Propensity Modelling

Predict customer behaviour

Predict customer behaviour

Propensity modelling helps you understand what your customers are going to do next. It enables you to analyse customers’ known features, past behaviours and purchasing history. You can then determine if they will take the action you want them to, such as make a purchase or respond to a discount.
 
By knowing what your customers are up to, you can optimise different channels in your marketing mix and increase conversions. Do you want to understand when to launch an email sequence? Find out if a prospect has budgetary constraints? Or discover which personalised promotions will retain the most active shoppers.
How we can help you

How we can help you

Our data science team can help you bring together your transaction history and customer activity data. This means you can then create a dynamic, scalable and ROI-driven propensity model. This can provide you with insights that will drive more results with less spend.

Our propensity modelling service

Scope definition and analysis

In our initial meeting, we’ll go through your business objectives and suggest the best modelling options. We can develop propensity models for first-time and repeat buyers. We can set up predictive lead scoring or create more specific models to estimate the prospect’s share of wallet. Or their likelihood to convert/churn. We can even devise a model suggesting the best discount strategies for different customers.

Data collection and preparation

Our team will schedule a deep inventory of your current data sources. We will recommend and implement the most suitable data warehousing and consolidation solutions. Then we’ll start the data cleansing process.

Roll out & monitoring

Our team will keep a close eye on the model’s early performance. We’ll ensure that new information flows in smoothly so that you can avoid any confirmation bias in the future. At this point, you will gain access to a reliable stream of insights and can take immediate action to improve your marketing campaigns.

Frequently asked questions

Propensity modelling predicts the likelihood of users, leads and customers taking specific actions. It analyses the behaviour of your audiences and compares this against the patterns of similar users. It can then predict how likely individual prospects are to buy from you. Or respond to specific types of messaging. For example: a lead nurturing campaign, a coupon offer or even suggesting a more expensive product.

Propensity modelling scores the quality of leads and customers as they progress along the consumer journey. It can predict the likelihood of leads converting into paying customers. This allows you to target low-scoring leads with campaigns designed to increase incentive. This means you can convert more of them into customers.
 
After the initial purchase, propensity modelling can predict the likelihood of future purchases, customer lifetime value and customer churn. You can use this data to recommend products that are most likely to secure the next purchase. Or take action as soon as customers show signs of churning and keep them engaged with your brand.

An agency that specialises in propensity modelling has the tools and processes to make predictions about your target audiences. Here at Vertical Leap, we’ve built a large database of user behaviour. We can use this to compare your visitors to make reliable predictions from day one.

Yes – this is the key benefit of using propensity modelling. It uses big data to inform marketing decisions so that you can respond to user needs faster and with greater accuracy. For example, when a prospect starts showing signs of interest, you can trigger remarketing or email campaigns to capture the sale before they buy from one of your rivals.
 
Likewise, you can make data-driven product recommendations to your existing customers. This will maximise repeat purchases and help you maximise customer retention.
Propensity modelling uses a wide range of datasets to predict user behaviours, including:
  • Demographics: Age, gender, location and any other accessible data that helps paint a picture of who users are.
  • Behavioural data: Such as product pages viewed, content engaged with, conversions completed, etc.
  • Purchase history: Historical patterns that can be used to make recommendations and predict future purchases.
  • Brand engagement: Such as website visits, app usage, email engagement and purchase frequency. These can be used to predict engagement and potential churn.
  • Third-party data: By comparing the data with similar audiences, we can predict actions with greater accuracy. This will help you make better marketing decisions.
With a deeper understanding of what your customers are doing, you can better optimise campaigns. And respond to their changing needs with greater speed and accuracy.