How to redefine personalisation for your users

June 9, 2023

7 min read

Share on

Shaan Bassi

Shaan is a neuroscientist turned entrepreneur. During his time as a venture builder at Prehype, he became familiar with the many problems & challenges of product development. By connecting his two passions, neuroscience and product development, he is solving those challenges with Kouo, an Emotion Analytics Platform he built to help companies gather customer feedback in real-time, without requiring user input.

Decoding the status quo

Product teams today have access to an abundance of tools and data. However, despite the excess of analytics tools or the size of the user base, there remains a void in swiftly gathering insights into the ‘Why’ behind the ‘What’ — the rationale governing user behaviour.

There are attempts to fill out the blanks on the ‘Why’ using methods like NPS, CSATs, interviews and surveys, but unfortunately, they are unreliable sources of truth.

As a result, our personalisations tend to fall into a handful of categories stunting the growth of our products:

More is more: This algorithm-based approach continuously serves large volumes of similar content until a shift in behaviour prompts a reassessment. This approach might perform well within eCommerce but falters in other areas like Wellness, Music, Education, Gaming, Health, and prominently in Social Media.
Faux personalisations: These create an illusion of personalisation, usually during the sign-up flow, with users ultimately accessing the same content journey as everyone else, no matter their sign-up preference.
Notification-only: These semi-personalised prompts are based on behaviour. But without insights into why users are leaving your app and the emotional impact of the in-app experience & content, these nudges could prove counterproductive, even driving users away. Not naming names, but we all know that one app that keeps nudging us to learn more. (Grateful for sleep mode blocking out some notifications 🙏.)
User-Controlled Personalisations: Though valuable, these customisations disguised as personalisation might confuse users who don’t yet know how to leverage your app, potentially leading to churn.

The drawbacks of current user feedback tools

CSATs date back to the mid-90s, and NPS to 2003. That means our latest invention in understanding what our customers might feel about us, in a quantifiable way, is 20 years old.

While user interview methods have adapted to our digital lives, overall our toolkit often proves to be dated, cumbersome, or unreliable, suffering from a range of issues:

Inordinate qualitative feedback: Time-consuming to align with analytics.
Post-facto data: Most tools can’t provide mid-experience insights without disrupting or skewing the experience.
Delayed insights: The prolonged process of setting up interviews and focus groups, analysing data, testing and reassessing hypotheses, and building products makes user insight and hypothesis testing at scale unfeasible for most companies.
Bias in feedback: Qualitative data carries inherent bias from both users and analysts.
Insufficient data points: There aren’t enough valuable data points collected during the feedback-gathering process to make confident, data-driven decisions.
Lack of a North Star metric: Qualitative data cannot provide a standard metric to measure success.

Unfortunately, these constraints have skewed our attitudes towards customer insights affecting our ability to personalise and meaningfully adapt our products. Some companies have de-prioritised user feedback, while others have hired a single user experience researcher (UXR) for large regions, thus solely relying on user habit data from backend analytics for personalisation.

We need more clarity in our approach

When delving into the topic of user feedback with product managers and product leaders, the responses seemed mixed at first. Some confessed that they ‘don’t yet focus on user feedback enough’ at their organisation, while others suggested that they ‘can build without focusing so much on user feedback’ and that ‘six user interviews a week is enough data and insights into what users think’. It would be easy to assume that user feedback is unimportant and that running interviews and processing qualitative feedback for patterns is the best way to get insights.

However, when asked about the success of products and how they saw products evolve to meet consumer needs, one thing was clear – everyone focussed heavily on the value of personalisations. And for that, they needed insights into their users’ lives. But with tools so cumbersome and offering so little clarity, everyone hoped to get user feedback some other way, using a more quantitative data source.

The risk of relying solely on behavioural data

Because of the above reasons, many companies use analytics data on behaviour and habits to understand changing user attitudes. However, reliance on behavioural data alone can be risky, as waiting for behaviour change places us on the back foot.

Behaviour change reveals ‘what happened’ and ‘where,’ but not ‘why’. Focusing on behaviour change alone broadens the search space for potential personalisations, prolongs the testing and hypotheses-building cycles and can ultimately mislead us.

Instagram’s ill-fated attempt at platform changes, informed only by user habit data, underlines the value of understanding the ‘why’ behind user behaviour. The changes resulted in massive user backlash, forcing Instagram to revert some updates. And while you might think that having Meta as a parent company has all to do with the bad reaction, across various products, around 80% of development time gets wasted on features that never get used because they don’t serve a user need correctly.

The solution and how to unlock the potential of personalisation

To effectively improve personalisation, we need to foster meaningful relationships with users by understanding and adapting to their needs. For that, we must invest in the right tools that enable swift customer insights, necessary for unlocking better personalisations.

The solution to all our personalisation troubles lies in automating feedback collection and moving towards more digital empathy. By reinventing feedback systems, we can balance user experience, technology, and business needs while solving customer problems more quickly.

Currently, with the help of AI and incredible ML models, we can leverage existing data sources like wearables and other smart sensors to measure user physiology to get quantitative insights on user emotions which serves as feedback on their experience while using an app or engaging with content, all without any bias or user input. Rich insights on users’ emotions when connected to behaviour, triggers and trends pave the way for dynamic content and experience personalisations that are actually able to respond to users’ varying moods.

The future

Going beyond individual-level personalisation to dynamic emotion-responsive personalisation that accommodates users varying needs will be the next seismic shift in the tech industry. The future leaders across sectors will be those who grasp the essential role of personalisation in enhancing user engagement. And this journey begins with redefining our approach to collecting and understanding user feedback.