This article will review the Top Spenders model and attributes to help you discover your most valuable customers ahead of time based on deposit/purchase amount and net revenue.
Identifying your future top spenders will help you strategically allocate your marketing spend and optimize promotion generosity to improve profitability.
Introduction
Nurturing customers that spend the most money with you will help to drive loyalty and generate higher revenue. Using our Top Spender Predictive Model, you can now take the guesswork out of your marketing activities when deciding how to reward high-value customers.
Top Spenders are classified as customers in the top 5% of your customer base in terms of deposit/purchase amount and net revenue.
With the Top Spenders attributes, you will be able to:
- Target customers based on their probability of becoming top spenders
- Identify your potential VIP customers
- Allocate your promotional budget in the most optimal way
How to Use
The Top Spenders model can be used to define and explore target groups.
- Start creating a Target Group from Customer Explorer or the Target Groups page.
- When selecting your Target Group criteria, click on the ‘Top Spenders’ criterion.
- Select your prediction categories from Highest to Lowest
The Top Spenders attribute uses a machine learning model to assign customers a prediction score between 0-100 to represent how likely they are to become a Top Spender. Optimove then helps you segment customers based their prediction score with 5 categories: Highest, High, Medium, Low, and Lowest.
You can choose more than one category. For example, select the Highest and High categories to segment customers who are likely to become top spenders.
Understanding Prediction Categories
When calculating a customer’s probability to become a Top Spender, they will receive a prediction score from 1-100. The higher the score, the higher the likelihood the customer will become a Top Spender.
Customers are then split based on percentiles which ranks each customer against each other based on their probability score. For example, if you are on the 90th percentile, you ranked higher than 90% of customers in terms of your likelihood to become a top spender.
Each category from Highest – Lowest represents a percentile:
- Highest (Top 1%): This category includes customers who fall in the 99th percentile, meaning they rank higher than 99% of all customers.
- High (Next 9%): Customers in this group are in the 90th to 98th percentiles. The Highest and High categories therefore represent the top 10% of customers.
- Medium (Next 20%): This segment includes customers in the 70th to 89th percentiles. These customers rank higher than 70% to 89% of other customers
- Low (Next 20%): Customers in this category fall within the 50th to 69th percentiles, ranking higher than 50% to 69% of other customers
- Lowest (Bottom 50%): This group comprises customers in the 1st to 49th percentiles. They rank lower than 50% of all customers and have the least likelihood of becoming top spenders.
You can then see the prediction score of customers in the category. In the example below, customers in the Highest group have a prediction score of 32-88% and there are 12 customers in this category.
To create your own customized prediction categories, use the Becoming Top Spender Score or Rank in Becoming Top Spender Customer Attributes.
Additional Attributes
You can also find four attributes related to the Top Spenders model in the ‘Customer Attributes’ dropdown, each with different criteria.
Is Top Spender
- “Is a Top Spender = Yes” to segment top spenders
- "Is a Top Spender = No” to segment customers that are not top spenders
Becoming Top Spender Score
- “Becoming Top Spender Score >= 80%” will segment customers with a likelihood to become a top spender of 80% or above.
Rank in Becoming Top Spender
- “Rank in Becoming Top Spender = 100” will segment customers in the top 1% meaning they have the highest likelihood to become a Top Spender
- “Rank in Becoming Top Spender >= 80” will segment customers in the top 20% in terms of their likelihood to become a Top Spender
Rank by LCS Becoming Top Spender
When using “Rank in Becoming Top Spender”, you may find that all customers in the top 1% are in the Active lifecycle stage. Therefore, you can use this attribute to make sure you are targeting customers that are most likely to become Top Spenders within their respective lifecycle stages.
- “Rank by LCS Becoming Top Spender = 100” will segment customers in the top 1% within the ‘New’ lifecycle stage, top 1% within the ‘Active’ lifecycle stage, top 1% within the ‘Churn’ lifecycle stage and so forth.
Use Cases
Using the Top Spender predictive attribute, you can easily identify which customers will deliver the highest ROI. This insight can help you to effectively allocate your marketing budget by segmenting customers based on their likelihood of becoming a top spender and personalizing your campaigns accordingly.
For example, using the Top Spender predictive attribute, you can:
- Identify your potential VIP customers. Defining which customers are considered VIPs can be a difficult task with complex rules and calculations. By integrating our Top Spender model with your VIP strategy, you can easily identify your most valuable customers early on and target them with campaigns enticing them to become VIPs. This way you can ensure a long-term and beneficial relationship with high-value customers.
- Offer promotions on a sliding scale. Send customers that have a lower likelihood of becoming a top spender with special promo codes to encourage them to spend more money with your brand. Meanwhile, send customers with a higher likelihood of becoming a top spender lower promotions since they need less encouragement to spend.
Model Calculation
Optimove classifies Top Spender as customers in the top 5% of your customer base in terms of deposit/purchase amount and net revenue.
Please Note: if you do not have the Net Revenue metric, we will calculate Top Spenders only by deposit/purchase amount.
The Top Spender model is an AI-generated predictive model that is measured for every customer in your database in all lifecycle stages. Our Machine Learning models analyze behaviors, attributes, or signals of top spending customers and use this data as an input to the prediction model. Our models look at every relevant data point, from frequency and recency of visits/purchases/deposits, monetary values (purchase/deposit amount), as well as demographic information. As a result, our Machine Learning can analyze the behaviors that led to a customer becoming a Top Spender and identify customers that mirror these behaviors.
The model then predicts a customer’s probability of becoming a Top Spender within the next 3-6 months and assigns every customer a prediction score between 1-100.
If your period length is 14 days, the model will predict a customer’s probability of becoming a Top Spender within the next 3 months. If your period length is 30 days, the model will predict a customer’s probability of becoming a Top Spender within the next 6 months.