Think AI-first: The journey from segmentation to hyper personalization.
Two people are going down street, one of them (let’s name her Jane) enters store A and the other (let’s name him Matt) store B. Jane, who just entered store A is looking for a black dress, however, the store owners decided to showcase all their clothes under one space, menswear, women’s clothing, kids, accessories all placed tougher with no real order or logic. On the other side of the street Matt, who entered store B is looking for black pants, he enters the store and sees a sign “Men floor-1”, as he was heading towards the top floor Jane was just starting to browse through different items in store A.
Although Jane and Matt both started their exploration at the same time it was far from being the same shopping experience.
Now, if you doubled check the title of this article when reading this story, don’t worry, you are in the right place, I’m not going to write about clothes nor accessories, in fact the story above represents a simple form of segmentation, where Store A items were showcased with no order, causing confusion and time spent, Store B directed customers to the relevant location of the items they were looking for.
From segmentation to hyper personalization
As the story above shows, segmentation can be defined as a process where the target market (i.e. people who are looking for clothes) is being broken down to groups that have, or are perceived to have, common interests, needs and priorities (i.e. Matt, is a man who is interested in menswear). The goal of a segmentation strategy is to get better insights on how homogenous groups (to some extent) behave, operate and communicate in order to reduce friction, understand, target, and increase the likelihood for these groups to engage with your brand, product or message.
Segmentation is not personalization! While segmentation makes your offering more relevant to certain groups, it is still very broad and requires more granularity and knowledge about your customers or users. The level of the personalized experience depends on the quality and quantity of data you collect about your products, users and their environment.
Understanding personalization
Let’s think about a mobile push message campaign for Store B. With a very basic segmentation we will be able to create two templates one for men and one for women and send the relevant template based on the customer gender.
Now, as mentioned above the more data we have about our users and their behavior the higher the level of the personalized experience that we can offer. So, if we know (based on purchase history) that certain users in our men segment are interested in shoes we could adjust our message to contain ‘shoes deals’ and send it to the relevant audience and by that increase the likelihood of this segment to create a transaction.
Personalization vs Hyper-Personalization
Hyper personalization can be considered as an additional layer on top of the personalization layer, it uses a variety of different data sources, multiple channels and touchpoints with the users to create extremely personalized user experience. So if we think about our push message campaign, by analyzing the open rates of the push message from previous campaigns as a factor of the time it was sent we can optimize our campaign by sending the message in different time frames for each user.
Think AI-first. Think hyper-personalization.
Advanced Hyper personalization can only be achieved by implementing AI and machine learning models that will use internal and external data sources to create relevant products. Hyper personalization does not evaluate users as groups but individuals, it takes users data, personal information, real time information, contextual data and more to create an experience that is tailored to the need of each user.
Two key elements for AI Based hyper-personalization:
- Collect data from day one: Even if you are not able to invest resources on advanced AI processes in early stages it doesn’t mean that you shouldn’t collect the right data. Having historical data is a key component in building machine learning models, and thinking about it in early stages can have a huge impact on your product or business in the future (from technical, financial and competitive point of view).
- Collect feedback: Having historical data is a fundamental element that is needed to train machine learning models, however without feedback-loops the AI system will not know if the historical data is still relevant or not, such feedback-loops are essential for building a robust system that can adapt and flex according to the users behavior and needs.
3 examples for using AI-Based personalization
AI based personalization helps to create better user experience, with it we are able to target each user's specific interests and challenges and tailor our product to his or her needs. Here are three areas where AI-based hyper-personalization can help improve the user experience:
Gaming
In the world of social gaming, having an optimal game economy can mean success or failure, it is one of the key components that creates the users experience and contributes to their sense of achievement. Hyper personalization of the users game progression, challenges or even pricing can bring KPIs such as retention rates, game counts and transactions to increase significantly.
Ecommerce
Much like gaming, eCommerce holds great potential for Hyper personalization. As someone that personally scaled up products in both worlds I always find great lines of similarities between the two. AI-first mindset in eCommerce can come in different forms and the level of personalization can vary a lot based on the type of data you collected. Think about ‘product recommendation’ features or ‘customer centric search’ that are powered by hyper-personalization and the impact it can have on your users shopping behavior.
Marketing
It is almost natural to think about personalization in the context of marketing, hyper-personalization can become a strategic competitive advantage if you can master methodologies in retargeting, identification of high quality prospects, cross platform segmentation, optimization of paid channel ROI and more.
Takeaway
It’s really easy to associate personalization with the eCommerce world but in fact this is relevant for every business, product, marketing campaign or even in-person interaction. Understanding the two key elements behind AI based personalization and how to adjust them to fit your product will give you advantage when addressing the problem you are trying to solve and can be a key component in your ability to not only know your users better but also increase their engagement throughout the entire user journey.
