue till up to 760%. Now I must admit that segmentation isn’t always the way to go, especially when it comes to email marketing or campaigns where your audience expects highly personalized communication. So how can you decide which approach to go for?
Some practical tips
1. Estimate how much data you can collect
Your ability to create detailed segments and move towards personalization depends on how much data you are able to gather. The story above takes place at an airport with millions of passengers traveling through on a daily basis. There is a high volume of total passengers but a low frequency of specific passengers returning again shortly. This means that you have access to lots of aggregated data which you can use to your advantage. Alternatively, a customer subscribed to a newsletter interacts frequently with that platform leaving a trail of personal data behind that can be collected. Therefore, context matters.
2.Identify your segmentation criteria
Once you know what data you have; you can decide which factors you want to focus on to create better segments. Identifying which data needs to be treated as individual data or a specific segment can help make your step towards data analysis and expressions of that data into valuable experiences, much easier.
And that’s a wrap!
I hope that this article was useful to you and I would love to hear about your experiences on the topic. Since this is my first time writing here on Medium, I am open to any constructive feedback on my article!
