Do You Know These 4 Types of Personalized Marketing?

If you’re considering implementing a personalization system in your marketing organization or expanding upon your existing infrastructure, you should know your options. Just as one-size-fits all marketing doesn’t adequately address the needs of individual customers, a one-size-fits all approach to personalization isn’t optimal for marketers. Below, we lay out the major approaches and offer some advice for selecting the right one for you.

Prescriptive Personalization

Prescriptive personalization (also known as rules-based personalization) operates according to rules that are, at least in theory, based on logic relevant to the business. For example, a clothing retailer may know that certain customers tend to open emails or browse the website but rarely purchase online, instead preferring to purchase in store. This business may put these customers into a segment called “Window Shoppers.” A prudent rules-based personalization in this case could be to:

  1. Determine which customers qualify as a “Window Shoppers” — for example, moderate engagement but no purchase within a 60 day window.
  2. Assign them to the Window Shoppers segment.
  3. Engage these customers with emails and mobile push notifications designed to drive them to the store.

Prescriptive personalization works in so far as it accurately anticipates the needs and wants of customers in relation to those of the business. If the rules used by such a system are aligned with the goals and needs of the business and its customers, such systems can and do benefit retailers. The inverse is also true: if the rules are irrelevant to the business and customers, these systems will perform poorly.

Recommender Engines

In contrast to prescriptive personalization systems, which rely on marketers to figure out how their outreach should respond to the behavior of individual customers, recommender engines are automated systems. They determine which product recommendations should be served to which customers according to their own internal programming without the need for manual intervention–although many recommender engines do allow marketers to input business rules. There are three main types of recommender systems:

Collaborative Filtering
You probably know these systems by the descriptions they use to present themselves to users: “customers who bought this also bought that” or “customers who viewed this also viewed that”.

Systems that use the collaborative filtering method gather and analyze large amounts of users’ preferences to determine which items a given user will like based on the behavior of users similar to them in some important way, such as purchasing or browsing history.

Content Filtering
Content filtering systems compare items against individual user preferences. For example, such a system might “see” that a user tends to purchase neckties online more than he purchases shoes; therefore, it would recommend neckties to him rather than shoes.

Hybrid Systems
Hybrid recommender engines combine the approaches of both collaborative filtering and content filtering to make product recommendations, looking at both the individual user in relation to items as well as the user in relation to other users.

Real Time Personalization

Real time personalization, as its name suggests, personalizes the user experience in real time. This could be something as simple as a popup message saying “Welcome back [First Name]” or as complex as adapting the color scheme of the site to match each user’s particular tastes in an effort to get that user to engage more. Many real-time systems also provide product recommendations, which calculate their recommendations using the methods described above.

Real time personalization draws on both real time data, such as time of day, geographic location, device, in addition to historical data like purchase history or time since last purchase.

Customer Journey Mapping and Optimization

A customer journey refers to the series of interactions a customer takes with a brand over time. At Coherent Path, we define the customer journey as the trajectory along which the customer moves within a brand’s product, content and transactional space. Items and content that intersect with that trajectory are those which the customer is on a path to consume.

There are a few different approaches to mapping customer journeys. One method is similar to the prescriptive personalization discussed above. It involves marketers relying on their knowledge of the customer and their business to aggregate customer behavior into various personas and create generalized models of the customer journey. For example, a workflow might regard people who have clicked on items but not purchased as ‘explorers’ who need more information about an item or set of items before they commit to a purchase. This workflow could send such customers an email providing more information (if the email of the customer in question is known).

Another method of customer journey mapping and optimization is to use machine learning and data science to model each individual journey. This is what we do at Coherent Path. We believe that this data-driven, machine-learned approach is superior to the method described above because it allows customer journeys to be mapped as they actually are rather than what marketers reckon they are.

Moreover, it allows you to map each individual, unique journey rather than rely on a model which generalizes customer behaviour. Generalized models can overlook important variation among individuals whereas personalized journey mapping enables brands to identify the products most likely to resonate with each person.

Another benefit of customer journey optimization is that it accounts for a more comprehensive range of engagement between customers and brands than do recommender engines, which tend to focus only on the next step. The ability to focus on the long term enables brands to drive progress towards strategic objectives such as growing lifetime value and earning loyalty by aligning outreach with the trajectories of each customer over time.

Recommender Engine vs Customer Journey Optimization - small grey text

So Which Type of Personalization Is Best for You?

Working with our clients, we’ve found that many marketing organizations benefit from using multiple approaches rather than a single one. Product recommendations can work to boost near term conversions. If you’re running a small operation with only a few thousand customers whose behavior is already well known then a rules-based system could be sufficient. If you want a website or app that will adapt to the tastes of individual customers, real-time personalization is worth looking into. If you want to understand where customers are heading and how to strengthen your relationship with them, customer journey optimization is invaluable.

Case Study: Automated Customer Journey Optimization Delivers 30% Lift