Cloud computing · Empirical machine-learning study implementing a hierarchical ensembled customer-lifetime-value model; supervised models are compared with simpler heuristics, traditional CLV approaches, and time-series methods using a random 70% training/30% test spl
A novel approach to predicting customer lifetime value in B2B SaaS companies
Journal of Marketing Analytics · 2023
Summary
The publisher reports that the hierarchical ensemble improves practical CLV prediction over simpler heuristics and traditional CLV approaches, and that the predictions can be used to optimize marketing spend and ROI and generate managerial insights.
For your marketing
Possible founder application
Use it as a B2B mechanism or sales-process design reference. Validate the relationship in your deal size, buying committee and sales cycle.
Reported result
What the study found
The publisher reports that the hierarchical ensemble improves practical CLV prediction over simpler heuristics and traditional CLV approaches, and that the predictions can be used to optimize marketing spend and ROI and generate managerial insights.
Research question
What the authors investigated
Can a hierarchical, ensembled machine-learning framework improve customer-lifetime-value prediction and marketing resource allocation in the heterogeneous, temporally constrained B2B SaaS setting?
Design
Empirical machine-learning study implementing a hierarchical ensembled customer-lifetime-value model; supervised models are compared with simpler heuristics, traditional CLV approaches, and time-series methods using a random 70% training/30% test split.
Sample and setting
Data from a major B2B SaaS provider: customers with active licenses on 1 January 2021, using revenue/MRR, product-license, product-usage, and firmographic features. The two-year CLV label covers 2021–2022 and the longer-horizon model uses a five-year window ending 2022.
Scope and transfer
Limitations to keep in view
The publisher page does not state an explicit limitations section in the fetched article. It does explicitly describe the empirical setting's challenges as heterogeneous customers, multiple products, and constrained temporal data; no broader generalizability claim is made here.
Research sources
The Springer publisher abstract identifies the B2B SaaS setting, hierarchical ensembled framework, empirical implementation on a B2B SaaS company, improvement over simpler/traditional approaches, and applications to marketing spend and ROI. The methods page specifies active-license customers, the 70/30 train-test split, and the revenue, usage, license, and firmographic data. Direct publisher-page retrieval was recorded on 2026-10-09; profile summarizes the accessible record and does not claim independent review of inaccessible full text.