top of page

Chip

Customer LTV Model

tickl built a customer lifetime value model (LTV) for fintech Chip, accounting for multiple variables that influence the likelihood and timing of churn, enabling Chip to predict with confidence the LTV of individual customers and broader cohorts.

Project Overview

About Chip and their business relationship with tickl.

Chip is a rapidly growing and multi award-winning London-based fintech, offering savings and investment products to its customer base of over 300,000 active users, with assets under administration (AUA) exceeding £5 billion.

tickl partnered with Chip in 2025 on a cross-functional project to produce a customer LTV model that could forecast revenue / churn rates and inform strategic decision-making regarding customer acquisitions.

Tickl Decorative Image_edited.jpg

The Challenge

The challenges faced by Chip before tickl intervened.

Chip’s customer acquisition cost (CAC) had risen sharply, with conflicting theories internally as to whether acquiring customers at that higher CAC remained profitable in the long-term. Chip therefore wanted to calculate customer LTV to understand when / if individual customers and cohorts would become profitable – taking account of churn rates and the impact of compounding interest on deposits – to steer their decisions around acquisitions and retention. This challenge was compounded by the competitive nature of the environment in which Chip operates, in which there’s an antagonism between generally seeking to reduce churn rates, while simultaneously actively seeking to churn customers who it is predicted will always be loss-making as quickly (and therefore cheaply) as possible.

The Solution

We built an integrated data model by connecting Chip’s Databricks data into a semantic PowerBI model, and then layering the forecasting in two stages.

This produced expected LTV at both an individual customer and cohort level, so Chip could accurately determine profitability. It also surfaced features that predicted churn, enabling Chip to identify customers who were at a higher risk of leaving.

  • A naïve forecast that projected revenue and AUA based on customer acquisition costs while assuming no customers would churn


  • A mature forecast that applied individual churn risk – using drivers like teaser rate period remaining, tenure, number of active products, product types, and total balance – to the forecast while also handling compounding interest dynamics.

Tickl Decorative Image.avif

In the Words of The Client

TBD

TBD

Alex Latham, Chief Marketing Officer, Chip

Project Outcomes

Chip are no longer flying build, with an operational LTV model that:


  • Has identified historic mistakes that Chip can now consider when devising marketing and pricing strategy going forward.

  • Allows them to reduce the likelihood of inadvertently acquiring customers that will always be loss-making.

  • Gives Chip the ability to intervene when profitable customers are at a higher risk of leaving, thereby reducing churn rates.

Ready to make your business faster and smarter?

Speak to tickl about how we can help make your business one that embraces efficiency and makes decisions more confidently.

bottom of page